input_filter: dead-key sequence support

Issue #3483

repos/gems/recipes/raw/drivers_managed-pc/input_filter.config

@@ -22,7 +22,6 @@
 			</remap>
 			<mod1>
 				<key name="KEY_LEFTSHIFT"/> <key name="KEY_RIGHTSHIFT"/>
-				<rom name="capslock"/>
 			</mod1>
 			<mod2>
 				<key name="KEY_LEFTCTRL"/> <key name="KEY_RIGHTCTRL"/>
@@ -30,6 +29,9 @@
 			<mod3>
 				<key name="KEY_RIGHTALT"/> <!-- AltGr -->
 			</mod3>
+			<mod4>
+				<rom name="capslock"/>
+			</mod4>
 			<repeat delay_ms="230" rate_ms="40"/>
 			<include rom="en_us.chargen"/>
 			<include rom="special.chargen"/>

repos/os/run/input_filter.run

@@ -241,9 +241,9 @@ append config {
 						<remap>
 							<input name="usb"/>
 						</remap>
-						<mod1> <rom name="capslock"/> </mod1>
+						<mod4> <rom name="capslock"/> </mod4>
 						<map>            <key name="KEY_A" char="a"/> </map>
-						<map mod1="yes"> <key name="KEY_A" char="A"/> </map>
+						<map mod4="yes"> <key name="KEY_A" char="A"/> </map>
 					</chargen>
 				</output>
 			</filter_config>
@@ -268,6 +268,56 @@ append config {
 			<expect_press code="KEY_A" char="a"/> <expect_release code="KEY_A"/>
 
 
+			<message string="sequence handling"/>
+
+			<filter_config>
+				<input label="usb"/>
+				<output>
+					<chargen>
+						<remap>
+							<input name="usb"/>
+						</remap>
+						<map>
+							<key name="KEY_GRAVE" code="0x0300"/> <!-- dead_grave -->
+							<key name="KEY_A"     char="a"/>
+							<key name="KEY_E"     char="e"/>
+							<key name="KEY_X"     char="x"/>
+						</map>
+						<sequence first="0x0300" second="0x0061" code="0x00e0"/> <!-- LATIN SMALL LETTER A WITH GRAVE -->
+						<sequence first="0x0300" second="0x0065" code="0x00e8"/> <!-- LATIN SMALL LETTER E WITH GRAVE -->
+					</chargen>
+				</output>
+			</filter_config>
+			<sleep ms="250"/>
+			<usb>
+				<press code="KEY_GRAVE"/> <release code="KEY_GRAVE"/> <!-- invalid char -->
+				<press code="KEY_A"/>     <release code="KEY_A"/>     <!-- generate a-grave -->
+				<press code="KEY_GRAVE"/> <release code="KEY_GRAVE"/> <!-- invalid char -->
+				<press code="KEY_E"/>     <release code="KEY_E"/>     <!-- generate e-grave -->
+				<press code="KEY_GRAVE"/> <release code="KEY_GRAVE"/> <!-- invalid char -->
+				<press code="KEY_X"/>     <release code="KEY_X"/>     <!-- abort sequence (invalid char) -->
+				<press code="KEY_X"/>     <release code="KEY_X"/>     <!-- generate x -->
+			</usb>
+			<expect_press   code="KEY_GRAVE" codepoint="0xfffe"/>
+			<expect_release code="KEY_GRAVE"/>
+			<expect_press   code="KEY_A"     codepoint="0x00e0"/>
+			<expect_release code="KEY_A"/>
+			<expect_press   code="KEY_GRAVE"/>
+			<expect_release code="KEY_GRAVE"/>
+			<expect_press   code="KEY_E"     codepoint="0x00e8"/>
+			<expect_release code="KEY_E"/>
+			<expect_press   code="KEY_GRAVE"/>
+			<expect_release code="KEY_GRAVE"/>
+			<message string="1"/>
+			<expect_press   code="KEY_X"     codepoint="0xfffe"/>
+			<expect_release code="KEY_X"/>
+			<message string="2"/>
+			<expect_press   code="KEY_X" char="x"/>
+			<expect_release code="KEY_X"/>
+			<message string="3"/>
+			<sleep ms="250"/>
+
+
 			<message string="button-scroll feature"/>
 
 			<filter_config>

repos/os/src/server/input_filter/README

@@ -78,14 +78,17 @@ sub nodes:
 :<mod1>/<mod2>/<mod3>/<mod4>:
 
   Defines which physical keys are interpreted as modifier keys. Usually,
-  '<mod1>' corresponds to shift, '<mod2>' to control, and '<mod3>' to altgr
+  '<mod1>' corresponds to shift, '<mod2>' to control, '<mod3>' to altgr
-  (on German keyboards). Each modifier node may host any number of '<key>'
+  (on German keyboards), and '<mod4>' to Caps Lock. Each modifier node
-  nodes with their corresponding 'name' attribute. For example:
+  may host any number of '<key>' nodes with their corresponding 'name'
+  attribute. For example:
 
   ! <mod1>
   !   <key name="KEY_LEFTSHIFT"/> <key name="KEY_RIGHTSHIFT"/>
-  !   <rom name="capslock"/>
   ! </mod1>
+  ! <mod4>
+  !   <rom name="capslock"/>
+  ! </mod4>
 
   The '<rom>' node incorporates the content of the ROM module of the
   specified name into the modifier state. If the ROM module contains a
@@ -103,10 +106,48 @@ sub nodes:
 
   Each '<map>' may contain any number of '<key>' subnodes. Each '<key>'
   must have the key name as 'name' attribute. The to-be-emitted character
-  is defined by the attributes 'ascii', 'char', or 'b0/b1/b2/b3'. The
+  is defined by the attributes 'ascii', 'char', 'code', or 'b0/b1/b2/b3'.
-  'ascii' attribute accepts an integer value between 0 and 127, the
+
-  'char' attribute accepts a single ASCII character, the 'b0/b1/b2/b3'
+  :'ascii': accepts an integer value between 0 and 127
-  attributes define the individual bytes of an UTF-8 character.
+  :'char': accepts a single ASCII character
+  :'code': defines the Unicode codepoint as integer value
+  :'b0'/'b1'/'b2'/'b3': define the individual bytes of an UTF-8 character
+
+:<sequence first="..." second="..." third="..." fourth="..." code="..."/>:
+
+  A sequence node permits the definition of dead-key/composing
+  character sequences. With such sequences the character is not
+  generated instantly on key press but only after the sequence is
+  completed. If an unfinished sequence can't be completed due to an
+  unmatched character, the sequence is aborted and no character is
+  generated. input_filter supports sequences of up to four characters.
+
+  For example, the French AZERTY keyboard layout [1] has a dead key
+  for Circumflex Accent "^" right of the P key (which is bracket left
+  "[" on US keyboards). When Circumflex is pressed no visible
+  character should be generated instantly but the accent must be
+  combined with a follow-up character, e.g., Circumflex plus "a"
+  generates â.
+
+  [1] https://docs.microsoft.com/en-us/globalization/keyboards/kbdfr.html
+
+  Dead keys can be defined in the <key> nodes of any <map> by using
+  codepoints not used for direct output, for example, Combining
+  Diacritical Marks beginning at U+0300. The French Circumflex example
+  can be configured like follows.
+
+  ! <mod1>
+  !   <key name="KEY_LEFTSHIFT"/> <key name="KEY_RIGHTSHIFT"/>
+  ! </mod1>
+  ! <map>
+  !   <key name="KEY_Q"         code="0x0061"/> <!-- a -->
+  !   <key name="KEY_LEFTBRACE" code="0x0302"/> <!-- dead_circumflex -->
+  ! </map>
+  ! <map mod1="true">
+  !   <key name="KEY_Q"         code="0x0041"/> <!-- A -->
+  ! </map>
+  ! <sequence first="0x0302" second="0x0061" code="0x00e2"/> <!-- â -->
+  ! <sequence first="0x0302" second="0x0041" code="0x00c2"/> <!-- Â -->
 
 :<repeat delay_ms="500" rate_ms="250">:
 

repos/os/src/server/input_filter/chargen_source.h

@@ -238,6 +238,49 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 			}
 		};
 
+		struct Missing_character_definition { };
+
+		/**
+		 * Return Unicode codepoint defined in XML node attributes
+		 *
+		 * \throw Missing_character_definition
+		 */
+		static Codepoint _codepoint_from_xml_node(Xml_node node)
+		{
+			if (node.has_attribute("ascii"))
+				return Codepoint { node.attribute_value<uint32_t>("ascii", 0) };
+
+			if (node.has_attribute("code"))
+				return Codepoint { node.attribute_value<uint32_t>("code", 0) };
+
+			if (node.has_attribute("char")) {
+
+				typedef String<2> Value;
+				Value value = node.attribute_value("char", Value());
+
+				unsigned char const ascii = value.string()[0];
+
+				if (ascii < 128)
+					return Codepoint { ascii };
+
+				warning("char attribute with non-ascii character "
+				        "'", value, "'");
+				throw Missing_character_definition();
+			}
+
+			if (node.has_attribute("b0")) {
+				char const b0 = node.attribute_value("b0", 0L),
+				           b1 = node.attribute_value("b1", 0L),
+				           b2 = node.attribute_value("b2", 0L),
+				           b3 = node.attribute_value("b3", 0L);
+
+				char const buf[5] { b0, b1, b2, b3, 0 };
+				return Utf8_ptr(buf).codepoint();
+			}
+
+			throw Missing_character_definition();
+		}
+
 		/**
 		 * Map of the states of the physical keys
 		 */
@@ -286,49 +329,6 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 					               : Key::Rule::Conditions::Modifier::RELEASED;
 				}
 
-				struct Missing_character_definition { };
-
-				/**
-				 * Return UTF8 character defined in XML node attributes
-				 *
-				 * \throw Missing_character_definition
-				 */
-				static Codepoint _codepoint_from_xml_node(Xml_node node)
-				{
-					if (node.has_attribute("ascii"))
-						return Codepoint { node.attribute_value<uint32_t>("ascii", 0) };
-
-					if (node.has_attribute("code"))
-						return Codepoint { node.attribute_value<uint32_t>("code", 0) };
-
-					if (node.has_attribute("char")) {
-
-						typedef String<2> Value;
-						Value value = node.attribute_value("char", Value());
-
-						unsigned char const ascii = value.string()[0];
-
-						if (ascii < 128)
-							return Codepoint { ascii };
-
-						warning("char attribute with non-ascii character "
-						        "'", value, "'");
-						throw Missing_character_definition();
-					}
-
-					if (node.has_attribute("b0")) {
-						char const b0 = node.attribute_value("b0", 0L),
-						           b1 = node.attribute_value("b1", 0L),
-						           b2 = node.attribute_value("b2", 0L),
-						           b3 = node.attribute_value("b3", 0L);
-
-						char const buf[5] { b0, b1, b2, b3, 0 };
-						return Utf8_ptr(buf).codepoint();
-					}
-
-					throw Missing_character_definition();
-				}
-
 				void import_map(Xml_node map)
 				{
 					/* obtain modifier conditions from map attributes */
@@ -368,6 +368,150 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 					mod_rom.enabled(); });
 		}
 
+		/**
+		 * Generate characters from codepoint sequences
+		 */
+		class Sequencer
+		{
+			private:
+
+				Allocator &_alloc;
+
+				struct Sequence
+				{
+					Codepoint seq[4] { Codepoint::INVALID, Codepoint::INVALID,
+					                   Codepoint::INVALID, Codepoint::INVALID };
+
+					unsigned len { 0 };
+
+					enum Match { MISMATCH , UNFINISHED, COMPLETED };
+
+					Sequence() { }
+
+					Sequence(Codepoint c0, Codepoint c1, Codepoint c2, Codepoint c3)
+					: seq { c0, c1, c2, c3 }, len { 4 } { }
+
+					void append(Codepoint c)
+					{
+						/* excess codepoints are just dropped */
+						if (len < 4)
+							seq[len++] = c;
+					}
+
+					/**
+					 * Match 'other' to 'this' until first invalid codepoint in
+					 * 'other', completion, or mismatch
+					 */
+					Match match(Sequence const &o) const
+					{
+						/* first codepoint must match */
+						if (o.seq[0].value != seq[0].value) return MISMATCH;
+
+						for (unsigned i = 1; i < sizeof(seq)/sizeof(*seq); ++i) {
+							/* end of this sequence means COMPLETED */
+							if (!seq[i].valid()) break;
+
+							/* end of other sequence means UNFINISHED */
+							if (!o.seq[i].valid()) return UNFINISHED;
+
+							if (o.seq[i].value != seq[i].value) return MISMATCH;
+
+							/* continue until completion with both valid and equal */
+						}
+						return COMPLETED;
+					}
+				};
+
+				struct Rule
+				{
+					typedef Sequence::Match Match;
+
+					Registry<Rule>::Element element;
+					Sequence          const sequence;
+					Codepoint         const code;
+
+					Rule(Registry<Rule> &registry, Sequence const &sequence, Codepoint code)
+					:
+						element(registry, *this),
+						sequence(sequence),
+						code(code)
+					{ }
+				};
+
+				Registry<Rule> _rules { };
+
+				Sequence _curr_sequence { };
+
+			public:
+
+				Sequencer(Allocator &alloc) : _alloc(alloc) { }
+
+				~Sequencer()
+				{
+					_rules.for_each([&] (Rule &rule) {
+						destroy(_alloc, &rule); });
+				}
+
+				void import_sequence(Xml_node node)
+				{
+					unsigned const invalid { Codepoint::INVALID };
+
+					Sequence sequence {
+						Codepoint { node.attribute_value("first",  invalid) },
+						Codepoint { node.attribute_value("second", invalid) },
+						Codepoint { node.attribute_value("third",  invalid) },
+						Codepoint { node.attribute_value("fourth", invalid) } };
+
+					new (_alloc) Rule(_rules, sequence, _codepoint_from_xml_node(node));
+				}
+
+				Codepoint process(Codepoint codepoint)
+				{
+					Codepoint const invalid { Codepoint::INVALID };
+					Rule::Match best_match  { Sequence::MISMATCH };
+					Codepoint result        { codepoint };
+					Sequence seq            { _curr_sequence };
+
+					seq.append(codepoint);
+
+					_rules.for_each([&] (Rule const &rule) {
+						/* early return if completed match was found already */
+						if (best_match == Sequence::COMPLETED) return;
+
+						Rule::Match const match { rule.sequence.match(seq) };
+						switch (match) {
+						case Sequence::MISMATCH:
+							return;
+						case Sequence::UNFINISHED:
+							best_match = match;
+							result     = invalid;
+							return;
+						case Sequence::COMPLETED:
+							best_match = match;
+							result     = rule.code;
+							return;
+						}
+					});
+
+					switch (best_match) {
+					case Sequence::MISMATCH:
+						/* drop cancellation codepoint of unfinished sequence */
+						if (_curr_sequence.len > 0)
+							result = invalid;
+						_curr_sequence = Sequence();
+						break;
+					case Sequence::UNFINISHED:
+						_curr_sequence = seq;
+						break;
+					case Sequence::COMPLETED:
+						_curr_sequence = Sequence();
+						break;
+					}
+
+					return result;
+				}
+		} _sequencer;
+
 		Owner _owner;
 
 		Source::Sink &_destination;
@@ -442,6 +586,8 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 				/* supplement codepoint information to press event */
 				key.apply_best_matching_rule(_mod_map, [&] (Codepoint codepoint) {
 
+					codepoint = _sequencer.process(codepoint);
+
 					ev = Event(Input::Press_char{keycode, codepoint});
 
 					if (_char_repeater.constructed())
@@ -499,9 +645,25 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 			 * Handle map nodes
 			 */
 			if (node.type() == "map") {
+				try {
 					_key_map.import_map(node);
 					return;
 				}
+				catch (Missing_character_definition) {
+					throw Invalid_config(); }
+			}
+
+			/*
+			 * Handle sequence nodes
+			 */
+			if (node.type() == "sequence") {
+				try {
+					_sequencer.import_sequence(node);
+					return;
+				}
+				catch (Missing_character_definition) {
+					throw Invalid_config(); }
+			}
 
 			/*
 			 * Instantiate character repeater on demand
@@ -555,6 +717,7 @@ class Input_filter::Chargen_source : public Source, Source::Sink
 			_timer_accessor(timer_accessor),
 			_include_accessor(include_accessor),
 			_key_map(_alloc),
+			_sequencer(_alloc),
 			_owner(factory),
 			_destination(destination),
 			_source(factory.create_source(_owner, input_sub_node(config), *this))

repos/os/src/test/input_filter/main.cc

@@ -375,11 +375,14 @@ struct Test::Main : Input_from_filter::Event_handler
 		ev.handle_press([&] (Input::Keycode key, Codepoint codepoint) {
 
 			auto codepoint_of_step = [] (Xml_node step) {
-				return Utf8_ptr(step.attribute_value("char", Value()).string()).codepoint(); };
+				if (step.has_attribute("codepoint"))
+					return Codepoint { step.attribute_value("codepoint", 0U) };
+				return Utf8_ptr(step.attribute_value("char", Value()).string()).codepoint();
+			};
 
 			if (step.type() == "expect_press"
 			 && step.attribute_value("code", Value()) == Input::key_name(key)
-			 && (!step.has_attribute("char") ||
+			 && ((!step.has_attribute("char") && !step.has_attribute("codepoint")) ||
 			     codepoint_of_step(step).value == codepoint.value))
 				step_succeeded = true;
 		});