Changeset 2780

Timestamp:

Aug 15, 2007, 11:02:56 PM (15 years ago)

Author:

joshe

Message:

Rework utf-8 string concat function a little and make it more widely available.

Location:

trunk/libtransmission

Files:

• metainfo.c (modified) (7 diffs)
• utils.c (modified) (1 diff)
• utils.h (modified) (1 diff)

trunk/libtransmission/metainfo.c

@@ -57 +57 @@
 static int parseFiles( tr_info_t * inf, benc_val_t * name,
                        benc_val_t * files, benc_val_t * length );
-static void strcatUTF8( char *, int, const char *, int );
 
 /***********************************************************************
@@ -206 +205 @@
     if( NULL != val && TYPE_STR == val->type )
     {
-        strcatUTF8( inf->comment, sizeof( inf->comment ), val->val.s.s, 0 );
+        strlcat_utf8( inf->comment, val->val.s.s, sizeof( inf->comment ), 0 );
     }
     
@@ -213 +212 @@
     if( NULL != val && TYPE_STR == val->type )
     {
-        strcatUTF8( inf->creator, sizeof( inf->creator ), val->val.s.s, 0 );
+        strlcat_utf8( inf->creator, val->val.s.s, sizeof( inf->creator ), 0 );
     }
     
@@ -371 +370 @@
     }
 
-    strcatUTF8( buf, size, prefix, 0 );
+    strlcat_utf8( buf, prefix, size, 0 );
     for( ii = 0; jj > ii; ii++ )
     {
-        strcatUTF8( buf, size, TR_PATH_DELIMITER_STR, 0 );
-        strcatUTF8( buf, size, list[ii], 1 );
+        strlcat_utf8( buf, TR_PATH_DELIMITER_STR, size, 0 );
+        strlcat_utf8( buf, list[ii], size, TR_PATH_DELIMITER );
     }
     free( list );
@@ -691 +690 @@
     }
 
-    strcatUTF8( inf->name, sizeof( inf->name ), name->val.s.s, 1 );
+    strlcat_utf8( inf->name, name->val.s.s, sizeof( inf->name ),
+                  TR_PATH_DELIMITER );
     if( '\0' == inf->name[0] )
     {
@@ -745 +745 @@
         }
 
-        strcatUTF8( inf->files[0].name, sizeof( inf->files[0].name ),
-                    name->val.s.s, 1 );
+        strlcat_utf8( inf->files[0].name, name->val.s.s,
+                      sizeof( inf->files[0].name ), TR_PATH_DELIMITER );
 
         inf->files[0].length = length->val.i;
@@ -760 +760 @@
     return TR_OK;
 }
-
-/***********************************************************************
- * strcatUTF8
- ***********************************************************************
- * According to the official specification, all strings in the torrent
- * file are supposed to be UTF-8 encoded. However, there are
- * non-compliant torrents around... If we encounter an invalid UTF-8
- * character, we assume it is ISO 8859-1 and convert it to UTF-8.
- **********************************************************************/
-#define WANTBYTES( want, got ) \
-    if( (want) > (got) ) { return; } else { (got) -= (want); }
-static void strcatUTF8( char * s, int len, const char * append, int deslash )
-{
-    const char * p;
-
-    /* don't overwrite the nul at the end */
-    len--;
-
-    /* Go to the end of the destination string */
-    while( s[0] )
-    {
-        s++;
-        len--;
-    }
-
-    /* Now start appending, converting on the fly if necessary */
-    for( p = append; p[0]; )
-    {
-        /* skip over / if requested */
-        if( deslash && '/' == p[0] )
-        {
-            p++;
-            continue;
-        }
-
-        if( !( p[0] & 0x80 ) )
-        {
-            /* ASCII character */
-            WANTBYTES( 1, len );
-            *(s++) = *(p++);
-            continue;
-        }
-
-        if( ( p[0] & 0xE0 ) == 0xC0 && ( p[1] & 0xC0 ) == 0x80 )
-        {
-            /* 2-bytes UTF-8 character */
-            WANTBYTES( 2, len );
-            *(s++) = *(p++); *(s++) = *(p++);
-            continue;
-        }
-
-        if( ( p[0] & 0xF0 ) == 0xE0 && ( p[1] & 0xC0 ) == 0x80 &&
-            ( p[2] & 0xC0 ) == 0x80 )
-        {
-            /* 3-bytes UTF-8 character */
-            WANTBYTES( 3, len );
-            *(s++) = *(p++); *(s++) = *(p++);
-            *(s++) = *(p++);
-            continue;
-        }
-
-        if( ( p[0] & 0xF8 ) == 0xF0 && ( p[1] & 0xC0 ) == 0x80 &&
-            ( p[2] & 0xC0 ) == 0x80 && ( p[3] & 0xC0 ) == 0x80 )
-        {
-            /* 4-bytes UTF-8 character */
-            WANTBYTES( 4, len );
-            *(s++) = *(p++); *(s++) = *(p++);
-            *(s++) = *(p++); *(s++) = *(p++);
-            continue;
-        }
-
-        /* ISO 8859-1 -> UTF-8 conversion */
-        WANTBYTES( 2, len );
-        *(s++) = 0xC0 | ( ( *p & 0xFF ) >> 6 );
-        *(s++) = 0x80 | ( *(p++) & 0x3F );
-    }
-}

trunk/libtransmission/utils.c

@@ -730 +730 @@
 #endif
 }
+
+#define WANTBYTES( want, got ) \
+    if( (want) > (got) ) { return; } else { (got) -= (want); }
+void
+strlcat_utf8( void * dest, const void * src, size_t len, char skip )
+{
+    char       * s      = dest;
+    const char * append = src;
+    const char * p;
+
+    /* don't overwrite the nul at the end */
+    len--;
+
+    /* Go to the end of the destination string */
+    while( s[0] )
+    {
+        s++;
+        len--;
+    }
+
+    /* Now start appending, converting on the fly if necessary */
+    for( p = append; p[0]; )
+    {
+        /* skip over the requested character */
+        if( skip == p[0] )
+        {
+            p++;
+            continue;
+        }
+
+        if( !( p[0] & 0x80 ) )
+        {
+            /* ASCII character */
+            WANTBYTES( 1, len );
+            *(s++) = *(p++);
+            continue;
+        }
+
+        if( ( p[0] & 0xE0 ) == 0xC0 && ( p[1] & 0xC0 ) == 0x80 )
+        {
+            /* 2-bytes UTF-8 character */
+            WANTBYTES( 2, len );
+            *(s++) = *(p++); *(s++) = *(p++);
+            continue;
+        }
+
+        if( ( p[0] & 0xF0 ) == 0xE0 && ( p[1] & 0xC0 ) == 0x80 &&
+            ( p[2] & 0xC0 ) == 0x80 )
+        {
+            /* 3-bytes UTF-8 character */
+            WANTBYTES( 3, len );
+            *(s++) = *(p++); *(s++) = *(p++);
+            *(s++) = *(p++);
+            continue;
+        }
+
+        if( ( p[0] & 0xF8 ) == 0xF0 && ( p[1] & 0xC0 ) == 0x80 &&
+            ( p[2] & 0xC0 ) == 0x80 && ( p[3] & 0xC0 ) == 0x80 )
+        {
+            /* 4-bytes UTF-8 character */
+            WANTBYTES( 4, len );
+            *(s++) = *(p++); *(s++) = *(p++);
+            *(s++) = *(p++); *(s++) = *(p++);
+            continue;
+        }
+
+        /* ISO 8859-1 -> UTF-8 conversion */
+        WANTBYTES( 2, len );
+        *(s++) = 0xC0 | ( ( *p & 0xFF ) >> 6 );
+        *(s++) = 0x80 | ( *(p++) & 0x3F );
+    }
+}
+
+size_t
+bufsize_utf8( const void * vstr, int * changed )
+{
+    const char * str = vstr;
+    size_t       ii, grow;
+
+    if( NULL != changed )
+        *changed = 0;
+
+    ii   = 0;
+    grow = 1;
+    while( '\0' != str[ii] )
+    {
+        if( !( str[ii] & 0x80 ) )
+            /* ASCII character */
+            ii++;
+        else if( ( str[ii]   & 0xE0 ) == 0xC0 && ( str[ii+1] & 0xC0 ) == 0x80 )
+            /* 2-bytes UTF-8 character */
+            ii += 2;
+        else if( ( str[ii]   & 0xF0 ) == 0xE0 && ( str[ii+1] & 0xC0 ) == 0x80 &&
+                 ( str[ii+2] & 0xC0 ) == 0x80 )
+            /* 3-bytes UTF-8 character */
+            ii += 3;
+        else if( ( str[ii]   & 0xF8 ) == 0xF0 && ( str[ii+1] & 0xC0 ) == 0x80 &&
+                 ( str[ii+2] & 0xC0 ) == 0x80 && ( str[ii+3] & 0xC0 ) == 0x80 )
+            /* 4-bytes UTF-8 character */
+            ii += 4;
+        else
+        {
+            /* ISO 8859-1 -> UTF-8 conversion */
+            ii++;
+            grow++;
+            if( NULL != changed )
+                *changed = 1;
+        }
+    }
+
+    return ii + grow;
+}

trunk/libtransmission/utils.h

@@ -89 +89 @@
 void tr_wait( uint64_t delay_milliseconds );
 
+/***********************************************************************
+ * strlcat_utf8
+ ***********************************************************************
+ * According to the official specification, all strings in the torrent
+ * file are supposed to be UTF-8 encoded. However, there are
+ * non-compliant torrents around... If we encounter an invalid UTF-8
+ * character, we assume it is ISO 8859-1 and convert it to UTF-8.
+ **********************************************************************/
+void strlcat_utf8( void *, const void *, size_t, char );
+size_t bufsize_utf8( const void *, int * );
+
 /***
 ****