Merge pull request #3613 from masatake/optscript-params

main: introduce --_paramdef-<LANG>=<NAME>,<DESCRIPTION> option

Tmain/list-params.d/run.sh

@@ -27,3 +27,7 @@ echo
 echo '# CPP MACHINABLE NOHEADER'
 ${C} --with-list-header=no  --machinable --list-params=CPreProcessor
 echo
+
+echo '# CPP MACHINABLE NOHEADER + PARAM DEFINE WITH CMDLINE'
+${C} --_paramdef-CPreProcessor='pragma,handle program' --with-list-header=no  --machinable --list-params=CPreProcessor
+echo

Tmain/list-params.d/stdout-expected.txt

@@ -59,3 +59,10 @@ define	define replacement for an identifier (name(params,...)=definition)
 if0	examine code within "#if 0" branch (true or [false])
 ignore	a token to be specially handled
 
+# CPP MACHINABLE NOHEADER + PARAM DEFINE WITH CMDLINE
+_expand	expand macros if their definitions are in the current C/C++/CUDA input file (true or [false])
+define	define replacement for an identifier (name(params,...)=definition)
+if0	examine code within "#if 0" branch (true or [false])
+ignore	a token to be specially handled
+pragma	handle program
+

Units/paramdef.r/no-set.d/args.ctags

@@ -0,0 +1,11 @@
+--langdef=Foo
+--map-Foo=.foo
+--kinddef-Foo=v,xval,externally defined values
+--_paramdef-Foo=VAR,desc
+--regex-Foo=/(x)//{{
+    /VAR _param {
+        /xval _tag _commit pop
+    } {
+        \1 /xval _tag _commit pop
+    } ifelse
+}}

Units/paramdef.r/no-set.d/expected.tags

@@ -0,0 +1 @@
+x	input.foo	/^x$/;"	v

Units/paramdef.r/no-set.d/input.foo

@@ -0,0 +1 @@
+x

Units/paramdef.r/simple.d/args.ctags

@@ -0,0 +1,10 @@
+--langdef=Foo
+--map-Foo=.foo
+--kinddef-Foo=v,xval,externally defined values
+--_paramdef-Foo=VAR,desc
+--regex-Foo=/(x)//{{
+    /VAR _param {
+        /xval _tag _commit pop
+    } if
+}}
+--param-Foo.VAR=y

Units/paramdef.r/simple.d/expected.tags

@@ -0,0 +1 @@
+y	input.foo	/^x$/;"	v

Units/paramdef.r/simple.d/input.foo

@@ -0,0 +1 @@
+x

docs/optscript.rst

@@ -232,6 +232,8 @@ Related options
 
 	--list-fields
 
+	--_paramdef-<LANG>=<NAME>,<DESCRIPTION>
+
 You can run optscript code fragments when pattern specified with
 options matches successfully. The options are ``--regex-<LANG>``,
 ``--mline-regex-<LANG>``, and ``--_mtable-regex-<LANG>`` as you
@@ -269,12 +271,18 @@ represents the field has an operator for reading
 (``:fieldname``). ``w`` represents the field has an operator for
 writing (``fieldname:``).
 
+``--_paramdef-<LANG>=<NAME>,<DESCRIPTION>`` defines a language specific
+parameter named  ``<NAME>``. ``--param-<LANG>.<NAME>=<VALUE>`` assigns
+a value ``<VALUE>`` to the parameter. You can access the value from
+Optscript code with ``_param`` operator. Don't use ``{`` character
+in the description. The character is reserved for future extension.
+
 Operators
 ............................
 
 **.** -> ``-`` **.** ``corkIndex:int``
 
-    Push the cork index for the tag
+    Push the cork index for the tag.
 
 **\\n** -> ``-`` **\\n** ``matchedString:string``
 
@@ -324,7 +332,45 @@ Data types
 Recipes
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-TBW
+Parameters
+...........................................
+With ``--_paramdef-<LANG>=<param>,<description>`` option, you can
+define a parser-specific parameter to your optlib parser.
+
+In the following example, we define a parser, ``Foo`` with a
+parser-specific parameter ``VAR``. If the parser sees "x" in the
+current input stream, a code fragment attached to the pattern "/(x)/"
+runs. The code fragment does (1) check whether a value for ``VAR`` is
+given, (2) emit a tag with the value given to parameter ``VAR`` as
+name and with ``xval`` kind, and (3) remove the cork index from the
+operand stack.
+
+foo.ctags:
+.. code-block::
+
+	--langdef=Foo
+	--map-Foo=.foo
+	--kinddef-Foo=v,xval,externally defined values
+	--_paramdef-Foo=VAR,desc
+	--regex-Foo=/(x)//{{
+		/VAR _param { % if VAR is defined ....
+			/xval _tag _commit pop
+		} if
+		% if VAR is not defined, we do nothing.
+	}}
+
+input.foo:
+.. code-block::
+
+	x
+
+When running the parser, we can give a value ("hereiam" in the following example)
+for the parameter ``VAR`` from the command line:
+
+.. code-block:: console
+
+	$ ./ctags --options=args.ctags --param-Foo.VAR=hereiam -o - input.foo
+	hereiam	input.foo	/^x$/;"	v
 
 Difference between Optscript and PostScript
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

main/lregex.c

@@ -231,6 +231,7 @@ struct lregexControlBlock {
 	struct guestRequest *guest_req;
 
 	EsObject *local_dict;
+	hashTable*param_dict;
 
 	ptrArray *hook[SCRIPT_HOOK_MAX];
 	ptrArray *hook_code[SCRIPT_HOOK_MAX];
@@ -341,6 +342,8 @@ extern struct lregexControlBlock* allocLregexControlBlock (parserDefinition *par
 	lcb->tstack = ptrArrayNew(NULL);
 	lcb->guest_req = guestRequestNew ();
 	lcb->local_dict = es_nil;
+	lcb->param_dict = hashTableNew (3, hashCstrhash, hashCstreq,
+									eFree, eFree);
 
 	for (int i = 0; i< SCRIPT_HOOK_MAX; i++)
 	{
@@ -373,6 +376,9 @@ extern void freeLregexControlBlock (struct lregexControlBlock* lcb)
 	es_object_unref (lcb->local_dict);
 	lcb->local_dict = es_nil;
 
+	hashTableDelete (lcb->param_dict);
+	lcb->param_dict = NULL;
+
 	for (int i = 0; i < SCRIPT_HOOK_MAX; i++)
 	{
 		ptrArrayDelete (lcb->hook[i]);
@@ -3157,6 +3163,13 @@ extern void	addOptscriptToHook (struct lregexControlBlock *lcb, enum scriptHook
 	ptrArrayAdd (lcb->hook[hook], eStrdup (code));
 }
 
+extern void propagateParamToOptscript (struct lregexControlBlock *lcb, const char *param, const char *value)
+{
+	Assert (param);
+	Assert (value);
+	hashTablePutItem (lcb->param_dict, eStrdup (param), eStrdup (value));
+}
+
 /* Return true if available. */
 extern bool checkRegex (void)
 {
@@ -4000,6 +4013,32 @@ static EsObject *lrop_makepromise (OptVM *vm, EsObject *name)
 	return es_false;
 }
 
+static EsObject *lrop_param (OptVM *vm, EsObject *name)
+{
+	struct lregexControlBlock *lcb = opt_vm_get_app_data (vm);
+	EsObject *key = opt_vm_ostack_top (vm);
+	if (es_object_get_type (key) != OPT_TYPE_NAME)
+		return OPT_ERR_TYPECHECK;
+	EsObject *key_sym = es_pointer_get (key);
+	const char *keyc = es_symbol_get (key_sym);
+	const char *valuec = hashTableGetItem (lcb->param_dict, keyc);
+
+	if (valuec)
+	{
+		opt_vm_ostack_pop (vm);
+		EsObject *value = opt_string_new_from_cstr (valuec);
+		opt_vm_ostack_push (vm, value);
+		es_object_unref (value);
+		opt_vm_ostack_push (vm, es_true);
+	}
+	else
+	{
+		opt_vm_ostack_pop (vm);
+		opt_vm_ostack_push (vm, es_false);
+	}
+	return false;
+}
+
 static struct optscriptOperatorRegistration lropOperators [] = {
 	{
 		.name     = "_matchstr",
@@ -4165,6 +4204,13 @@ static struct optscriptOperatorRegistration lropOperators [] = {
 		.help_str = "lang:string start:matchloc end:matchloc _MAKEPROMISE promise:int true%"
 		"lang:string start:matchloc end:matchloc _MAKEPROMISE false",
 	},
+	{
+		.name     = "_param",
+		.fn       = lrop_param,
+		.arity    = 1,
+		.help_str = "param:name _PARAM value:string true%"
+		"param:name _PARAM false",
+	},
 };
 
 extern void initRegexOptscript (void)

main/lregex_p.h

@@ -111,6 +111,7 @@ extern void initRegexOptscript (void);
 extern void listRegexOpscriptOperators (FILE *fp);
 
 extern void addOptscriptToHook (struct lregexControlBlock *lcb, enum scriptHook hook, const char *code);
+extern void propagateParamToOptscript (struct lregexControlBlock *lcb, const char *param, const char *value);
 
 extern void printMultitableStatistics (struct lregexControlBlock *lcb);
 

main/mini-geany.c

@@ -184,6 +184,10 @@ static unsigned int ctagsGetLangCount(void)
 void addIgnoreSymbol(const char *value)
 {
 	langType lang = getNamedLanguage ("CPreProcessor", 0);
+	/*
+	 * In the future, we will rename applyParameter to
+	 * applyLanguageParam.
+	 */
 	applyParameter (lang, "ignore", value);
 }
 

main/options.c

@@ -381,6 +381,8 @@ static optionDescription LongOptionDescription [] = {
  {1,1,"       Copy patterns of a regex table to another regex table."},
  {1,1,"  --_mtable-regex-<LANG>=<table>/<line_pattern>/<name_pattern>/[<flags>]"},
  {1,1,"       Define multitable regular expression for locating tags in specific language."},
+ {1,1,"  --_paramdef-<LANG>=<name>,<description>"},
+ {1,1,"       Define new param for <LANG>."},
  {1,1,"  --_prelude-<LANG>={{ optscript-code }}"},
  {1,1,"       Specify code run before parsing with <LANG> parser."},
  {1,1,"  --_pretend-<NEWLANG>=<OLDLANG>"},
@@ -1709,7 +1711,7 @@ static void processIf0Option (const char *const option,
 	langType lang = getNamedLanguage ("CPreProcessor", 0);
 	const char *arg = if0? "true": "false";
 
-	applyParameter (lang, "if0", arg);
+	applyLanguageParam (lang, "if0", arg);
 }
 
 static void processLanguageForceOption (
@@ -2018,31 +2020,6 @@ extern bool processMapOption (
 	return true;
 }
 
-extern bool processParamOption (
-			const char *const option, const char *const value)
-{
-	langType language;
-	const char* name;
-	const char* sep;
-
-	language = getLanguageComponentInOption (option, "param-");
-	if (language == LANG_IGNORE)
-		return false;
-
-	sep = option + strlen ("param-") + strlen (getLanguageName (language));
-	/* `:' is only for keeping self compatibility */
-	if (! (*sep == '.' || *sep == ':' ))
-		error (FATAL, "no separator(.) is given for %s=%s", option, value);
-	name = sep + 1;
-
-	if (value == NULL || value [0] == '\0')
-		error (FATAL, "no value is given for %s", option);
-
-	applyParameter (language, name, value);
-
-	return true;
-}
-
 static void processLicenseOption (
 		const char *const option CTAGS_ATTR_UNUSED,
 		const char *const parameter CTAGS_ATTR_UNUSED)
@@ -2129,18 +2106,18 @@ static void processListParametersOption (const char *const option,
 										 const char *const parameter)
 {
 	if (parameter [0] == '\0' || strcasecmp (parameter, RSV_LANG_ALL) == 0)
-		printLanguageParameters (LANG_AUTO,
-								 localOption.withListHeader, localOption.machinable,
-								 stdout);
+		printLanguageParams (LANG_AUTO,
+							 localOption.withListHeader, localOption.machinable,
+							 stdout);
 	else
 	{
 		langType language = getNamedLanguage (parameter, 0);
 		if (language == LANG_IGNORE)
 			error (FATAL, "Unknown language \"%s\" in \"%s\" option", parameter, option);
 		else
-			printLanguageParameters (language,
-									 localOption.withListHeader, localOption.machinable,
-									 stdout);
+			printLanguageParams (language,
+								 localOption.withListHeader, localOption.machinable,
+								 stdout);
 	}
 	exit (0);
 }
@@ -2595,7 +2572,7 @@ static void readIgnoreList (const char *const list)
 
 	while (token != NULL)
 	{
-		applyParameter (lang, "ignore", token);
+		applyLanguageParam (lang, "ignore", token);
 		token = strtok (NULL, IGNORE_SEPARATORS);
 	}
 	eFree (newList);
@@ -2615,7 +2592,7 @@ static void addIgnoreListFromFile (const char *const fileName)
 	for(i=0;i<c;i++)
 	{
 		vString * s = stringListItem(tokens,i);
-		applyParameter (lang, "ignore", vStringValue(s));
+		applyLanguageParam (lang, "ignore", vStringValue(s));
 	}
 
 	stringListDelete(tokens);
@@ -2626,7 +2603,7 @@ static void processIgnoreOption (const char *const list, int IgnoreOrDefine)
 	langType lang = getNamedLanguage ("CPreProcessor", 0);
 
 	if (IgnoreOrDefine == 'D')
-		applyParameter (lang, "define", list);
+		applyLanguageParam (lang, "define", list);
 	else if (strchr ("@./\\", list [0]) != NULL)
 	{
 		const char* fileName = (*list == '@') ? list + 1 : list;
@@ -2637,7 +2614,7 @@ static void processIgnoreOption (const char *const list, int IgnoreOrDefine)
 		addIgnoreListFromFile (list);
 #endif
 	else if (strcmp (list, "-") == 0)
-		applyParameter (lang, "ignore", NULL);
+		applyLanguageParam (lang, "ignore", NULL);
 	else
 		readIgnoreList (list);
 }
@@ -3352,6 +3329,8 @@ static void processLongOption (
 		;
 	else if (processMapOption (option, parameter))
 		;
+	else if (processParamdefOption (option, parameter))
+		;
 	else if (processParamOption (option, parameter))
 		;
 	else if (processTabledefOption (option, parameter))

main/options_p.h

@@ -162,6 +162,7 @@ extern langType getLanguageComponentInOptionFull (const char *const option,
 
 extern void processLanguageDefineOption (const char *const option, const char *const parameter);
 extern bool processMapOption (const char *const option, const char *const parameter);
+extern bool processParamdefOption (const char *const option, const char *const value);
 extern bool processParamOption (const char *const option, const char *const value);
 extern bool processKinddefOption (const char *const option, const char *const parameter);
 extern bool processKindsOption (const char *const option, const char *const parameter);