text: move regex related functions to separate file

Eventually this should probably be rewritten to use an iternal
regex engine, currently it has unacceptable memory usage, it
copies the whole text.

6 files changed, 88 insertions, 77 deletions

diff --git a/editor.h b/editor.h
index a2ab3f5..76a3136 100644
--- a/editor.h
+++ b/editor.h
@@ -17,6 +17,7 @@ typedef struct Win Win;
 #include "syntax.h"
 #include "ring-buffer.h"
 #include "map.h"
+#include "text-regex.h"
 
 typedef union {
 	bool b;
diff --git a/text-motions.h b/text-motions.h
index 0f99076..3c7c9c5 100644
--- a/text-motions.h
+++ b/text-motions.h
@@ -7,6 +7,7 @@
 
 #include <stddef.h>
 #include "text.h"
+#include "text-regex.h"
 
 size_t text_begin(Text*, size_t pos);
 size_t text_end(Text*, size_t pos);
diff --git a/text-regex.c b/text-regex.c
new file mode 100644
index 0000000..cff8587
--- /dev/null
+++ b/text-regex.c
@@ -0,0 +1,71 @@
+#include <stdlib.h>
+#include <regex.h>
+
+#include "text-regex.h"
+
+struct Regex {
+	const char *string;
+	regex_t regex;
+};
+
+Regex *text_regex_new(void) {
+	Regex *r = calloc(1, sizeof(Regex));
+	if (!r)
+		return NULL;
+	regcomp(&r->regex, "\0\0", 0); /* this should not match anything */
+	return r;
+}
+
+int text_regex_compile(Regex *regex, const char *string, int cflags) {
+	regex->string = string;
+	int r = regcomp(&regex->regex, string, cflags);
+	if (r)
+		regcomp(&regex->regex, "\0\0", 0);
+	return r;
+}
+
+void text_regex_free(Regex *r) {
+	if (!r)
+		return;
+	regfree(&r->regex);
+	free(r);
+}
+
+int text_search_range_forward(Text *txt, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags) {
+	char *buf = malloc(len + 1);
+	if (!buf)
+		return REG_NOMATCH;
+	len = text_bytes_get(txt, pos, len, buf);
+	buf[len] = '\0';
+	regmatch_t match[nmatch];
+	int ret = regexec(&r->regex, buf, nmatch, match, eflags);
+	if (!ret) {
+		for (size_t i = 0; i < nmatch; i++) {
+			pmatch[i].start = match[i].rm_so == -1 ? EPOS : pos + match[i].rm_so;
+			pmatch[i].end = match[i].rm_eo == -1 ? EPOS : pos + match[i].rm_eo;
+		}
+	}
+	free(buf);
+	return ret;
+}
+
+int text_search_range_backward(Text *txt, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags) {
+	char *buf = malloc(len + 1);
+	if (!buf)
+		return REG_NOMATCH;
+	len = text_bytes_get(txt, pos, len, buf);
+	buf[len] = '\0';
+	regmatch_t match[nmatch];
+	char *cur = buf;
+	int ret = REG_NOMATCH;
+	while (!regexec(&r->regex, cur, nmatch, match, eflags)) {
+		ret = 0;
+		for (size_t i = 0; i < nmatch; i++) {
+			pmatch[i].start = match[i].rm_so == -1 ? EPOS : pos + (size_t)(cur - buf) + match[i].rm_so;
+			pmatch[i].end = match[i].rm_eo == -1 ? EPOS : pos + (size_t)(cur - buf) + match[i].rm_eo;
+		}
+		cur += match[0].rm_eo;
+	}
+	free(buf);
+	return ret;
+}
diff --git a/text-regex.h b/text-regex.h
new file mode 100644
index 0000000..d304d57
--- /dev/null
+++ b/text-regex.h
@@ -0,0 +1,15 @@
+#ifndef TEXT_REGEX_H
+#define TEXT_REGEX_H
+
+#include "text.h"
+
+typedef struct Regex Regex;
+typedef Filerange RegexMatch;
+
+Regex *text_regex_new(void);
+int text_regex_compile(Regex *r, const char *regex, int cflags);
+void text_regex_free(Regex *r);
+int text_search_range_forward(Text*, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags);
+int text_search_range_backward(Text*, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags);
+
+#endif
\ No newline at end of file
diff --git a/text.c b/text.c
index fee7aef..be9a8f5 100644
--- a/text.c
+++ b/text.c
@@ -20,7 +20,6 @@
 #include <time.h>
 #include <fcntl.h>
 #include <errno.h>
-#include <regex.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <sys/mman.h>
@@ -41,11 +40,6 @@
  * results in havoc. */
 #define BUFFER_MMAP_SIZE (1 << 23)
 
-struct Regex {
-	const char *string;
-	regex_t regex;
-};
-
 /* Buffer holding the file content, either readonly mmap(2)-ed from the original
  * file or heap allocated to store the modifications.
  */
@@ -1465,68 +1459,6 @@ size_t text_history_get(Text *txt, size_t index) {
 	return EPOS;
 }
 
-Regex *text_regex_new(void) {
-	Regex *r = calloc(1, sizeof(Regex));
-	if (!r)
-		return NULL;
-	regcomp(&r->regex, "\0\0", 0); /* this should not match anything */
-	return r;
-}
-
-int text_regex_compile(Regex *regex, const char *string, int cflags) {
-	regex->string = string;
-	int r = regcomp(&regex->regex, string, cflags);
-	if (r)
-		regcomp(&regex->regex, "\0\0", 0);
-	return r;
-}
-
-void text_regex_free(Regex *r) {
-	if (!r)
-		return;
-	regfree(&r->regex);
-	free(r);
-}
-
-int text_search_range_forward(Text *txt, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags) {
-	char *buf = malloc(len + 1);
-	if (!buf)
-		return REG_NOMATCH;
-	len = text_bytes_get(txt, pos, len, buf);
-	buf[len] = '\0';
-	regmatch_t match[nmatch];
-	int ret = regexec(&r->regex, buf, nmatch, match, eflags);
-	if (!ret) {
-		for (size_t i = 0; i < nmatch; i++) {
-			pmatch[i].start = match[i].rm_so == -1 ? EPOS : pos + match[i].rm_so;
-			pmatch[i].end = match[i].rm_eo == -1 ? EPOS : pos + match[i].rm_eo;
-		}
-	}
-	free(buf);
-	return ret;
-}
-
-int text_search_range_backward(Text *txt, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags) {
-	char *buf = malloc(len + 1);
-	if (!buf)
-		return REG_NOMATCH;
-	len = text_bytes_get(txt, pos, len, buf);
-	buf[len] = '\0';
-	regmatch_t match[nmatch];
-	char *cur = buf;
-	int ret = REG_NOMATCH;
-	while (!regexec(&r->regex, cur, nmatch, match, eflags)) {
-		ret = 0;
-		for (size_t i = 0; i < nmatch; i++) {
-			pmatch[i].start = match[i].rm_so == -1 ? EPOS : pos + (size_t)(cur - buf) + match[i].rm_so;
-			pmatch[i].end = match[i].rm_eo == -1 ? EPOS : pos + (size_t)(cur - buf) + match[i].rm_eo;
-		}
-		cur += match[0].rm_eo;
-	}
-	free(buf);
-	return ret;
-}
-
 bool text_range_valid(Filerange *r) {
 	return r->start != EPOS && r->end != EPOS && r->start <= r->end;
 }
diff --git a/text.h b/text.h
index 3b6344b..c322a10 100644
--- a/text.h
+++ b/text.h
@@ -129,13 +129,4 @@ ssize_t text_range_write(Text*, Filerange*, int fd);
 /* release all ressources associated with this text instance */
 void text_free(Text*);
 
-typedef struct Regex Regex;
-typedef Filerange RegexMatch;
-
-Regex *text_regex_new(void);
-int text_regex_compile(Regex *r, const char *regex, int cflags);
-void text_regex_free(Regex *r);
-int text_search_range_forward(Text*, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags);
-int text_search_range_backward(Text*, size_t pos, size_t len, Regex *r, size_t nmatch, RegexMatch pmatch[], int eflags);
-
 #endif