Implement the remaining regex types

Closes: https://github.com/tavianator/bfs/issues/21

15 files changed, 98 insertions, 6 deletions

diff --git a/src/parse.c b/src/parse.c
index 539aa05..58a0209 100644
--- a/src/parse.c
+++ b/src/parse.c
@@ -2253,16 +2253,27 @@ static struct bfs_expr *parse_regextype(struct bfs_parser *parser, int arg1, int
 	// See https://www.gnu.org/software/gnulib/manual/html_node/Predefined-Syntaxes.html
 	const char *type = expr->argv[1];
 	if (strcmp(type, "posix-basic") == 0
+	    || strcmp(type, "posix-minimal-basic") == 0
 	    || strcmp(type, "ed") == 0
 	    || strcmp(type, "sed") == 0) {
 		parser->regex_type = BFS_REGEX_POSIX_BASIC;
 	} else if (strcmp(type, "posix-extended") == 0) {
 		parser->regex_type = BFS_REGEX_POSIX_EXTENDED;
 #if BFS_WITH_ONIGURUMA
+	} else if (strcmp(type, "awk") == 0
+	           || strcmp(type, "posix-awk") == 0) {
+		parser->regex_type = BFS_REGEX_AWK;
+	} else if (strcmp(type, "gnu-awk") == 0) {
+		parser->regex_type = BFS_REGEX_GNU_AWK;
 	} else if (strcmp(type, "emacs") == 0) {
 		parser->regex_type = BFS_REGEX_EMACS;
 	} else if (strcmp(type, "grep") == 0) {
 		parser->regex_type = BFS_REGEX_GREP;
+	} else if (strcmp(type, "egrep") == 0
+	           || strcmp(type, "posix-egrep") == 0) {
+		parser->regex_type = BFS_REGEX_EGREP;
+	} else if (strcmp(type, "findutils-default") == 0) {
+		parser->regex_type = BFS_REGEX_GNU_FIND;
 #endif
 	} else if (strcmp(type, "help") == 0) {
 		parser->just_info = true;
@@ -2277,14 +2288,23 @@ static struct bfs_expr *parse_regextype(struct bfs_parser *parser, int arg1, int
 
 list_types:
 	cfprintf(cfile, "Supported types are:\n\n");
-	cfprintf(cfile, "  ${bld}posix-basic${rs}:    POSIX basic regular expressions (BRE)\n");
-	cfprintf(cfile, "  ${bld}posix-extended${rs}: POSIX extended regular expressions (ERE)\n");
-	cfprintf(cfile, "  ${bld}ed${rs}:             Like ${grn}ed${rs} (same as ${bld}posix-basic${rs})\n");
+	cfprintf(cfile, "        ${bld}posix-basic${rs}: POSIX basic regular expressions (BRE)\n");
+	cfprintf(cfile, "                 ${bld}ed${rs}: Like ${grn}ed${rs} (same as ${bld}posix-basic${rs})\n");
+	cfprintf(cfile, "                ${bld}sed${rs}: Like ${grn}sed${rs} (same as ${bld}posix-basic${rs})\n\n");
+
+	cfprintf(cfile, "     ${bld}posix-extended${rs}: POSIX extended regular expressions (ERE)\n\n");
+
 #if BFS_WITH_ONIGURUMA
-	cfprintf(cfile, "  ${bld}emacs${rs}:          Like ${grn}emacs${rs}\n");
-	cfprintf(cfile, "  ${bld}grep${rs}:           Like ${grn}grep${rs}\n");
+	cfprintf(cfile, "        [${bld}posix-${rs}]${bld}awk${rs}: Like ${grn}awk${rs}\n");
+	cfprintf(cfile, "            ${bld}gnu-awk${rs}: Like GNU ${grn}awk${rs}\n\n");
+
+	cfprintf(cfile, "              ${bld}emacs${rs}: Like ${grn}emacs${rs}\n\n");
+
+	cfprintf(cfile, "               ${bld}grep${rs}: Like ${grn}grep${rs}\n");
+	cfprintf(cfile, "      [${bld}posix-${rs}]${bld}egrep${rs}: Like ${grn}grep${rs} ${cyn}-E${rs}\n\n");
+
+	cfprintf(cfile, "  ${bld}findutils-default${rs}: Like GNU ${grn}find${rs}\n");
 #endif
-	cfprintf(cfile, "  ${bld}sed${rs}:            Like ${grn}sed${rs} (same as ${bld}posix-basic${rs})\n");
 	return NULL;
 }
 
diff --git a/src/xregex.c b/src/xregex.c
index 414c366..2d089b2 100644
--- a/src/xregex.c
+++ b/src/xregex.c
@@ -37,7 +37,11 @@ struct bfs_regex {
 static int bfs_onig_status;
 static OnigEncoding bfs_onig_enc;
 
+static OnigSyntaxType bfs_onig_syntax_awk;
+static OnigSyntaxType bfs_onig_syntax_gnu_awk;
 static OnigSyntaxType bfs_onig_syntax_emacs;
+static OnigSyntaxType bfs_onig_syntax_egrep;
+static OnigSyntaxType bfs_onig_syntax_gnu_find;
 
 /** pthread_once() callback. */
 static void bfs_onig_once(void) {
@@ -106,9 +110,34 @@ static void bfs_onig_once(void) {
 		bfs_onig_enc = NULL;
 	}
 
+	// Compute the GNU extensions
+	OnigSyntaxType *ere = ONIG_SYNTAX_POSIX_EXTENDED;
+	OnigSyntaxType *gnu = ONIG_SYNTAX_GNU_REGEX;
+	unsigned int gnu_op = gnu->op & ~ere->op;
+	unsigned int gnu_op2 = gnu->op2 & ~ere->op2;
+	unsigned int gnu_behavior = gnu->behavior & ~ere->behavior;
+
+	onig_copy_syntax(&bfs_onig_syntax_awk, ONIG_SYNTAX_POSIX_EXTENDED);
+	bfs_onig_syntax_awk.behavior |= ONIG_SYN_ALLOW_INVALID_INTERVAL;
+	bfs_onig_syntax_awk.behavior |= ONIG_SYN_BACKSLASH_ESCAPE_IN_CC;
+
+	onig_copy_syntax(&bfs_onig_syntax_gnu_awk, &bfs_onig_syntax_awk);
+	bfs_onig_syntax_gnu_awk.op |= gnu_op;
+	bfs_onig_syntax_gnu_awk.op2 |= gnu_op2;
+	bfs_onig_syntax_gnu_awk.behavior |= gnu_behavior;
+	bfs_onig_syntax_gnu_awk.behavior &= ~ONIG_SYN_CONTEXT_INDEP_REPEAT_OPS;
+	bfs_onig_syntax_gnu_awk.behavior &= ~ONIG_SYN_CONTEXT_INVALID_REPEAT_OPS;
+
 	// https://github.com/kkos/oniguruma/issues/296
 	onig_copy_syntax(&bfs_onig_syntax_emacs, ONIG_SYNTAX_EMACS);
 	bfs_onig_syntax_emacs.op2 |= ONIG_SYN_OP2_QMARK_GROUP_EFFECT;
+
+	onig_copy_syntax(&bfs_onig_syntax_egrep, ONIG_SYNTAX_POSIX_EXTENDED);
+	bfs_onig_syntax_egrep.behavior |= ONIG_SYN_ALLOW_INVALID_INTERVAL;
+	bfs_onig_syntax_egrep.behavior &= ~ONIG_SYN_CONTEXT_INVALID_REPEAT_OPS;
+
+	onig_copy_syntax(&bfs_onig_syntax_gnu_find, &bfs_onig_syntax_emacs);
+	bfs_onig_syntax_gnu_find.options |= ONIG_OPTION_MULTILINE;
 }
 
 /** Initialize Oniguruma. */
@@ -149,12 +178,24 @@ int bfs_regcomp(struct bfs_regex **preg, const char *pattern, enum bfs_regex_typ
 	case BFS_REGEX_POSIX_EXTENDED:
 		syntax = ONIG_SYNTAX_POSIX_EXTENDED;
 		break;
+	case BFS_REGEX_AWK:
+		syntax = &bfs_onig_syntax_awk;
+		break;
+	case BFS_REGEX_GNU_AWK:
+		syntax = &bfs_onig_syntax_gnu_awk;
+		break;
 	case BFS_REGEX_EMACS:
 		syntax = &bfs_onig_syntax_emacs;
 		break;
 	case BFS_REGEX_GREP:
 		syntax = ONIG_SYNTAX_GREP;
 		break;
+	case BFS_REGEX_EGREP:
+		syntax = &bfs_onig_syntax_egrep;
+		break;
+	case BFS_REGEX_GNU_FIND:
+		syntax = &bfs_onig_syntax_gnu_find;
+		break;
 	}
 	bfs_assert(syntax, "Invalid regex type");
 
diff --git a/src/xregex.h b/src/xregex.h
index 998a2b0..750db24 100644
--- a/src/xregex.h
+++ b/src/xregex.h
@@ -15,8 +15,12 @@ struct bfs_regex;
 enum bfs_regex_type {
 	BFS_REGEX_POSIX_BASIC,
 	BFS_REGEX_POSIX_EXTENDED,
+	BFS_REGEX_AWK,
+	BFS_REGEX_GNU_AWK,
 	BFS_REGEX_EMACS,
 	BFS_REGEX_GREP,
+	BFS_REGEX_EGREP,
+	BFS_REGEX_GNU_FIND,
 };
 
 /**
diff --git a/tests/gnu/regextype_awk.out b/tests/gnu/regextype_awk.out
new file mode 100644
index 0000000..0f32fc4
--- /dev/null
+++ b/tests/gnu/regextype_awk.out
@@ -0,0 +1,2 @@
+weirdnames/*/m
+weirdnames/[/k
diff --git a/tests/gnu/regextype_awk.sh b/tests/gnu/regextype_awk.sh
new file mode 100644
index 0000000..3718473
--- /dev/null
+++ b/tests/gnu/regextype_awk.sh
@@ -0,0 +1,3 @@
+invoke_bfs -regextype awk -quit || skip
+
+bfs_diff weirdnames -regextype awk -regex '.*/[\[\*]/.*'
diff --git a/tests/gnu/regextype_egrep.out b/tests/gnu/regextype_egrep.out
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/tests/gnu/regextype_egrep.out
diff --git a/tests/gnu/regextype_egrep.sh b/tests/gnu/regextype_egrep.sh
new file mode 100644
index 0000000..281d9c0
--- /dev/null
+++ b/tests/gnu/regextype_egrep.sh
@@ -0,0 +1,3 @@
+invoke_bfs -regextype egrep -quit || skip
+
+bfs_diff weirdnames -regextype egrep -regex '*.*/{l'
diff --git a/tests/gnu/regextype_findutils_default.out b/tests/gnu/regextype_findutils_default.out
new file mode 100644
index 0000000..709a7ba
--- /dev/null
+++ b/tests/gnu/regextype_findutils_default.out
@@ -0,0 +1,3 @@
+/n
+weirdnames/
+weirdnames/*/m
diff --git a/tests/gnu/regextype_findutils_default.sh b/tests/gnu/regextype_findutils_default.sh
new file mode 100644
index 0000000..c870312
--- /dev/null
+++ b/tests/gnu/regextype_findutils_default.sh
@@ -0,0 +1,3 @@
+invoke_bfs -regextype findutils-default -quit || skip
+
+bfs_diff weirdnames -regextype findutils-default -regex '.*/./\(m\|n\)'
diff --git a/tests/gnu/regextype_gnu_awk.out b/tests/gnu/regextype_gnu_awk.out
new file mode 100644
index 0000000..0f32fc4
--- /dev/null
+++ b/tests/gnu/regextype_gnu_awk.out
@@ -0,0 +1,2 @@
+weirdnames/*/m
+weirdnames/[/k
diff --git a/tests/gnu/regextype_gnu_awk.sh b/tests/gnu/regextype_gnu_awk.sh
new file mode 100644
index 0000000..6b66496
--- /dev/null
+++ b/tests/gnu/regextype_gnu_awk.sh
@@ -0,0 +1,3 @@
+invoke_bfs -regextype gnu-awk -quit || skip
+
+bfs_diff weirdnames -regextype gnu-awk -regex '.*/[\[\*]/(\<.\>)'
diff --git a/tests/gnu/regextype_posix_awk.out b/tests/gnu/regextype_posix_awk.out
new file mode 100644
index 0000000..0f32fc4
--- /dev/null
+++ b/tests/gnu/regextype_posix_awk.out
@@ -0,0 +1,2 @@
+weirdnames/*/m
+weirdnames/[/k
diff --git a/tests/gnu/regextype_posix_awk.sh b/tests/gnu/regextype_posix_awk.sh
new file mode 100644
index 0000000..86377d7
--- /dev/null
+++ b/tests/gnu/regextype_posix_awk.sh
@@ -0,0 +1,3 @@
+invoke_bfs -regextype posix-awk -quit || skip
+
+bfs_diff weirdnames -regextype posix-awk -regex '.*/[\[\*]/.*'
diff --git a/tests/gnu/regextype_posix_minimal_basic.out b/tests/gnu/regextype_posix_minimal_basic.out
new file mode 100644
index 0000000..0f0971e
--- /dev/null
+++ b/tests/gnu/regextype_posix_minimal_basic.out
@@ -0,0 +1 @@
+./(
diff --git a/tests/gnu/regextype_posix_minimal_basic.sh b/tests/gnu/regextype_posix_minimal_basic.sh
new file mode 100644
index 0000000..ee324f3
--- /dev/null
+++ b/tests/gnu/regextype_posix_minimal_basic.sh
@@ -0,0 +1,2 @@
+cd weirdnames
+bfs_diff -regextype posix-minimal-basic -regex '\./\((\)'