Merge branch 1.0-bugfix

include/vlc_url.h

@@ -48,8 +48,9 @@ struct vlc_url_t
 VLC_EXPORT( char *, unescape_URI_duplicate, ( const char *psz ) );
 VLC_EXPORT( void, unescape_URI, ( char *psz ) );
 VLC_EXPORT( char *, decode_URI_duplicate, ( const char *psz ) );
-VLC_EXPORT( void, decode_URI, ( char *psz ) );
+VLC_EXPORT( char *, decode_URI, ( char *psz ) );
 VLC_EXPORT( char *, encode_URI_component, ( const char *psz ) );
+VLC_EXPORT( char *, make_URI, ( const char *path ) );
 
 /*****************************************************************************
  * vlc_UrlParse:

modules/misc/playlist/xspf.c

@@ -34,11 +34,14 @@
 #include <vlc_playlist.h>
 #include <vlc_input.h>
 #include <vlc_strings.h>
-#include <vlc_charset.h>
+#include <vlc_url.h>
 #include "xspf.h"
 
 #include <assert.h>
 
+static void xspf_export_item( playlist_item_t *, FILE *, int * );
+static void xspf_extension_item( playlist_item_t *, FILE *, int * );
+
 /**
  * \brief Prints the XSPF header to file, writes each item by xspf_export_item()
  * and closes the open xml elements
@@ -139,7 +142,7 @@ static void xspf_export_item( playlist_item_t *p_item, FILE *p_file,
 
     if( psz_uri && *psz_uri )
     {
-        psz = assertUTF8URI( psz_uri );
+        psz = make_URI( psz_uri );
         fprintf( p_file, "\t\t\t<location>%s</location>\n", psz );
         free( psz );
     }
@@ -209,7 +212,7 @@ static void xspf_export_item( playlist_item_t *p_item, FILE *p_file,
     if( psz == NULL ) psz = strdup( "" );
     if( !EMPTY_STR( psz ) )
     {
-        psz_uri = assertUTF8URI( psz );
+        psz_uri = make_URI( psz );
         fprintf( p_file, "\t\t\t<image>%s</image>\n", psz_uri );
         free( psz_uri );
     }
@@ -284,89 +287,3 @@ static void xspf_extension_item( playlist_item_t *p_item, FILE *p_file,
 
     return;
 }
-
-/**
- * \param psz_name the location of the media ressource (e.g. local file,
- *        device, network stream, etc.)
- * \return a new char buffer which asserts that the location is valid UTF-8
- *         and a valid URI
- * \note the returned buffer must be freed, when it isn't used anymore
- */
-static char *assertUTF8URI( char *psz_name )
-{
-    char *psz_ret = NULL;              /**< the new result buffer to return */
-    char *psz_s = NULL, *psz_d = NULL; /**< src & dest pointers for URI conversion */
-    bool b_uri_is_file = false; /**< we do additional %-encoding if the URI is a file:// one */
-
-    if( !psz_name || !*psz_name )
-        return NULL;
-
-    /* check that string is valid UTF-8 */
-    /* XXX: Why do we even need to do that ? (all strings in core are UTF-8 encoded */
-    if( !( psz_s = EnsureUTF8( psz_name ) ) )
-        return NULL;
-
-    /* max. 3x for URI conversion (percent escaping) and
-       8 bytes for "file://" and NULL-termination */
-    psz_ret = (char *)malloc( strlen(psz_name)*6*3+8 );
-    if( !psz_ret )
-        return NULL;
-
-    /** \todo check for a valid scheme part preceding the colon */
-    if( strstr( psz_s, "://") != NULL )
-    {
-        size_t i_delim = strcspn( psz_s, ":" );
-        i_delim++; /* skip the ':' */
-        strncpy( psz_ret, psz_s, i_delim );
-        psz_d = psz_ret + i_delim;
-
-        if( !strncmp( psz_s, "file://", 7 ) )
-            b_uri_is_file = true;
-
-        psz_s += i_delim;
-    }
-    /* assume "file" scheme if no scheme-part is included */
-    else
-    {
-        strcpy( psz_ret, "file://" );
-        psz_d = psz_ret + 7;
-        b_uri_is_file = true;
-    }
-
-    while( *psz_s )
-    {
-        /* percent-encode all non-ASCII and the XML special characters and the percent sign itself */
-        if( *psz_s & B10000000 ||
-            *psz_s == '<' ||
-            *psz_s == '>' ||
-            *psz_s == '&' ||
-            *psz_s == ' ' ||
-            *psz_s == '+' ||
-            *psz_s == '%' ||
-            *psz_s == '\\' ||
-            ( b_uri_is_file && (
-            *psz_s == ':' ||
-            *psz_s == '"' ||
-            *psz_s == '?' ||
-            *psz_s == '#' ||
-            *psz_s == '[' ||
-            *psz_s == ']' ||
-            *psz_s == '@' )
-            )
-          )
-        {
-            *psz_d++ = '%';
-            *psz_d++ = hexchars[(*psz_s >> 4) & B00001111];
-            *psz_d++ = hexchars[*psz_s & B00001111];
-        }
-        else
-        {
-            *psz_d++ = *psz_s;
-        }
-
-        psz_s++;
-    }
-    *psz_d = '\0';
-
-    return (char *)realloc( psz_ret, strlen( psz_ret ) + 1 );
-}

modules/misc/playlist/xspf.h

@@ -33,6 +33,3 @@ const char hexchars[16] = "0123456789ABCDEF";
 
 /* prototypes */
 int xspf_export_playlist( vlc_object_t * );
-static void xspf_export_item( playlist_item_t *, FILE *, int * );
-static void xspf_extension_item( playlist_item_t *, FILE *, int * );
-static char *assertUTF8URI( char * );

src/libvlccore.sym

@@ -216,6 +216,7 @@ libvlc_InternalInit
 libvlc_InternalWait
 libvlc_Quit
 LocaleFree
+make_URI
 mdate
 module_config_free
 module_config_get

src/test/url.c

@@ -29,6 +29,7 @@
 
 #include <stdio.h>
 #include <stdlib.h>
+#include <assert.h>
 
 typedef char * (*conv_t) (const char *);
 
@@ -60,8 +61,15 @@ static inline void test_b64 (const char *in, const char *out)
     test (vlc_b64_encode, in, out);
 }
 
+static inline void test_path (const char *in, const char *out)
+{
+    test (make_URI, in, out);
+}
+
 int main (void)
 {
+    int val;
+
     (void)setvbuf (stdout, NULL, _IONBF, 0);
     test_decode ("this_should_not_be_modified_1234",
                  "this_should_not_be_modified_1234");
@@ -93,5 +101,24 @@ int main (void)
     test_b64 ("fooba", "Zm9vYmE=");
     test_b64 ("foobar", "Zm9vYmFy");
 
+    /* Path test */
+    test_path ("file:///", "file:///");
+    test_path ("http://www.example.com/%7Ejohn/",
+               "http://www.example.com/%7Ejohn/");
+    test_path ("/", "file:///");
+    test_path ("/home/john/", "file:///home/john/");
+    test_path ("/home/john/music.ogg", "file:///home/john/music.ogg");
+    //test_path ("\\\\server/pub/music.ogg", "file://server/pub/music.ogg");
+
+    /*int fd = open (".", O_RDONLY);
+    assert (fd != -1);*/
+    val = chdir ("/tmp");
+    assert (val != -1);
+    test_path ("movie.ogg", "file:///tmp/movie.ogg");
+    test_path (".", "file:///tmp/.");
+    test_path ("", "file:///tmp/");
+    /*val = fchdir (fd);
+    assert (val != -1);*/
+
     return 0;
 }

src/text/strings.c

@@ -2,6 +2,7 @@
  * strings.c: String related functions
  *****************************************************************************
  * Copyright (C) 2006 the VideoLAN team
+ * Copyright (C) 2008-2009 Rémi Denis-Courmont
  * $Id$
  *
  * Authors: Antoine Cellerier <dionoea at videolan dot org>
@@ -130,7 +131,7 @@ void unescape_URI( char *psz )
 }
 
 /**
- * Decode encoded URI string
+ * Decode encoded URI component. See also decode_URI().
  * \return decoded duplicated string
  */
 char *decode_URI_duplicate( const char *psz )
@@ -141,14 +142,23 @@ char *decode_URI_duplicate( const char *psz )
 }
 
 /**
- * Decode encoded URI string in place
- * \return nothing
+ * Decode an encoded URI component in place.
+ * <b>This function does NOT decode entire URIs.</b>
+ * It decodes components (e.g. host name, directory, file name).
+ * Decoded URIs do not exist in the real world (see RFC3986 §2.4).
+ * Complete URIs are always "encoded" (or they are syntaxically invalid).
+ *
+ * Note that URI encoding is different from Javascript escaping. Especially,
+ * white spaces and Unicode non-ASCII code points are encoded differently.
+ *
+ * \return psz on success, NULL if it was not properly encoded
  */
-void decode_URI( char *psz )
+char *decode_URI( char *psz )
 {
     unsigned char *in = (unsigned char *)psz, *out = in, c;
+
     if( psz == NULL )
-        return;
+        return NULL;
 
     while( ( c = *in++ ) != '\0' )
     {
@@ -160,14 +170,14 @@ void decode_URI( char *psz )
 
                 if( ( ( hex[0] = *in++ ) == 0 )
                  || ( ( hex[1] = *in++ ) == 0 ) )
-                    return;
+                    return NULL;
 
                 hex[2] = '\0';
                 *out++ = (unsigned char)strtoul( hex, NULL, 0x10 );
                 break;
             }
 
-            case '+':
+            case '+': /* This is HTTP forms, not URI decoding... */
                 *out++ = ' ';
                 break;
 
@@ -182,6 +192,7 @@ void decode_URI( char *psz )
     }
     *out = '\0';
     EnsureUTF8( psz );
+    return psz;
 }
 
 static inline bool isurisafe( int c )
@@ -193,23 +204,13 @@ static inline bool isurisafe( int c )
             || ( strchr( "-._~", c ) != NULL );
 }
 
-/**
- * Encodes an URI component (RFC3986 §2).
- *
- * @param psz_uri nul-terminated UTF-8 representation of the component.
- * Obviously, you can't pass an URI containing a nul character, but you don't
- * want to do that, do you?
- *
- * @return encoded string (must be free()'d), or NULL for ENOMEM.
- */
-char *encode_URI_component( const char *psz_uri )
+static char *encode_URI_bytes (const char *psz_uri, size_t len)
 {
-    char *psz_enc = malloc ((3 * strlen (psz_uri)) + 1), *out = psz_enc;
-
+    char *psz_enc = malloc (3 * len + 1), *out = psz_enc;
     if (psz_enc == NULL)
         return NULL;
 
-    while (*psz_uri)
+    for (size_t i = 0; i < len; i++)
     {
         static const char hex[16] = "0123456789ABCDEF";
         uint8_t c = *psz_uri;
@@ -232,6 +233,21 @@ char *encode_URI_component( const char *psz_uri )
     return out ? out : psz_enc; /* realloc() can fail (safe) */
 }
 
+/**
+ * Encodes an URI component (RFC3986 §2).
+ *
+ * @param psz_uri nul-terminated UTF-8 representation of the component.
+ * Obviously, you can't pass an URI containing a nul character, but you don't
+ * want to do that, do you?
+ *
+ * @return encoded string (must be free()'d), or NULL for ENOMEM.
+ */
+char *encode_URI_component( const char *psz_uri )
+{
+    return encode_URI_bytes (psz_uri, strlen (psz_uri));
+}
+
+
 static const struct xml_entity_s
 {
     char    psz_entity[8];
@@ -1120,3 +1136,78 @@ void path_sanitize( char *str )
         str++;
     }
 }
+
+#include <vlc_url.h>
+
+/**
+ * Convert a file path to an URI. If already an URI, do nothing.
+ */
+char *make_URI (const char *path)
+{
+    if (path == NULL)
+        return NULL;
+    if (strstr (path, "://") != NULL)
+        return strdup (path); /* Already an URI */
+    /* Note: VLC cannot handle URI schemes without double slash after the
+     * scheme name (such as mailto: or news:). */
+
+    char *buf;
+#ifdef WIN32
+    if (isalpha (path[0]) && (path[1] == ':'))
+    {
+        if (asprintf (&buf, "file:///%c:", path[0]) == -1)
+            buf = NULL;
+        path += 2;
+    }
+    else
+#endif
+#if 0
+    /* Windows UNC paths (file://host/share/path instead of file:///path) */
+    if (!strncmp (path, "\\\\", 2))
+    {
+        path += 2;
+        buf = strdup ("file://");
+    }
+    else
+#endif
+    if (path[0] != DIR_SEP_CHAR)
+    {   /* Relative path: prepend the current working directory */
+        char cwd[PATH_MAX];
+
+        if (getcwd (cwd, sizeof (cwd)) == NULL) /* FIXME: UTF8? */
+            return NULL;
+        if (asprintf (&buf, "%s/%s", cwd, path) == -1)
+            return NULL;
+        char *ret = make_URI (buf);
+        free (buf);
+        return ret;
+    }
+    else
+        buf = strdup ("file://");
+    if (buf == NULL)
+        return NULL;
+
+    assert (path[0] == DIR_SEP_CHAR);
+
+    /* Absolute file path */
+    for (const char *ptr = path + 1;; ptr++)
+    {
+        size_t len = strcspn (ptr, DIR_SEP);
+        char *component = encode_URI_bytes (ptr, len);
+        if (component == NULL)
+        {
+            free (buf);
+            return NULL;
+        }
+        char *uri;
+        int val = asprintf (&uri, "%s/%s", buf, component);
+        free (component);
+        free (buf);
+        if (val == -1)
+            return NULL;
+        buf = uri;
+        ptr += len;
+        if (*ptr == '\0')
+            return buf;
+    }
+}