tahoe-lafs/contrib/fuse/runtests.py

909 lines
32 KiB
Python
Raw Normal View History

#! /usr/bin/env python
'''
Unit and system tests for tahoe-fuse.
'''
# Note: It's always a SetupFailure, not a TestFailure if a webapi
# operation fails, because this does not indicate a fuse interface
# failure.
# TODO: Unmount after tests regardless of failure or success!
# TODO: Test mismatches between tahoe and fuse/posix. What about nodes
# with crazy names ('\0', unicode, '/', '..')? Huuuuge files?
# Huuuuge directories... As tahoe approaches production quality, it'd
# be nice if the fuse interface did so also by hardening against such cases.
# FIXME: Only create / launch necessary nodes. Do we still need an introducer and three nodes?
# FIXME: This framework might be replaceable with twisted.trial,
# especially the "layer" design, which is a bit cumbersome when
# using recursion to manage multiple clients.
# FIXME: Identify all race conditions (hint: starting clients, versus
# using the grid fs).
import sys, os, shutil, unittest, subprocess
import tempfile, re, time, random, httplib, urllib
#import traceback
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
from twisted.python import usage
if sys.platform.startswith('darwin'):
UNMOUNT_CMD = ['umount']
else:
# linux, and until we hear otherwise, all other platforms with fuse, by assumption
UNMOUNT_CMD = ['fusermount', '-u']
# Import fuse implementations:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
#FuseDir = os.path.join('.', 'contrib', 'fuse')
#if not os.path.isdir(FuseDir):
# raise SystemExit('''
#Could not find directory "%s". Please run this script from the tahoe
#source base directory.
#''' % (FuseDir,))
FuseDir = '.'
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
### Load each implementation
sys.path.append(os.path.join(FuseDir, 'impl_a'))
import tahoe_fuse as impl_a
sys.path.append(os.path.join(FuseDir, 'impl_b'))
import pyfuse.tahoe as impl_b
fuse/impl_c: move mac tahoefuse impl out into contrib/fuse For a variety of reasons, high amongst them the fact that many people interested in fuse support for tahoe seem to have missed its existence, the existing fuse implementation for tahoe, previously 'mac/tahoefuse.py' has been renamed and moved. It was suggested that, even though the mac build depends upon it, that the mac/tahoefuse implementation be moved into contrib/fuse along with the other fuse implementations. The fact that it's not as extensively covered by unit tests as mainline tahoe was given as corroboration. In a bid to try and stem the confusion inherent in having tahoe_fuse, tfuse and tahoefuse jumbled together (not necessarily helped by referring to them as impl_a, b and c respectively) I'm hereby renaming tahoefuse as 'blackmatch' (black match is, per wikipedia "a type of crude fuse" hey, I'm a punny guy) Maybe one day it'll be promoted to be 'quickmatch' instead... Anyway, this patch moves mac/tahoefuse.py out to contrib/fuse/impl_c/ as blackmatch.py, and makes appropriate changes to the mac build process to transclude blackmatch therein. this leaves the extant fuse.py and fuseparts business in mac/ as-is and doesn't attempt to address such issues in contrib/fuse/impl_c. it is left as an exercise to the reader (or the reader of a message to follow) as to how to deal with the 'fuse' python module on the mac. as of this time, blackmatch should work on both mac and linux, and passes the four extant tests in runtests. (fwiw neither impl_a nor impl_b have I managed to get working on the mac yet) since blackmatch supports a read-write and caching fuse interface to tahoe, some write tests obviously need to be added to runtests.
2008-09-25 01:42:14 +00:00
sys.path.append(os.path.join(FuseDir, 'impl_c'))
import blackmatch as impl_c
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
### config info about each impl, including which make sense to run
implementations = {
'impl_a': dict(module=impl_a,
mount_args=['--basedir', '%(nodedir)s', '%(mountpath)s', ],
mount_wait=True,
suites=['read', ]),
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
'impl_b': dict(module=impl_b,
todo=True,
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
mount_args=['--basedir', '%(nodedir)s', '%(mountpath)s', ],
mount_wait=False,
suites=['read', ]),
fuse/impl_c: move mac tahoefuse impl out into contrib/fuse For a variety of reasons, high amongst them the fact that many people interested in fuse support for tahoe seem to have missed its existence, the existing fuse implementation for tahoe, previously 'mac/tahoefuse.py' has been renamed and moved. It was suggested that, even though the mac build depends upon it, that the mac/tahoefuse implementation be moved into contrib/fuse along with the other fuse implementations. The fact that it's not as extensively covered by unit tests as mainline tahoe was given as corroboration. In a bid to try and stem the confusion inherent in having tahoe_fuse, tfuse and tahoefuse jumbled together (not necessarily helped by referring to them as impl_a, b and c respectively) I'm hereby renaming tahoefuse as 'blackmatch' (black match is, per wikipedia "a type of crude fuse" hey, I'm a punny guy) Maybe one day it'll be promoted to be 'quickmatch' instead... Anyway, this patch moves mac/tahoefuse.py out to contrib/fuse/impl_c/ as blackmatch.py, and makes appropriate changes to the mac build process to transclude blackmatch therein. this leaves the extant fuse.py and fuseparts business in mac/ as-is and doesn't attempt to address such issues in contrib/fuse/impl_c. it is left as an exercise to the reader (or the reader of a message to follow) as to how to deal with the 'fuse' python module on the mac. as of this time, blackmatch should work on both mac and linux, and passes the four extant tests in runtests. (fwiw neither impl_a nor impl_b have I managed to get working on the mac yet) since blackmatch supports a read-write and caching fuse interface to tahoe, some write tests obviously need to be added to runtests.
2008-09-25 01:42:14 +00:00
'impl_c': dict(module=impl_c,
mount_args=['--cache-timeout', '0', '--root-uri', '%(root-uri)s',
'--node-directory', '%(nodedir)s', '%(mountpath)s', ],
mount_wait=True,
suites=['read', 'write', ]),
fuse/blackmatch: split into client/server (twisted server) This implements a client/server split for blackmatch, where the client implements the fuse_main bindings and a simple blocking rpc client mechanism. The server implements the other half of that rpc mechanism, and contains all the actual logic for interpreting fuse requests in the context of the on disk cache and requests to the tahoe node. The server is based on a twisted reactor. The rpc mechanism implements a simple method dispatch including marshalling, using json, of basic inert data types, in a flat namespace (no objects). The client side is written in a blocking idiom, to interface with the threading model used by the fuse_main bindings, whereas the server side is written for a twisted reactor-based environment, intended to facilitate implementing more sophisticated logic in that paradigm. The two communicate over a unix domain socket, allocated within the nodedir. Command line usage is unchanged; the server is launched automatically by the client. The server daemonizes itself, to avoid preventing the original parent process (e.g. 'runtests') from waiting upon the server exiting. The client keeps open a 'keepalive' connection to the server; upon loss thereof the server will exit. This addresses the fact that the python-fuse bindings provide no notification of exit of the client process upon unmount. The client thus provides a relatively thin 'shim' proxying requests from the fuse_main bindings across the rpc to the server process, which handles the logic behind each request. For the time being, a '--no-split' option is provided to surpress the splitting into client/server, yielding the prior behaviour. Once the server logic gets more complex and more entrenched in a twisted idiom, this might be removed. The 'runtests' test harness currently tests both modes, as 'impl_c' and 'impl_c_no_split'
2008-10-16 15:08:46 +00:00
'impl_c_no_split': dict(module=impl_c,
mount_args=['--cache-timeout', '0', '--root-uri', '%(root-uri)s',
'--no-split',
'--node-directory', '%(nodedir)s', '%(mountpath)s', ],
mount_wait=True,
suites=['read', 'write', ]),
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
}
if sys.platform == 'darwin':
del implementations['impl_a']
del implementations['impl_b']
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
default_catch_up_pause = 0
if sys.platform == 'linux2':
default_catch_up_pause = 2
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
class FuseTestsOptions(usage.Options):
optParameters = [
["test-type", None, "both",
"Type of test to run; unit, system or both"
],
["implementations", None, "all",
"Comma separated list of implementations to test, or 'all'"
],
["suites", None, "all",
"Comma separated list of test suites to run, or 'all'"
],
["tests", None, None,
"Comma separated list of specific tests to run"
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
],
["path-to-tahoe", None, "../../bin/tahoe",
"Which 'tahoe' script to use to create test nodes"],
["tmp-dir", None, "/tmp",
"Where the test should create temporary files"],
# Note; this is '/tmp' because on leopard, tempfile.mkdtemp creates
# directories in a location which leads paths to exceed what macfuse
# can handle without leaking un-umount-able fuse processes.
["catch-up-pause", None, str(default_catch_up_pause),
"Pause between tahoe operations and fuse tests thereon"],
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
]
optFlags = [
["debug-wait", None,
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
"Causes the test system to pause at various points, to facilitate debugging"],
["web-open", None,
"Opens a web browser to the web ui at the start of each impl's tests"],
["no-cleanup", False,
"Prevents the cleanup of the working directories, to allow analysis thereof"],
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
]
def postOptions(self):
if self['suites'] == 'all':
self.suites = ['read', 'write']
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
# [ ] todo: deduce this from looking for test_ in dir(self)
else:
self.suites = map(str.strip, self['suites'].split(','))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
if self['implementations'] == 'all':
self.implementations = implementations.keys()
else:
self.implementations = map(str.strip, self['implementations'].split(','))
if self['tests']:
self.tests = map(str.strip, self['tests'].split(','))
else:
self.tests = None
self.catch_up_pause = float(self['catch-up-pause'])
### Main flow control:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def main(args):
config = FuseTestsOptions()
config.parseOptions(args[1:])
target = 'all'
if len(args) > 1:
target = args.pop(1)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
test_type = config['test-type']
if test_type not in ('both', 'unit', 'system'):
raise usage.error('test-type %r not supported' % (test_type,))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
if test_type in ('both', 'unit'):
run_unit_tests([args[0]])
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
if test_type in ('both', 'system'):
return run_system_test(config)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def run_unit_tests(argv):
print 'Running Unit Tests.'
try:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
unittest.main(argv=argv)
except SystemExit, se:
pass
print 'Unit Tests complete.\n'
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def run_system_test(config):
return SystemTest(config).run()
def drepr(obj):
r = repr(obj)
if len(r) > 200:
return '%s ... %s [%d]' % (r[:100], r[-100:], len(r))
else:
return r
### System Testing:
class SystemTest (object):
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def __init__(self, config):
self.config = config
# These members represent test state:
self.cliexec = None
self.testroot = None
# This test state is specific to the first client:
self.port = None
self.clientbase = None
## Top-level flow control:
# These "*_layer" methods call each other in a linear fashion, using
# exception unwinding to do cleanup properly. Each "layer" invokes
# a deeper layer, and each layer does its own cleanup upon exit.
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def run(self):
print '\n*** Setting up system tests.'
try:
results = self.init_cli_layer()
print '\n*** System Tests complete:'
total_failures = todo_failures = 0
for result in results:
impl_name, failures, total = result
if implementations[impl_name].get('todo'):
todo_failures += failures
else:
total_failures += failures
print 'Implementation %s: %d failed out of %d.' % result
if total_failures:
print '%s total failures, %s todo' % (total_failures, todo_failures)
return 1
else:
return 0
except SetupFailure, sfail:
print
print sfail
print '\n*** System Tests were not successfully completed.'
return 1
def maybe_wait(self, msg='waiting', or_if_webopen=False):
if self.config['debug-wait'] or or_if_webopen and self.config['web-open']:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
print msg
raw_input()
def maybe_webopen(self, where=None):
if self.config['web-open']:
import webbrowser
url = self.weburl
if where is not None:
url += urllib.quote(where)
webbrowser.open(url)
def maybe_pause(self):
time.sleep(self.config.catch_up_pause)
def init_cli_layer(self):
'''This layer finds the appropriate tahoe executable.'''
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
#self.cliexec = os.path.join('.', 'bin', 'tahoe')
self.cliexec = self.config['path-to-tahoe']
version = self.run_tahoe('--version')
print 'Using %r with version:\n%s' % (self.cliexec, version.rstrip())
return self.create_testroot_layer()
def create_testroot_layer(self):
print 'Creating test base directory.'
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
#self.testroot = tempfile.mkdtemp(prefix='tahoe_fuse_test_')
#self.testroot = tempfile.mkdtemp(prefix='tahoe_fuse_test_', dir='/tmp/')
tmpdir = self.config['tmp-dir']
if tmpdir:
self.testroot = tempfile.mkdtemp(prefix='tahoe_fuse_test_', dir=tmpdir)
else:
self.testroot = tempfile.mkdtemp(prefix='tahoe_fuse_test_')
try:
return self.launch_introducer_layer()
finally:
if not self.config['no-cleanup']:
print 'Cleaning up test root directory.'
try:
shutil.rmtree(self.testroot)
except Exception, e:
print 'Exception removing test root directory: %r' % (self.testroot, )
print 'Ignoring cleanup exception: %r' % (e,)
else:
print 'Leaving test root directory: %r' % (self.testroot, )
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def launch_introducer_layer(self):
print 'Launching introducer.'
introbase = os.path.join(self.testroot, 'introducer')
# NOTE: We assume if tahoe exits with non-zero status, no separate
# tahoe child process is still running.
createoutput = self.run_tahoe('create-introducer', '--basedir', introbase)
self.check_tahoe_output(createoutput, ExpectedCreationOutput, introbase)
startoutput = self.run_tahoe('start', '--basedir', introbase)
try:
self.check_tahoe_output(startoutput, ExpectedStartOutput, introbase)
return self.launch_clients_layer(introbase)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
finally:
print 'Stopping introducer node.'
self.stop_node(introbase)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def set_tahoe_option(self, base, key, value):
import re
filename = os.path.join(base, 'tahoe.cfg')
content = open(filename).read()
content = re.sub('%s = (.+)' % key, '%s = %s' % (key, value), content)
open(filename, 'w').write(content)
TotalClientsNeeded = 3
def launch_clients_layer(self, introbase, clientnum = 0):
if clientnum >= self.TotalClientsNeeded:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
self.maybe_wait('waiting (launched clients)')
ret = self.create_test_dirnode_layer()
self.maybe_wait('waiting (ran tests)', or_if_webopen=True)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
return ret
tmpl = 'Launching client %d of %d.'
print tmpl % (clientnum,
self.TotalClientsNeeded)
base = os.path.join(self.testroot, 'client_%d' % (clientnum,))
output = self.run_tahoe('create-node', '--basedir', base)
self.check_tahoe_output(output, ExpectedCreationOutput, base)
if clientnum == 0:
# The first client is special:
self.clientbase = base
self.port = random.randrange(1024, 2**15)
self.set_tahoe_option(base, 'web.port', 'tcp:%d:interface=127.0.0.1' % self.port)
self.weburl = "http://127.0.0.1:%d/" % (self.port,)
print self.weburl
else:
self.set_tahoe_option(base, 'web.port', '')
introfurl = os.path.join(introbase, 'introducer.furl')
furl = open(introfurl).read().strip()
self.set_tahoe_option(base, 'introducer.furl', furl)
# NOTE: We assume if tahoe exist with non-zero status, no separate
# tahoe child process is still running.
startoutput = self.run_tahoe('start', '--basedir', base)
try:
self.check_tahoe_output(startoutput, ExpectedStartOutput, base)
return self.launch_clients_layer(introbase, clientnum+1)
finally:
print 'Stopping client node %d.' % (clientnum,)
self.stop_node(base)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def create_test_dirnode_layer(self):
print 'Creating test dirnode.'
cap = self.create_dirnode()
f = open(os.path.join(self.clientbase, 'private', 'root_dir.cap'), 'w')
f.write(cap)
f.close()
return self.mount_fuse_layer(cap)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def mount_fuse_layer(self, root_uri):
mpbase = os.path.join(self.testroot, 'mountpoint')
os.mkdir(mpbase)
results = []
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
if self.config['debug-wait']:
ImplProcessManager.debug_wait = True
#for name, kwargs in implementations.items():
for name in self.config.implementations:
kwargs = implementations[name]
#print 'instantiating %s: %r' % (name, kwargs)
implprocmgr = ImplProcessManager(name, **kwargs)
print '\n*** Testing impl: %r' % (implprocmgr.name)
implprocmgr.configure(self.clientbase, mpbase)
implprocmgr.mount()
try:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
failures, total = self.run_test_layer(root_uri, implprocmgr)
result = (implprocmgr.name, failures, total)
tmpl = '\n*** Test Results implementation %s: %d failed out of %d.'
print tmpl % result
results.append(result)
finally:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
implprocmgr.umount()
return results
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def run_test_layer(self, root_uri, iman):
self.maybe_webopen('uri/'+root_uri)
failures = 0
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
testnum = 0
numtests = 0
if self.config.tests:
tests = self.config.tests
else:
tests = list(set(self.config.suites).intersection(set(iman.suites)))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
self.maybe_wait('waiting (about to run tests)')
for test in tests:
testnames = [n for n in sorted(dir(self)) if n.startswith('test_'+test)]
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
numtests += len(testnames)
print 'running %s %r tests' % (len(testnames), test,)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
for testname in testnames:
testnum += 1
print '\n*** Running test #%d: %s' % (testnum, testname)
try:
testcap = self.create_dirnode()
dirname = '%s_%s' % (iman.name, testname)
self.attach_node(root_uri, testcap, dirname)
method = getattr(self, testname)
method(testcap, testdir = os.path.join(iman.mountpath, dirname))
print 'Test succeeded.'
except TestFailure, f:
print f
#print traceback.format_exc()
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
failures += 1
except:
print 'Error in test code... Cleaning up.'
raise
return (failures, numtests)
# Tests:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_read_directory_existence(self, testcap, testdir):
if not wrap_os_error(os.path.isdir, testdir):
raise TestFailure('Attached test directory not found: %r', testdir)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_read_empty_directory_listing(self, testcap, testdir):
listing = wrap_os_error(os.listdir, testdir)
if listing:
raise TestFailure('Expected empty directory, found: %r', listing)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_read_directory_listing(self, testcap, testdir):
names = []
filesizes = {}
for i in range(3):
fname = 'file_%d' % (i,)
names.append(fname)
body = 'Hello World #%d!' % (i,)
filesizes[fname] = len(body)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
cap = self.webapi_call('PUT', '/uri', body)
self.attach_node(testcap, cap, fname)
dname = 'dir_%d' % (i,)
names.append(dname)
cap = self.create_dirnode()
self.attach_node(testcap, cap, dname)
names.sort()
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
listing = wrap_os_error(os.listdir, testdir)
listing.sort()
if listing != names:
tmpl = 'Expected directory list containing %r but fuse gave %r'
raise TestFailure(tmpl, names, listing)
for file, size in filesizes.items():
st = wrap_os_error(os.stat, os.path.join(testdir, file))
if st.st_size != size:
tmpl = 'Expected %r size of %r but fuse returned %r'
raise TestFailure(tmpl, file, size, st.st_size)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_read_file_contents(self, testcap, testdir):
name = 'hw.txt'
body = 'Hello World!'
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
cap = self.webapi_call('PUT', '/uri', body)
self.attach_node(testcap, cap, name)
path = os.path.join(testdir, name)
try:
found = open(path, 'r').read()
except Exception, err:
tmpl = 'Could not read file contents of %r: %r'
raise TestFailure(tmpl, path, err)
if found != body:
tmpl = 'Expected file contents %r but found %r'
raise TestFailure(tmpl, body, found)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_read_in_random_order(self, testcap, testdir):
sz = 2**20
bs = 2**10
assert(sz % bs == 0)
name = 'random_read_order'
body = os.urandom(sz)
cap = self.webapi_call('PUT', '/uri', body)
self.attach_node(testcap, cap, name)
# XXX this should also do a test where sz%bs != 0, so that it correctly tests
# the edge case where the last read is a 'short' block
path = os.path.join(testdir, name)
try:
fsize = os.path.getsize(path)
if fsize != len(body):
tmpl = 'Expected file size %s but found %s'
raise TestFailure(tmpl, len(body), fsize)
except Exception, err:
tmpl = 'Could not read file size for %r: %r'
raise TestFailure(tmpl, path, err)
try:
f = open(path, 'r')
posns = range(0,sz,bs)
random.shuffle(posns)
data = [None] * (sz/bs)
for p in posns:
f.seek(p)
data[p/bs] = f.read(bs)
found = ''.join(data)
except Exception, err:
tmpl = 'Could not read file %r: %r'
raise TestFailure(tmpl, path, err)
if found != body:
tmpl = 'Expected file contents %s but found %s'
raise TestFailure(tmpl, drepr(body), drepr(found))
def get_file(self, dircap, path):
body = self.webapi_call('GET', '/uri/%s/%s' % (dircap, path))
return body
def test_write_tiny_file(self, testcap, testdir):
self._write_test_linear(testcap, testdir, name='tiny.junk', bs=2**9, sz=2**9)
def test_write_linear_small_writes(self, testcap, testdir):
self._write_test_linear(testcap, testdir, name='large_linear.junk', bs=2**9, sz=2**20)
def test_write_linear_large_writes(self, testcap, testdir):
# at least on the mac, large io block sizes are reduced to 64k writes through fuse
self._write_test_linear(testcap, testdir, name='small_linear.junk', bs=2**18, sz=2**20)
def _write_test_linear(self, testcap, testdir, name, bs, sz):
body = os.urandom(sz)
try:
path = os.path.join(testdir, name)
f = file(path, 'w')
except Exception, err:
tmpl = 'Could not open file for write at %r: %r'
raise TestFailure(tmpl, path, err)
try:
for posn in range(0,sz,bs):
f.write(body[posn:posn+bs])
f.close()
except Exception, err:
tmpl = 'Could not write to file %r: %r'
raise TestFailure(tmpl, path, err)
self.maybe_pause()
self._check_write(testcap, name, body)
def _check_write(self, testcap, name, expected_body):
uploaded_body = self.get_file(testcap, name)
if uploaded_body != expected_body:
tmpl = 'Expected file contents %s but found %s'
raise TestFailure(tmpl, drepr(expected_body), drepr(uploaded_body))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def test_write_overlapping_small_writes(self, testcap, testdir):
self._write_test_overlap(testcap, testdir, name='large_overlap', bs=2**9, sz=2**20)
def test_write_overlapping_large_writes(self, testcap, testdir):
self._write_test_overlap(testcap, testdir, name='small_overlap', bs=2**18, sz=2**20)
def _write_test_overlap(self, testcap, testdir, name, bs, sz):
body = os.urandom(sz)
try:
path = os.path.join(testdir, name)
f = file(path, 'w')
except Exception, err:
tmpl = 'Could not open file for write at %r: %r'
raise TestFailure(tmpl, path, err)
try:
for posn in range(0,sz,bs):
start = max(0, posn-bs)
end = min(sz, posn+bs)
f.seek(start)
f.write(body[start:end])
f.close()
except Exception, err:
tmpl = 'Could not write to file %r: %r'
raise TestFailure(tmpl, path, err)
self.maybe_pause()
self._check_write(testcap, name, body)
def test_write_random_scatter(self, testcap, testdir):
sz = 2**20
name = 'random_scatter'
body = os.urandom(sz)
def rsize(sz=sz):
return min(int(random.paretovariate(.25)), sz/12)
# first chop up whole file into random sized chunks
slices = []
posn = 0
while posn < sz:
size = rsize()
slices.append( (posn, body[posn:posn+size]) )
posn += size
random.shuffle(slices) # and randomise their order
try:
path = os.path.join(testdir, name)
f = file(path, 'w')
except Exception, err:
tmpl = 'Could not open file for write at %r: %r'
raise TestFailure(tmpl, path, err)
try:
# write all slices: we hence know entire file is ultimately written
# write random excerpts: this provides for mixed and varied overlaps
for posn,slice in slices:
f.seek(posn)
f.write(slice)
rposn = random.randint(0,sz)
f.seek(rposn)
f.write(body[rposn:rposn+rsize()])
f.close()
except Exception, err:
tmpl = 'Could not write to file %r: %r'
raise TestFailure(tmpl, path, err)
self.maybe_pause()
self._check_write(testcap, name, body)
def test_write_partial_overwrite(self, testcap, testdir):
name = 'partial_overwrite'
body = '_'*132
overwrite = '^'*8
position = 26
def write_file(path, mode, contents, position=None):
try:
f = file(path, mode)
if position is not None:
f.seek(position)
f.write(contents)
f.close()
except Exception, err:
tmpl = 'Could not write to file %r: %r'
raise TestFailure(tmpl, path, err)
def read_file(path):
try:
f = file(path, 'rb')
contents = f.read()
f.close()
except Exception, err:
tmpl = 'Could not read file %r: %r'
raise TestFailure(tmpl, path, err)
return contents
path = os.path.join(testdir, name)
#write_file(path, 'w', body)
cap = self.webapi_call('PUT', '/uri', body)
self.attach_node(testcap, cap, name)
self.maybe_pause()
contents = read_file(path)
if contents != body:
raise TestFailure('File contents mismatch (%r) %r v.s. %r', path, contents, body)
write_file(path, 'r+', overwrite, position)
contents = read_file(path)
expected = body[:position] + overwrite + body[position+len(overwrite):]
if contents != expected:
raise TestFailure('File contents mismatch (%r) %r v.s. %r', path, contents, expected)
# Utilities:
def run_tahoe(self, *args):
realargs = ('tahoe',) + args
status, output = gather_output(realargs, executable=self.cliexec)
if status != 0:
tmpl = 'The tahoe cli exited with nonzero status.\n'
tmpl += 'Executable: %r\n'
tmpl += 'Command arguments: %r\n'
tmpl += 'Exit status: %r\n'
tmpl += 'Output:\n%s\n[End of tahoe output.]\n'
raise SetupFailure(tmpl,
self.cliexec,
realargs,
status,
output)
return output
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def check_tahoe_output(self, output, expected, expdir):
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
ignorable_lines = map(re.compile, [
'.*site-packages/zope\.interface.*\.egg/zope/__init__.py:3: UserWarning: Module twisted was already imported from .*egg is being added to sys.path',
' import pkg_resources',
])
def ignore_line(line):
for ignorable_line in ignorable_lines:
if ignorable_line.match(line):
return True
else:
return False
output = '\n'.join( [ line
for line in output.split('\n')+['']
#if line not in ignorable_lines ] )
if not ignore_line(line) ] )
m = re.match(expected, output, re.M)
if m is None:
tmpl = 'The output of tahoe did not match the expectation:\n'
tmpl += 'Expected regex: %s\n'
tmpl += 'Actual output: %r\n'
self.warn(tmpl, expected, output)
elif expdir != m.group('path'):
tmpl = 'The output of tahoe refers to an unexpected directory:\n'
tmpl += 'Expected directory: %r\n'
tmpl += 'Actual directory: %r\n'
self.warn(tmpl, expdir, m.group(1))
def stop_node(self, basedir):
try:
self.run_tahoe('stop', '--basedir', basedir)
except Exception, e:
print 'Failed to stop tahoe node.'
print 'Ignoring cleanup exception:'
# Indent the exception description:
desc = str(e).rstrip()
print ' ' + desc.replace('\n', '\n ')
def webapi_call(self, method, path, body=None, **options):
if options:
path = path + '?' + ('&'.join(['%s=%s' % kv for kv in options.items()]))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
conn = httplib.HTTPConnection('127.0.0.1', self.port)
conn.request(method, path, body = body)
resp = conn.getresponse()
if resp.status != 200:
tmpl = 'A webapi operation failed.\n'
tmpl += 'Request: %r %r\n'
tmpl += 'Body:\n%s\n'
tmpl += 'Response:\nStatus %r\nBody:\n%s'
raise SetupFailure(tmpl,
method, path,
body or '',
resp.status, body)
return resp.read()
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def create_dirnode(self):
return self.webapi_call('PUT', '/uri', t='mkdir').strip()
def attach_node(self, dircap, childcap, childname):
body = self.webapi_call('PUT',
'/uri/%s/%s' % (dircap, childname),
body = childcap,
t = 'uri',
replace = 'false')
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
assert body.strip() == childcap, `body, dircap, childcap, childname`
def polling_operation(self, operation, polldesc, timeout = 10.0, pollinterval = 0.2):
totaltime = timeout # Fudging for edge-case SetupFailure description...
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
totalattempts = int(timeout / pollinterval)
starttime = time.time()
for attempt in range(totalattempts):
opstart = time.time()
try:
result = operation()
except KeyboardInterrupt, e:
raise
except Exception, e:
result = False
totaltime = time.time() - starttime
if result is not False:
#tmpl = '(Polling took over %.2f seconds.)'
#print tmpl % (totaltime,)
return result
elif totaltime > timeout:
break
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
else:
opdelay = time.time() - opstart
realinterval = max(0., pollinterval - opdelay)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
#tmpl = '(Poll attempt %d failed after %.2f seconds, sleeping %.2f seconds.)'
#print tmpl % (attempt+1, opdelay, realinterval)
time.sleep(realinterval)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
tmpl = 'Timeout while polling for: %s\n'
tmpl += 'Waited %.2f seconds (%d polls).'
raise SetupFailure(tmpl, polldesc, totaltime, attempt+1)
def warn(self, tmpl, *args):
print ('Test Warning: ' + tmpl) % args
# SystemTest Exceptions:
class Failure (Exception):
def __init__(self, tmpl, *args):
msg = self.Prefix + (tmpl % args)
Exception.__init__(self, msg)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
class SetupFailure (Failure):
Prefix = 'Setup Failure - The test framework encountered an error:\n'
class TestFailure (Failure):
Prefix = 'TestFailure: '
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
### Unit Tests:
class Impl_A_UnitTests (unittest.TestCase):
'''Tests small stand-alone functions.'''
def test_canonicalize_cap(self):
iopairs = [('http://127.0.0.1:3456/uri/URI:DIR2:yar9nnzsho6czczieeesc65sry:upp1pmypwxits3w9izkszgo1zbdnsyk3nm6h7e19s7os7s6yhh9y',
'URI:DIR2:yar9nnzsho6czczieeesc65sry:upp1pmypwxits3w9izkszgo1zbdnsyk3nm6h7e19s7os7s6yhh9y'),
('http://127.0.0.1:3456/uri/URI%3ACHK%3Ak7ktp1qr7szmt98s1y3ha61d9w%3A8tiy8drttp65u79pjn7hs31po83e514zifdejidyeo1ee8nsqfyy%3A3%3A12%3A242?filename=welcome.html',
'URI:CHK:k7ktp1qr7szmt98s1y3ha61d9w:8tiy8drttp65u79pjn7hs31po83e514zifdejidyeo1ee8nsqfyy:3:12:242?filename=welcome.html')]
for input, output in iopairs:
result = impl_a.canonicalize_cap(input)
self.failUnlessEqual(output, result, 'input == %r' % (input,))
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
### Misc:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
class ImplProcessManager(object):
debug_wait = False
def __init__(self, name, module, mount_args, mount_wait, suites, todo=False):
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
self.name = name
self.module = module
self.script = module.__file__
self.mount_args = mount_args
self.mount_wait = mount_wait
self.suites = suites
self.todo = todo
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def maybe_wait(self, msg='waiting'):
if self.debug_wait:
print msg
raw_input()
def configure(self, client_nodedir, mountpoint):
self.client_nodedir = client_nodedir
self.mountpath = os.path.join(mountpoint, self.name)
os.mkdir(self.mountpath)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def mount(self):
print 'Mounting implementation: %s (%s)' % (self.name, self.script)
rootdirfile = os.path.join(self.client_nodedir, 'private', 'root_dir.cap')
root_uri = file(rootdirfile, 'r').read().strip()
fields = {'mountpath': self.mountpath,
'nodedir': self.client_nodedir,
'root-uri': root_uri,
}
args = ['python', self.script] + [ arg%fields for arg in self.mount_args ]
print ' '.join(args)
self.maybe_wait('waiting (about to launch fuse)')
if self.mount_wait:
exitcode, output = gather_output(args)
if exitcode != 0:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
tmpl = '%r failed to launch:\n'
tmpl += 'Exit Status: %r\n'
tmpl += 'Output:\n%s\n'
raise SetupFailure(tmpl, self.script, exitcode, output)
else:
self.proc = subprocess.Popen(args)
def umount(self):
print 'Unmounting implementation: %s' % (self.name,)
args = UNMOUNT_CMD + [self.mountpath]
print args
self.maybe_wait('waiting (unmount)')
#print os.system('ls -l '+self.mountpath)
ec, out = gather_output(args)
if ec != 0 or out:
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
tmpl = '%r failed to unmount:\n' % (' '.join(UNMOUNT_CMD),)
tmpl += 'Arguments: %r\n'
tmpl += 'Exit Status: %r\n'
tmpl += 'Output:\n%s\n'
raise SetupFailure(tmpl, args, ec, out)
def gather_output(*args, **kwargs):
'''
This expects the child does not require input and that it closes
stdout/err eventually.
'''
p = subprocess.Popen(stdout = subprocess.PIPE,
stderr = subprocess.STDOUT,
*args,
**kwargs)
output = p.stdout.read()
exitcode = p.wait()
return (exitcode, output)
fuse/tests: slew of changes to fuse 'runtests' This patch makes a significant number of changes to the fuse 'runtests' script which stem from my efforts to integrate the third fuse implementation into this framework. Perhaps not all were necessary to that end, and I beg nejucomo's forebearance if I got too carried away. - cleaned up the blank lines; imho blank lines should be empty - made the unmount command switch based on platform, since macfuse just uses 'umount' not the 'fusermount' command (which doesn't exist) - made the expected working dir for runtests the contrib/fuse dir, not the top-level tahoe source tree - see also discussion of --path-to-tahoe below - significantly reworked the ImplProcManager class. rather than subclassing for each fuse implementation to be tested, the new version is based on instantiating objects and providing relevant config info to the constructor. this was motivated by a desire to eliminate the duplication of similar but subtly different code between instances, framed by consideration of increasing the number of platforms and implementations involved. each implementation to test is thus reduced to the pertinent import and an entry in the 'implementations' table defining how to handle that implementation. this also provides a way to specify which sets of tests to run for each implementation, more on that below. - significantly reworked the command line options parsing, using twisted.usage; what used to be a single optional argument is now represented by the --test-type option which allows one to choose between running unittests, the system tests, or both. the --implementations option allows for a specific (comma-separated) list of implemenations to be tested, or the default 'all' the --tests option allows for a specific (comma-separated) list of tests sets to be run, or the default 'all'. note that only the intersection of tests requested on the command line and tests relevant to each implementation will be run. see below for more on tests sets. the --path-to-tahoe open allows for the path to the 'tahoe' executable to be specified. it defaults to '../../bin/tahoe' which is the location of the tahoe script in the source tree relative to the contrib/fuse dir by default. the --tmp-dir option controls where temporary directories (and hence mountpoints) are created during the test. this defaults to /tmp - a change from the previous behaviour of using the system default dir for calls to tempfile.mkdtemp(), a behaviour which can be obtained by providing an empty value, e.g. "--tmp-dir=" the --debug-wait flag causes the test runner to pause waiting upon user input at various stages through the testing, which facilitates debugging e.g. by allowing the user to open a browser and explore or modify the contents of the ephemeral grid after it has been instantiated but before tests are run, or make environmental adjustments before actually triggering fuse mounts etc. note that the webapi url for the first client node is printed out upon its startup to facilitate this sort of debugging also. - the default tmp dir was changed, and made configurable. previously the default behaviour of tempfile.mkdtemp() was used. it turns out that, at least on the mac, that led to temporary directories to be created in a location which ultimately led to mountpoint paths longer than could be handled by macfuse - specifically mounted filesystems could not be unmounted and would 'leak'. by changing the default location to be rooted at /tmp this leads to mountpoint paths short enough to be supported without problems. - tests are now grouped into 'sets' by method name prefix. all the existing tests have been moved into the 'read' set, i.e. with method names starting 'test_read_'. this is intended to facilitate the fact that some implementations are read-only, and some support write, so the applicability of tests will vary by implementation. the 'implementations' table, which governs the configuration of the ImplProcManager responsible for a given implementation, provides a list of 'test' (i.e test set names) which are applicable to that implementation. note no 'write' tests yet exist, this is merely laying the groundwork. - the 'expected output' of the tahoe command, which is checked for 'surprising' output by regex match, can be confused by spurious output from libraries. specfically, testing on the mac produced a warning message about zope interface resolution various multiple eggs. the 'check_tahoe_output()' function now has a list of 'ignorable_lines' (each a regex) which will be discarded before the remainder of the output of the tahoe script is matched against expectation. - cleaned up a typo, and a few spurious imports caught by pyflakes
2008-09-24 18:36:01 +00:00
def wrap_os_error(meth, *args):
try:
return meth(*args)
except os.error, e:
raise TestFailure('%s', e)
ExpectedCreationOutput = r'(introducer|client) created in (?P<path>.*?)\n'
ExpectedStartOutput = r'(.*\n)*STARTING (?P<path>.*?)\n(introducer|client) node probably started'
if __name__ == '__main__':
sys.exit(main(sys.argv))