main: using regex for choosing a parser for the given file name

This change extends --map-<LANG> option to support regular
expression matching with the full file name.

The original --map-<LANG> option supports the glob based matching
and the extension comparison with the file basename.
However, two methods are not enough if the file names are too
generic. See #3287 .

The regular expression passed to --map-<LANG> must be surrounded
by % character like

   --map-RpmMacros='%(.*/)?macros\.d/macros\.([^/]+)$%'

If you want to match in a case-insensitive way, append `i' after
the second % like

   --map-RpmMacros='%(.*/)?macros\.d/macros\.([^/]+)$%i'

If you want to use % as part of an expression, put \ before %
for escaping.

TODO:

- [ ] update ctags.1
- [ ] add Tmain test cases
- [ ] add pcre backend
- [ ] update NEWS

Signed-off-by: Masatake YAMATO <[email protected]>

Author: masatake

Tmain/list-map-rexprs.d/exit-expected.txt

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

Tmain/list-map-rexprs.d/run.sh

@@ -0,0 +1,16 @@
+# Copyright: 2025 Masatake YAMATO
+# License: GPL-2
+
+CTAGS=$1
+
+$CTAGS --quiet --options=NONE \
+	   --langdef=Something \
+	   --map-Something='%\%ESCAPING\%%' \
+	   --map-Something=+'%ICASE%i' \
+	   --map-Something=+'%TEMP%' \
+	   --map-Something=-'%TEMP%' \
+	   --map-Something=+'%TEMPi%i' \
+	   --map-Something=-'%TEMPi%i' \
+	   --list-map-rexprs=all && \
+echo '## RpmMacros' && \
+$CTAGS --quiet --options=NONE --list-map-rexprs=RpmMacros

Tmain/list-map-rexprs.d/stderr-expected.txt

@@ -0,0 +1,7 @@
+#LANGUAGE  EXPRESSION                       CASE
+RpmMacros  (.*/)?macros\.d/macros\.([^/]+)$ sensitive
+Something  %ESCAPING%                       sensitive
+Something  ICASE                            insensitive
+## RpmMacros
+#EXPRESSION                       CASE
+(.*/)?macros\.d/macros\.([^/]+)$  sensitive

Tmain/list-map-rexprs.d/stdout-expected.txt

@@ -16,6 +16,10 @@ About TEST language
 enabled: yes
 version: 10.9
 
+Mappings/rexprs
+-------------------------------------------------------
+
+
 Mappings/patterns
 -------------------------------------------------------
  MYTEST

Tmain/versioning.d/stdout-expected.txt

@@ -39,6 +39,7 @@
 #include "interactive_p.h"
 #include "writer_p.h"
 #include "trace.h"
+#include "flags_p.h"
 
 #ifdef HAVE_JANSSON
 #include <jansson.h>
@@ -59,6 +60,9 @@
 /*  The following separators are permitted for list options.
  */
 #define EXTENSION_SEPARATOR '.'
+#define REXPR_START '%'
+#define REXPR_STOP '%'
+#define REXPR_ICASE 'i'
 #define PATTERN_START '('
 #define PATTERN_STOP  ')'
 #define IGNORE_SEPARATORS   ", \t\n"
@@ -303,10 +307,10 @@ static optionDescription LongOptionDescription [] = {
  {1,0,"  --langmap=<map>[,<map>[...]]"},
  {1,0,"       Override default mapping of language to input file extension."},
  {1,0,"       e.g. --langmap=c:.c.x,java:+.j,make:([Mm]akefile).mak"},
- {1,0,"  --map-<LANG>=[+|-]<extension>|<pattern>"},
+ {1,0,"  --map-<LANG>=[+|-]<extension>|<pattern>|<rexpr>"},
  {1,0,"       Set, add(+) or remove(-) the map for <LANG>."},
- {1,0,"       Unlike --langmap, this doesn't take a list; only one file name <pattern>"},
- {1,0,"       or one file <extension> can be specified at once."},
+ {1,0,"       Unlike --langmap, this doesn't take a list; only one file name <pattern>,"},
+ {1,0,"       one file name <extension>, or one file <rexpr> can be specified at once."},
  {1,0,"       Unlike --langmap the change with this option affects mapping of <LANG> only."},
  {1,0,""},
  {1,0,"Tags File Contents Options"},
@@ -436,6 +440,8 @@ static optionDescription LongOptionDescription [] = {
  {1,0,"       Output list of language extensions in mapping."},
  {1,0,"  --list-map-patterns[=(<language>|all)]"},
  {1,0,"       Output list of language patterns in mapping."},
+ {1,0,"  --list-map-rexprs[=(<language>|all)]"},
+ {1,0,"       Output list of language regular expressions in mapping."},
  {1,0,"  --list-maps[=(<language>|all)]"},
  {1,0,"       Output list of language mappings (both extensions and patterns)."},
  {1,0,"  --list-mline-regex-flags"},
@@ -1793,6 +1799,7 @@ static char* extractMapFromParameter (const langType language,
 		++parameter;
 		for (p = parameter  ;  *p != PATTERN_STOP  &&  *p != '\0'  ;  ++p)
 		{
+			/* TODO: Can this handle a pattern including ')' ? */
 			if (*p == '\\'  &&  *(p + 1) == PATTERN_STOP)
 				++p;
 		}
@@ -1808,9 +1815,46 @@ static char* extractMapFromParameter (const langType language,
 		return result;
 	}
 
+	if (first == REXPR_START)
+	{
+		*mapType = LMAP_REXPR;
+
+		++parameter;
+		const char* parameter_orig = parameter;
+		vString *rexpr = vStringNew ();
+		for (p = parameter  ;  *p != REXPR_STOP  &&  *p != '\0'  ;  ++p)
+		{
+			if (*p == '\\'  &&  *(p + 1) == REXPR_STOP)
+				p++;
+			vStringPut (rexpr, *p);
+		}
+		if (*p == '\0')
+			error (FATAL, "Unterminated file name regular expression for %s language: %s",
+				   getLanguageName (language), parameter_orig);
+
+		*tail = p + 1;
+		return vStringDeleteUnwrap (rexpr);
+	}
+
 	return NULL;
 }
 
+static void langmap_rexpr_icase_short (char c CTAGS_ATTR_UNUSED, void* data)
+{
+	bool *icase = data;
+	*icase = true;
+}
+
+static void langmap_rexpr_icase_long (const char* s CTAGS_ATTR_UNUSED, const char* const unused CTAGS_ATTR_UNUSED, void* data)
+{
+	langmap_rexpr_icase_short ('i', data);
+}
+
+static flagDefinition langmapRexprFlagDef[] = {
+	{ 'i', "icase",  langmap_rexpr_icase_short,  langmap_rexpr_icase_long,
+	  NULL, "applied in a case-insensitive manner"},
+};
+
 static char* addLanguageMap (const langType language, char* map_parameter,
 			     bool exclusiveInAllLanguages)
 {
@@ -1823,6 +1867,13 @@ static char* addLanguageMap (const langType language, char* map_parameter,
 		addLanguageExtensionMap (language, map, exclusiveInAllLanguages);
 	else if (map && map_type == LMAP_PATTERN)
 		addLanguagePatternMap (language, map, exclusiveInAllLanguages);
+	else if (map && map_type == LMAP_REXPR)
+	{
+		bool icase = false;
+
+		flagsEval (p, langmapRexprFlagDef, ARRAY_SIZE(langmapRexprFlagDef), &icase);
+		addLanguageRexprMap (language, map, icase, exclusiveInAllLanguages);
+	}
 	else
 		error (FATAL, "Badly formed language map for %s language",
 				getLanguageName (language));
@@ -1843,6 +1894,13 @@ static char* removeLanguageMap (const langType language, char* map_parameter)
 		removeLanguageExtensionMap (language, map);
 	else if (map && map_type == LMAP_PATTERN)
 		removeLanguagePatternMap (language, map);
+	else if (map && map_type == LMAP_REXPR)
+	{
+		bool icase = (*p == REXPR_ICASE);
+		removeLanguageRexprMap (language, map, icase);
+		if (icase)
+			p++;
+	}
 	else
 		error (FATAL, "Badly formed language map for %s language",
 		       getLanguageName (language));
@@ -2164,6 +2222,13 @@ static void processListMapPatternsOption (const char *const option,
 	processListMapsOptionForType (option, parameter, LMAP_PATTERN|LMAP_TABLE_OUTPUT);
 }
 
+static void processListMapRexprsOption (const char *const option,
+				       const char *const parameter)
+{
+	processListMapsOptionForType (option, parameter, LMAP_REXPR|LMAP_TABLE_OUTPUT);
+}
+
+
 static void processListMapsOption (
 		const char *const option CTAGS_ATTR_UNUSED,
 		const char *const parameter CTAGS_ATTR_UNUSED)
@@ -2327,6 +2392,13 @@ static void processDescribeLanguage(const char *const option,
 		   getLanguageVersionCurrent (language),
 		   getLanguageVersionAge (language));
 
+	puts("");
+	puts("Mappings/rexprs");
+	puts("-------------------------------------------------------");
+	printLanguageMaps (language, LMAP_REXPR|LMAP_NO_LANG_PREFIX,
+					   localOption.withListHeader, localOption.machinable,
+					   stdout);
+
 	puts("");
 	puts("Mappings/patterns");
 	puts("-------------------------------------------------------");
@@ -2999,6 +3071,7 @@ static parametricOption ParametricOptions [] = {
 	{ "list-maps",              processListMapsOption,          true,   STAGE_ANY },
 	{ "list-map-extensions",    processListMapExtensionsOption, true,   STAGE_ANY },
 	{ "list-map-patterns",      processListMapPatternsOption,   true,   STAGE_ANY },
+	{ "list-map-rexprs",        processListMapRexprsOption,     true, STAGE_ANY },
 	{ "list-mline-regex-flags", processListMultilineRegexFlagsOption, true, STAGE_ANY },
 	{ "list-output-formats",    processListOutputFormatsOption, true,   STAGE_ANY },
 	{ "list-params",            processListParametersOption,    true,   STAGE_ANY },