From d2004b15f1e90efafedc367335c07ad4636d291d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marc=20Andr=C3=A9=20Tanner?= <mat@brain-dump.org>
Date: Fri, 7 Apr 2017 16:04:40 +0200
Subject: text: add mem{r,}chr(3) based byte search functions

These are generally implemented efficiently in libc.
While memrchr(3) is non-standard, it is a common extension.
If it is not available, we use a simple C implementation from musl.
---
 Makefile  |  5 ++++-
 configure | 20 ++++++++++++++++++++
 text.c    | 29 +++++++++++++++++++++++++++++
 text.h    |  2 ++
 util.h    | 11 +++++++++++
 5 files changed, 66 insertions(+), 1 deletion(-)

diff --git a/Makefile b/Makefile
index 8e114e9..ba79108 100644
--- a/Makefile
+++ b/Makefile
@@ -32,8 +32,11 @@ CFLAGS_STD ?= -std=c99 -D_POSIX_C_SOURCE=200809L -D_XOPEN_SOURCE=700 -DNDEBUG
 CFLAGS_STD += -DVERSION=\"${VERSION}\"
 LDFLAGS_STD ?= -lc
 
+CFLAGS_LIBC ?= -DHAVE_MEMRCHR=0
+
 CFLAGS_VIS = $(CFLAGS_AUTO) $(CFLAGS_TERMKEY) $(CFLAGS_CURSES) $(CFLAGS_ACL) \
-	$(CFLAGS_SELINUX) $(CFLAGS_TRE) $(CFLAGS_LUA) $(CFLAGS_LPEG) $(CFLAGS_STD)
+	$(CFLAGS_SELINUX) $(CFLAGS_TRE) $(CFLAGS_LUA) $(CFLAGS_LPEG) $(CFLAGS_STD) \
+	$(CFLAGS_LIBC)
 
 CFLAGS_VIS += -DVIS_PATH=\"${SHAREPREFIX}/vis\"
 CFLAGS_VIS += -DCONFIG_HELP=${CONFIG_HELP}
diff --git a/configure b/configure
index f12b96b..8cfd512 100755
--- a/configure
+++ b/configure
@@ -594,6 +594,25 @@ EOF
 	fi
 fi
 
+printf "checking for memrchr... "
+
+cat > "$tmpc" <<EOF
+#define _GNU_SOURCE
+#include <string.h>
+
+int main(int argc, char *argv[]) {
+        return !memrchr("\n", '\n', 1);
+}
+EOF
+
+if $CC $CFLAGS "$tmpc" $LDFLAGS -o "$tmpo" >/dev/null 2>&1; then
+	HAVE_MEMRCHR=1
+	printf "%s\n" "yes"
+else
+	HAVE_MEMRCHR=0
+	printf "%s\n" "no"
+fi
+
 printf "completing config.mk... "
 
 exec 3>&1 1>>config.mk
@@ -621,6 +640,7 @@ LDFLAGS_ACL = $LDFLAGS_ACL
 CONFIG_SELINUX = $CONFIG_SELINUX
 CFLAGS_SELINUX = $CFLAGS_SELINUX
 LDFLAGS_SELINUX = $LDFLAGS_SELINUX
+CFLAGS_LIBC = -DHAVE_MEMRCHR=$HAVE_MEMRCHR
 EOF
 exec 1>&3 3>&-
 
diff --git a/text.c b/text.c
index f789339..1bf1940 100644
--- a/text.c
+++ b/text.c
@@ -1,3 +1,4 @@
+#define _GNU_SOURCE // memrchr(3) is non-standard
 #include <unistd.h>
 #include <stdio.h>
 #include <stdlib.h>
@@ -1455,6 +1456,34 @@ bool text_iterator_byte_prev(Iterator *it, char *b) {
 	return true;
 }
 
+bool text_iterator_byte_find_prev(Iterator *it, char b) {
+	while (it->text) {
+		const char *match = memrchr(it->start, b, it->text - it->start);
+		if (match) {
+			it->pos -= it->text - match;
+			it->text = match;
+			return true;
+		}
+		text_iterator_prev(it);
+	}
+	text_iterator_next(it);
+	return false;
+}
+
+bool text_iterator_byte_find_next(Iterator *it, char b) {
+	while (it->text) {
+		const char *match = memchr(it->text, b, it->end - it->text);
+		if (match) {
+			it->pos += match - it->text;
+			it->text = match;
+			return true;
+		}
+		text_iterator_next(it);
+	}
+	text_iterator_prev(it);
+	return false;
+}
+
 bool text_iterator_codepoint_next(Iterator *it, char *c) {
 	while (text_iterator_byte_next(it, NULL)) {
 		if (ISUTF8(*it->text)) {
diff --git a/text.h b/text.h
index 0e372c1..c2a5812 100644
--- a/text.h
+++ b/text.h
@@ -97,6 +97,8 @@ bool text_iterator_byte_prev(Iterator*, char *b);
 /* if the new position is at EOF a NUL byte (which is not actually
  * part of the file) is read. */
 bool text_iterator_byte_next(Iterator*, char *b);
+bool text_iterator_byte_find_prev(Iterator*, char b);
+bool text_iterator_byte_find_next(Iterator*, char b);
 /* move to the next/previous UTF-8 encoded Unicode codepoint
  * and set c (if it is non NULL) to the first byte */
 bool text_iterator_codepoint_next(Iterator *it, char *c);
diff --git a/util.h b/util.h
index a53deda..9048c82 100644
--- a/util.h
+++ b/util.h
@@ -23,4 +23,15 @@ static inline bool addu(size_t a, size_t b, size_t *c) {
 }
 #endif
 
+#if !HAVE_MEMRCHR
+/* MIT licensed implementation from musl libc */
+static void *memrchr(const void *m, int c, size_t n)
+{
+	const unsigned char *s = m;
+	c = (unsigned char)c;
+	while (n--) if (s[n]==c) return (void *)(s+n);
+	return 0;
+}
 #endif
+
+#endif /* UTIL_H */
-- 
cgit v1.2.3