Paolo Bonzini wrote:
> If there is one, use the faster u8_strchr algorithm (you can
> just use u8_strchr, even though that does a useless conversion back to
> UTF-8).
Nice suggestion. Implemented as follows:
2010-07-31 Bruno Haible <br...@clisp.org>
unistr/u8-strstr, unistr/u16-strstr: Optimize the one-character case.
* lib/unistr/u-strstr.h (FUNC): When the needle contains only one
character, perform the search using U_STRCHR.
* lib/unistr/u8-strstr.c (U_STRMBTOUC): New macro.
* lib/unistr/u16-strstr.c (U_STRMBTOUC): Likewise.
* modules/unistr/u8-strstr (Depends-on): Add unistr/u8-strmbtouc.
* modules/unistr/u16-strstr (Depends-on): Add unistr/u16-strmbtouc.
Suggested by Paolo Bonzini.
--- lib/unistr/u-strstr.h.orig Sat Jul 31 22:05:02 2010
+++ lib/unistr/u-strstr.h Sat Jul 31 21:59:37 2010
@@ -1,5 +1,5 @@
/* Substring test for UTF-8/UTF-16/UTF-32 strings.
- Copyright (C) 1999, 2002, 2006, 2009-2010 Free Software Foundation, Inc.
+ Copyright (C) 1999, 2002, 2006, 2010 Free Software Foundation, Inc.
Written by Bruno Haible <br...@clisp.org>, 2002.
This program is free software: you can redistribute it and/or modify it
@@ -24,10 +24,20 @@
if (first == 0)
return (UNIT *) haystack;
- /* Is needle nearly empty? */
+ /* Is needle nearly empty (only one unit)? */
if (needle[1] == 0)
return U_STRCHR (haystack, first);
+#ifdef U_STRMBTOUC
+ /* Is needle nearly empty (only one character)? */
+ {
+ ucs4_t first_uc;
+ int count = U_STRMBTOUC (&first_uc, needle);
+ if (count > 0 && needle[count] == 0)
+ return U_STRCHR (haystack, first_uc);
+ }
+#endif
+
/* Search for needle's first unit. */
for (; *haystack != 0; haystack++)
if (*haystack == first)
--- lib/unistr/u16-strstr.c.orig Sat Jul 31 22:05:02 2010
+++ lib/unistr/u16-strstr.c Sat Jul 31 21:56:25 2010
@@ -1,5 +1,5 @@
/* Substring test for UTF-16 strings.
- Copyright (C) 1999, 2002, 2006, 2009-2010 Free Software Foundation, Inc.
+ Copyright (C) 1999, 2002, 2006, 2010 Free Software Foundation, Inc.
Written by Bruno Haible <br...@clisp.org>, 2002.
This program is free software: you can redistribute it and/or modify it
@@ -25,4 +25,5 @@
#define FUNC u16_strstr
#define UNIT uint16_t
#define U_STRCHR u16_strchr
+#define U_STRMBTOUC u16_strmbtouc
#include "u-strstr.h"
--- lib/unistr/u8-strstr.c.orig Sat Jul 31 22:05:02 2010
+++ lib/unistr/u8-strstr.c Sat Jul 31 21:56:25 2010
@@ -1,5 +1,5 @@
/* Substring test for UTF-8 strings.
- Copyright (C) 1999, 2002, 2006, 2009-2010 Free Software Foundation, Inc.
+ Copyright (C) 1999, 2002, 2006, 2010 Free Software Foundation, Inc.
Written by Bruno Haible <br...@clisp.org>, 2002.
This program is free software: you can redistribute it and/or modify it
@@ -25,4 +25,5 @@
#define FUNC u8_strstr
#define UNIT uint8_t
#define U_STRCHR u8_strchr
+#define U_STRMBTOUC u8_strmbtouc
#include "u-strstr.h"
--- modules/unistr/u16-strstr.orig Sat Jul 31 22:05:02 2010
+++ modules/unistr/u16-strstr Sat Jul 31 22:04:36 2010
@@ -8,6 +8,7 @@
Depends-on:
unistr/base
unistr/u16-strchr
+unistr/u16-strmbtouc
configure.ac:
gl_LIBUNISTRING_MODULE([0.9], [unistr/u16-strstr])
--- modules/unistr/u8-strstr.orig Sat Jul 31 22:05:02 2010
+++ modules/unistr/u8-strstr Sat Jul 31 22:04:43 2010
@@ -8,6 +8,7 @@
Depends-on:
unistr/base
unistr/u8-strchr
+unistr/u8-strmbtouc
configure.ac:
gl_LIBUNISTRING_MODULE([0.9], [unistr/u8-strstr])
