Author: stsp
Date: Wed Feb 1 16:58:50 2012
New Revision: 1239203
URL: http://svn.apache.org/viewvc?rev=1239203&view=rev
Log:
For issue #2487, "mod_dav_svn and locales fail to play nicely together",
add a new configuration option which allows Subversion libraries called
from mod_dav_svn to process UTF-8 strings, instead of just ASCII strings.
Because httpd is not locale-aware, we cannot assume that a native character
set other than ASCII is present in the environment that mod_dav_svn runs in.
UTF-8 is backward compatible to ASCII, and all paths within Subversion
repositories are already encoded in UTF-8. Thus, we can safely allow UTF-8
without running into character set conversion issues in code called from
mod_dav_svn.
Basic testing shows that httpd is prepared to tolerate non-ASCII characters,
even though it doesn't seem to interpret them. E.g. it logs such characters
as hex codes preceded by a backslash.
The new configuration option is called SVNUseUTF8 and can be set to
"On" or "Off" (the default is "Off"). It is a global option that takes
effect for all repository locations defined in the configuration file.
When the option is "On", files that contain UTF-8 characters in their
names can be processed by the pre-lock hook, hooks are able to write
UTF-8 characters to stdout and stderr, and filesystem paths to Subversion
repositories may contain UTF-8 characters.
* subversion/mod_dav_svn/mod_dav_svn.c
(server_conf_t): Add use_utf8 boolean.
(init): Initialise the native character set of Subversion's UTF-8
conversion routines to either ASCII (default) or UTF-8.
(SVNUseUTF8_cmd): New option command processor.
(cmds): Add new "SVNUseUTF8" option.
* subversion/include/svn_utf.h
(svn_utf_initialize2): Declare. Includes the ablity to force the native
character set to UTF-8, regardless of locale.
(svn_utf_initialize): Deprecate.
* subversion/libsvn_subr/deprecated.c
(svn_utf_initialize): Re-implement as wrapper around svn_utf_initialize2().
* subversion/libsvn_subr/cmdline.c
(svn_cmdline_init): Update funtion call to svn_utf_initialize2(), keeping
the native character set based on the current locale.
* subversion/libsvn_subr/utf.c
(assume_native_charset_is_utf8): New global boolean variable. Defaults
to FALSE, and may be set to TRUE by svn_utf_initialize2() (which, like
its predecessor, is documented to not be thread-safe).
(get_ntou_xlate_handle_node, get_uton_xlate_handle_node): If the native
character set has been forced to UTF-8, use SVN_APR_UTF8_CHARSET instead
of SVN_APR_LOCALE_CHARSET.
Modified:
subversion/trunk/subversion/include/svn_utf.h
subversion/trunk/subversion/libsvn_subr/cmdline.c
subversion/trunk/subversion/libsvn_subr/deprecated.c
subversion/trunk/subversion/libsvn_subr/utf.c
subversion/trunk/subversion/mod_dav_svn/mod_dav_svn.c
Modified: subversion/trunk/subversion/include/svn_utf.h
URL:
http://svn.apache.org/viewvc/subversion/trunk/subversion/include/svn_utf.h?rev=1239203&r1=1239202&r2=1239203&view=diff
==============================================================================
--- subversion/trunk/subversion/include/svn_utf.h (original)
+++ subversion/trunk/subversion/include/svn_utf.h Wed Feb 1 16:58:50 2012
@@ -49,12 +49,28 @@ extern "C" {
* Initialize the UTF-8 encoding/decoding routines.
* Allocate cached translation handles in a subpool of @a pool.
*
+ * If @a assume_native_utf8 is TRUE, the native character set is
+ * assumed to be UTF-8, i.e. conversion is a no-op. This is useful
+ * in contexts where the native character set is ASCII but UTF-8
+ * should be used regardless (e.g. for mod_dav_svn which runs within
+ * httpd and always uses the "C" locale).
+ *
* @note It is optional to call this function, but if it is used, no other
* svn function may be in use in other threads during the call of this
* function or when @a pool is cleared or destroyed.
* Initializing the UTF-8 routines will improve performance.
*
- * @since New in 1.1.
+ * @since New in 1.8.
+ */
+void
+svn_utf_initialize2(apr_pool_t *pool,
+ svn_boolean_t assume_native_utf8);
+
+/**
+ * Like svn_utf_initialize but without the ability to force the
+ * native encoding to UTF-8.
+ *
+ * @deprecated Provided for backward compatibility with the 1.7 API.
*/
void
svn_utf_initialize(apr_pool_t *pool);
Modified: subversion/trunk/subversion/libsvn_subr/cmdline.c
URL:
http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_subr/cmdline.c?rev=1239203&r1=1239202&r2=1239203&view=diff
==============================================================================
--- subversion/trunk/subversion/libsvn_subr/cmdline.c (original)
+++ subversion/trunk/subversion/libsvn_subr/cmdline.c Wed Feb 1 16:58:50 2012
@@ -226,7 +226,7 @@ svn_cmdline_init(const char *progname, F
/* Create a pool for use by the UTF-8 routines. It will be cleaned
up by APR at exit time. */
pool = svn_pool_create(NULL);
- svn_utf_initialize(pool);
+ svn_utf_initialize2(pool, FALSE);
if ((err = svn_nls_init()))
{
Modified: subversion/trunk/subversion/libsvn_subr/deprecated.c
URL:
http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_subr/deprecated.c?rev=1239203&r1=1239202&r2=1239203&view=diff
==============================================================================
--- subversion/trunk/subversion/libsvn_subr/deprecated.c (original)
+++ subversion/trunk/subversion/libsvn_subr/deprecated.c Wed Feb 1 16:58:50
2012
@@ -39,6 +39,7 @@
#include "svn_pools.h"
#include "svn_dso.h"
#include "svn_mergeinfo.h"
+#include "svn_utf.h"
#include "svn_xml.h"
#include "opt.h"
@@ -1173,3 +1174,9 @@ svn_xml_make_header(svn_stringbuf_t **st
{
svn_xml_make_header2(str, NULL, pool);
}
+
+void
+svn_utf_initialize(apr_pool_t *pool)
+{
+ svn_utf_initialize2(pool, FALSE);
+}
Modified: subversion/trunk/subversion/libsvn_subr/utf.c
URL:
http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_subr/utf.c?rev=1239203&r1=1239202&r2=1239203&view=diff
==============================================================================
--- subversion/trunk/subversion/libsvn_subr/utf.c (original)
+++ subversion/trunk/subversion/libsvn_subr/utf.c Wed Feb 1 16:58:50 2012
@@ -23,6 +23,7 @@
+#include <stdlib.h>
#include <string.h>
#include <assert.h>
@@ -55,6 +56,7 @@ static const char *SVN_UTF_UTON_XLATE_HA
static const char *SVN_APR_UTF8_CHARSET = "UTF-8";
static svn_mutex__t *xlate_handle_mutex = NULL;
+static svn_boolean_t assume_native_charset_is_utf8 = FALSE;
/* The xlate handle cache is a global hash table with linked lists of xlate
* handles. In multi-threaded environments, a thread "borrows" an xlate
@@ -118,7 +120,8 @@ xlate_handle_node_cleanup(void *arg)
}
void
-svn_utf_initialize(apr_pool_t *pool)
+svn_utf_initialize2(apr_pool_t *pool,
+ svn_boolean_t assume_native_utf8)
{
if (!xlate_handle_hash)
{
@@ -141,6 +144,9 @@ svn_utf_initialize(apr_pool_t *pool)
apr_pool_cleanup_register(subpool, NULL, xlate_cleanup,
apr_pool_cleanup_null);
}
+
+ if (!assume_native_charset_is_utf8)
+ assume_native_charset_is_utf8 = assume_native_utf8;
}
/* Return a unique string key based on TOPAGE and FROMPAGE. TOPAGE and
@@ -442,7 +448,9 @@ static svn_error_t *
get_ntou_xlate_handle_node(xlate_handle_node_t **ret, apr_pool_t *pool)
{
return get_xlate_handle_node(ret, SVN_APR_UTF8_CHARSET,
- SVN_APR_LOCALE_CHARSET,
+ assume_native_charset_is_utf8
+ ? SVN_APR_UTF8_CHARSET
+ : SVN_APR_LOCALE_CHARSET,
SVN_UTF_NTOU_XLATE_HANDLE, pool);
}
@@ -455,7 +463,10 @@ get_ntou_xlate_handle_node(xlate_handle_
static svn_error_t *
get_uton_xlate_handle_node(xlate_handle_node_t **ret, apr_pool_t *pool)
{
- return get_xlate_handle_node(ret, SVN_APR_LOCALE_CHARSET,
+ return get_xlate_handle_node(ret,
+ assume_native_charset_is_utf8
+ ? SVN_APR_UTF8_CHARSET
+ : SVN_APR_LOCALE_CHARSET,
SVN_APR_UTF8_CHARSET,
SVN_UTF_UTON_XLATE_HANDLE, pool);
}
Modified: subversion/trunk/subversion/mod_dav_svn/mod_dav_svn.c
URL:
http://svn.apache.org/viewvc/subversion/trunk/subversion/mod_dav_svn/mod_dav_svn.c?rev=1239203&r1=1239202&r2=1239203&view=diff
==============================================================================
--- subversion/trunk/subversion/mod_dav_svn/mod_dav_svn.c (original)
+++ subversion/trunk/subversion/mod_dav_svn/mod_dav_svn.c Wed Feb 1 16:58:50
2012
@@ -22,6 +22,8 @@
* ====================================================================
*/
+#include <stdlib.h>
+
#include <apr_strings.h>
#include <httpd.h>
@@ -55,6 +57,7 @@
/* per-server configuration */
typedef struct server_conf_t {
const char *special_uri;
+ svn_boolean_t use_utf8;
} server_conf_t;
@@ -111,6 +114,8 @@ static int
init(apr_pool_t *p, apr_pool_t *plog, apr_pool_t *ptemp, server_rec *s)
{
svn_error_t *serr;
+ server_conf_t *conf;
+
ap_add_version_component(p, "SVN/" SVN_VER_NUMBER);
serr = svn_fs_initialize(p);
@@ -123,7 +128,8 @@ init(apr_pool_t *p, apr_pool_t *plog, ap
}
/* This returns void, so we can't check for error. */
- svn_utf_initialize(p);
+ conf = ap_get_module_config(s->module_config, &dav_svn_module);
+ svn_utf_initialize2(p, conf->use_utf8);
return OK;
}
@@ -510,6 +516,18 @@ SVNCompressionLevel_cmd(cmd_parms *cmd,
return NULL;
}
+static const char *
+SVNUseUTF8_cmd(cmd_parms *cmd, void *config, int arg)
+{
+ server_conf_t *conf;
+
+ conf = ap_get_module_config(cmd->server->module_config,
+ &dav_svn_module);
+ conf->use_utf8 = arg;
+
+ return NULL;
+}
+
/** Accessor functions for the module's configuration state **/
@@ -1021,6 +1039,11 @@ static const command_rec cmds[] =
"content over the network (0 for no compression, 9 for "
"maximum, 5 is default)."),
+ /* per server */
+ AP_INIT_FLAG("SVNUseUTF8",
+ SVNUseUTF8_cmd, NULL,
+ RSRC_CONF,
+ "use UTF-8 as native character encoding (default is ASCII)."),
{ NULL }
};
