Merge pull request #92 from ehlers/check_url

use GET requests instead of HEAD for checking URLs

appliances/microsoft-windows+ie.gns3a

@@ -3,7 +3,7 @@
     "category": "guest",
     "description": "Microsoft Windows (or simply Windows) is a graphical operating system developed, marketed, and sold by Microsoft.\n\nMicrosoft releases time limited VMs for testing Internet Explorer.\n\nOn the download site select the VM, as platform select VirtualBox, then download the zip file, afterwards unzip it.",
     "vendor_name": "Microsoft",
-    "vendor_url": "https://dev.microsoft.com/",
+    "vendor_url": "http://www.microsoft.com",
     "product_name": "Windows",
     "registry_version": 1,
     "status": "experimental",

check_urls.py

@@ -22,28 +22,39 @@ import pycurl
 
 err_list = []
 
+
+def data_abort(data):
+    return -1
+
+
 def check_url(url, appliance):
     print("   " + url)
 
     error = None
+    c = pycurl.Curl()
     try:
-        c = pycurl.Curl()
         c.setopt(c.URL, url)
         c.setopt(c.USERAGENT, 'Mozilla/5.0 (compatible; MSIE 10.0; Windows NT 6.1; Trident/6.0)')
-        c.setopt(c.NOBODY, True)
+        c.setopt(c.HTTPHEADER, ['Accept-Language: en-us'])
         c.setopt(c.FOLLOWLOCATION, True)
+        c.setopt(c.WRITEFUNCTION, data_abort)
         c.perform()
+    except pycurl.error as err:
+        errno, errstr = err.args
+        if errno != pycurl.E_WRITE_ERROR:
+            error = errstr
+
+    if not error:
         http_status = c.getinfo(c.RESPONSE_CODE)
         if http_status >= 400:
             error = 'HTTP status {}'.format(http_status)
-        c.close()
-    except pycurl.error:
-        error = c.errstr()
 
     if error:
         print("     " + error)
         err_list.append("{}: {} - {}".format(appliance, url, error))
 
+    c.close()
+
 
 def check_urls(appliance):
     try: