Index: src/file.c
==================================================================
--- src/file.c
+++ src/file.c
@@ -64,23 +64,24 @@
** Fill stat buf with information received from stat() or lstat().
** lstat() is called on Unix if isWd is TRUE and allow-symlinks setting is on.
**
*/
static int fossil_stat(const char *zFilename, struct stat *buf, int isWd){
+ int rc;
#if !defined(_WIN32)
+ char *zMbcs = fossil_utf8_to_filename(zFilename);
if( isWd && g.allowSymlinks ){
- return lstat(zFilename, buf);
+ rc = lstat(zMbcs, buf);
}else{
- return stat(zFilename, buf);
+ rc = stat(zMbcs, buf);
}
#else
- int rc = 0;
- wchar_t *zMbcs = fossil_utf8_to_unicode(zFilename);
+ wchar_t *zMbcs = fossil_utf8_to_filename(zFilename);
rc = _wstati64(zMbcs, buf);
- fossil_unicode_free(zMbcs);
+#endif
+ fossil_filename_free(zMbcs);
return rc;
-#endif
}
/*
** Fill in the fileStat variable for the file named zFilename.
** If zFilename==0, then use the previous value of fileStat if
@@ -303,16 +304,17 @@
/*
** Wrapper around the access() system call.
*/
int file_access(const char *zFilename, int flags){
#ifdef _WIN32
- wchar_t *zMbcs = fossil_utf8_to_unicode(zFilename);
+ wchar_t *zMbcs = fossil_utf8_to_filename(zFilename);
int rc = _waccess(zMbcs, flags);
- fossil_unicode_free(zMbcs);
#else
- int rc = access(zFilename, flags);
+ char *zMbcs = fossil_utf8_to_filename(zFilename);
+ int rc = access(zMbcs, flags);
#endif
+ fossil_filename_free(zMbcs);
return rc;
}
/*
** Find an unused filename similar to zBase with zSuffix appended.
@@ -402,19 +404,20 @@
#if !defined(_WIN32)
struct timeval tv[2];
memset(tv, 0, sizeof(tv[0])*2);
tv[0].tv_sec = newMTime;
tv[1].tv_sec = newMTime;
- utimes(zFilename, tv);
+ char *zMbcs = fossil_utf8_to_filename(zFilename);
+ utimes(zMbcs, tv);
#else
struct _utimbuf tb;
- wchar_t *zMbcs = fossil_utf8_to_unicode(zFilename);
+ wchar_t *zMbcs = fossil_utf8_to_filename(zFilename);
tb.actime = newMTime;
tb.modtime = newMTime;
_wutime(zMbcs, &tb);
- fossil_unicode_free(zMbcs);
#endif
+ fossil_filename_free(zMbcs);
}
/*
** COMMAND: test-set-mtime
**
@@ -441,16 +444,17 @@
/*
** Delete a file.
*/
void file_delete(const char *zFilename){
#ifdef _WIN32
- wchar_t *z = fossil_utf8_to_unicode(zFilename);
+ wchar_t *z = fossil_utf8_to_filename(zFilename);
_wunlink(z);
- fossil_unicode_free(z);
#else
+ char *z = fossil_utf8_to_filename(zFilename);
unlink(zFilename);
#endif
+ fossil_filename_free(z);
}
/*
** Create the directory named in the argument, if it does not already
** exist. If forceFlag is 1, delete any prior non-directory object
@@ -464,18 +468,18 @@
if( !forceFlag ) return 1;
file_delete(zName);
}
if( rc!=1 ){
#if defined(_WIN32)
- int rc;
- wchar_t *zMbcs = fossil_utf8_to_unicode(zName);
+ wchar_t *zMbcs = fossil_utf8_to_filename(zName);
rc = _wmkdir(zMbcs);
- fossil_unicode_free(zMbcs);
- return rc;
#else
- return mkdir(zName, 0755);
+ char *zMbcs = fossil_utf8_to_filename(zName);
+ rc = mkdir(zName, 0755);
#endif
+ fossil_filename_free(zMbcs);
+ return rc;
}
return 0;
}
/*
@@ -483,11 +487,10 @@
** a file in a repository. Valid filenames follow all of the
** following rules:
**
** * Does not begin with "/"
** * Does not contain any path element named "." or ".."
-** * Does not contain any of these characters in the path: "\"
** * Does not end with "/".
** * Does not contain two or more "/" characters in a row.
** * Contains at least one character
**
** Invalid UTF8 characters result in a false return if bStrictUtf8 is
@@ -547,12 +550,10 @@
if( (z[++i]&0xc0)!=0x80 ){
/* Invalid second continuation byte */
return 0;
}
}
- }else if( bStrictUtf8 && (c=='\\') ){
- return 0;
}
if( c=='/' ){
if( z[i+1]=='/' ) return 0;
if( z[i+1]=='.' ){
if( z[i+2]=='/' || z[i+2]==0 ) return 0;
@@ -580,11 +581,11 @@
}
/*
** Simplify a filename by
**
-** * Convert all \ into / on windows
+** * Convert all \ into / on windows and cygwin
** * removing any trailing and duplicate /
** * removing /./
** * removing /A/../
**
** Changes are made in-place. Return the new name length.
@@ -593,12 +594,12 @@
*/
int file_simplify_name(char *z, int n, int slash){
int i, j;
if( n<0 ) n = strlen(z);
- /* On windows convert all \ characters to / */
-#if defined(_WIN32)
+ /* On windows/cygwin convert all \ characters to / */
+#if defined(_WIN32) || defined(__CYGWIN__)
for(i=0; i<n; i++){
if( z[i]=='\\' ) z[i] = '/';
}
#endif
@@ -705,11 +706,13 @@
** Return true if zPath is an absolute pathname. Return false
** if it is relative.
*/
int file_is_absolute_path(const char *zPath){
if( zPath[0]=='/'
-#if defined(_WIN32)
+#if defined(__CYGWIN__)
+ || zPath[0]=='\\'
+#elif defined(_WIN32)
|| zPath[0]=='\\'
|| (strlen(zPath)>3 && zPath[1]==':'
&& (zPath[2]=='\\' || zPath[2]=='/'))
#endif
){
@@ -1137,14 +1140,14 @@
** Like fopen() but always takes a UTF8 argument.
*/
FILE *fossil_fopen(const char *zName, const char *zMode){
#ifdef _WIN32
wchar_t *uMode = fossil_utf8_to_unicode(zMode);
- wchar_t *uName = fossil_utf8_to_unicode(zName);
+ wchar_t *uName = fossil_utf8_to_filename(zName);
FILE *f = _wfopen(uName, uMode);
- fossil_unicode_free(uName);
+ fossil_filename_free(uName);
fossil_unicode_free(uMode);
#else
FILE *f = fopen(zName, zMode);
#endif
return f;
}
Index: src/rebuild.c
==================================================================
--- src/rebuild.c
+++ src/rebuild.c
@@ -841,11 +841,11 @@
Blob aContent; /* content of the just read artifact */
static int nFileRead = 0;
void *zUnicodePath;
char *zUtf8Name;
- zUnicodePath = fossil_utf8_to_unicode(zPath);
+ zUnicodePath = fossil_utf8_to_filename(zPath);
d = opendir(zUnicodePath);
if( d ){
while( (pEntry=readdir(d))!=0 ){
Blob path;
char *zSubpath;
@@ -875,11 +875,11 @@
closedir(d);
}else {
fossil_panic("encountered error %d while trying to open \"%s\".",
errno, g.argv[3]);
}
- fossil_unicode_free(zUnicodePath);
+ fossil_filename_free(zUnicodePath);
}
/*
** COMMAND: reconstruct*
**
Index: src/utf8.c
==================================================================
--- src/utf8.c
+++ src/utf8.c
@@ -111,19 +111,36 @@
** Translate text from the filename character set into
** to precomposed UTF8. Return a pointer to the translated text.
** Call fossil_filename_free() to deallocate any memory used to store the
** returned pointer when done.
*/
-char *fossil_filename_to_utf8(const void *zFilename){
+char *fossil_filename_to_utf8(void *zFilename){
#if defined(_WIN32)
- int nByte = WideCharToMultiByte(CP_UTF8, 0, zFilename, -1, 0, 0, 0, 0);
- char *zUtf = sqlite3_malloc( nByte );
+ int nByte;
+ char *zUtf;
+ WCHAR *wUnicode = zFilename;
+ while( *wUnicode != 0 ){
+ if ( (*wUnicode & 0xFF80) == 0xF000 ){
+ WCHAR converted = (*wUnicode & 0x7F);
+ /* Only really convert it when the resulting char is in the given range*/
+ if ( (converted < 32) || wcschr(L"\"*<>?|:", converted) ){
+ *wUnicode = converted;
+ }
+ }
+ ++wUnicode;
+ }
+ nByte = WideCharToMultiByte(CP_UTF8, 0, zFilename, -1, 0, 0, 0, 0);
+ zUtf = sqlite3_malloc( nByte );
if( zUtf==0 ){
return 0;
}
WideCharToMultiByte(CP_UTF8, 0, zFilename, -1, zUtf, nByte, 0, 0);
return zUtf;
+#elif defined(__CYGWIN__)
+ char *zOut;
+ zOut = fossil_strdup(zFilename);
+ return zOut;
#elif defined(__APPLE__) && !defined(WITHOUT_ICONV)
char *zIn = (char*)zFilename;
char *zOut;
iconv_t cd;
size_t n, x;
@@ -149,19 +166,71 @@
return zOut;
#else
return (char *)zFilename; /* No-op on non-mac unix */
#endif
}
+
+/*
+** Translate UTF8 to unicode for use in filename translations.
+** Return a pointer to the translated text.. Call fossil_filename_free()
+** to deallocate any memory used to store the returned pointer when done.
+**
+** On Windows, characters in the range U+0001 to U+0031 and the
+** characters '"', '*', ':', '<', '>', '?' and '|' are invalid
+** to be used. Therefore, translated those to characters in the
+** (private use area), in the range U+F001 - U+F07F, so those
+** characters never arrive in any Windows API. The filenames might
+** look strange in Windows explorer, but in the cygwin shell
+** everything looks as expected.
+**
+** See: <http://cygwin.com/cygwin-ug-net/using-specialnames.html>
+**
+*/
+void *fossil_utf8_to_filename(const char *zUtf8){
+#ifdef _WIN32
+ WCHAR *zUnicode = fossil_utf8_to_unicode(zUtf8);
+ WCHAR *wUnicode = zUnicode;
+ /* If path starts with "<drive>:/" or "<drive>:\", don't translate the ':' */
+ if( fossil_isalpha(zUtf8[0]) && zUtf8[1]==':'
+ && (zUtf8[2]=='\\' || zUtf8[2]=='/')) {
+ zUnicode[2] = '\\';
+ wUnicode += 3;
+ }
+ while( *wUnicode != '\0' ){
+ if ( (*wUnicode < 32) || wcschr(L"\"*<>?|:", *wUnicode) ){
+ *wUnicode |= 0xF000;
+ }else if( *wUnicode == '/' ){
+ *wUnicode = '\\';
+ }
+ ++wUnicode;
+ }
+
+ return zUnicode;
+#elif defined(__CYGWIN__)
+ char *zPath = fossil_strdup(zUtf8);
+ char *p = zPath;
+ while( (*p = *zUtf8++) != 0){
+ if (*p++ == '\\' ) {
+ p[-1] = '/';
+ }
+ }
+ return zPath;
+#elif defined(__APPLE__) && !defined(WITHOUT_ICONV)
+ return fossil_strdup(zUtf8);
+#else
+ return (void *)zUtf8; /* No-op on unix */
+#endif
+}
/*
** Deallocate any memory that was previously allocated by
-** fossil_filename_to_utf8().
+** fossil_filename_to_utf8() or fossil_utf8_to_filename().
*/
-void fossil_filename_free(char *pOld){
+void fossil_filename_free(void *pOld){
#if defined(_WIN32)
sqlite3_free(pOld);
-#elif defined(__APPLE__) && !defined(WITHOUT_ICONV)
+#elif (defined(__APPLE__) && !defined(WITHOUT_ICONV)) || defined(__CYGWIN__)
fossil_free(pOld);
#else
/* No-op on all other unix */
#endif
}
Index: src/vfile.c
==================================================================
--- src/vfile.c
+++ src/vfile.c
@@ -459,11 +459,11 @@
);
}
depth++;
zDir = blob_str(pPath);
- zNative = fossil_utf8_to_unicode(zDir);
+ zNative = fossil_utf8_to_filename(zDir);
d = opendir(zNative);
if( d ){
while( (pEntry=readdir(d))!=0 ){
char *zPath;
char *zUtf8;
@@ -491,11 +491,11 @@
fossil_filename_free(zUtf8);
blob_resize(pPath, origSize);
}
closedir(d);
}
- fossil_unicode_free(zNative);
+ fossil_filename_free(zNative);
depth--;
if( depth==0 ){
db_finalize(&ins);
}