Decode URL obtained from HTML tags.

diff --git a/src/libserver/html.c b/src/libserver/html.c
index 539ff555dc6475b96ec99252eac7b543197ef682..fdb6a11db465391b990a689a66568fb3fc459d29 100644 (file)
--- a/src/libserver/html.c
+++ b/src/libserver/html.c
@@ -867,6 +867,7 @@ parse_tag_url (struct rspamd_task *task,
 
                url_text = rspamd_mempool_alloc (task->task_pool, len + 1);
                rspamd_strlcpy (url_text, c, len + 1);
+               rspamd_url_unescape (url_text);
                decode_entitles (url_text, NULL);
 
                if (g_ascii_strncasecmp (url_text, "http://",

diff --git a/src/libserver/url.c b/src/libserver/url.c
index df4e3102dab4358b6f3f4a99b904fe9efedaf170..96b422ae8fe4cfeac5ffec1fa716401866072dd0 100644 (file)
--- a/src/libserver/url.c
+++ b/src/libserver/url.c
@@ -953,39 +953,32 @@ url_calculate_escaped_hostlen (gchar *host, guint hostlen)
        return result;
 }
 
-/* URL-unescape the string S.
-
-   This is done by transforming the sequences "%HH" to the character
-   represented by the hexadecimal digits HH.  If % is not followed by
-   two hexadecimal digits, it is inserted literally.
-
-   The transformation is done in place.  If you need the original
-   string intact, make a copy before calling this function.  */
-
-static void
-url_unescape (gchar *s)
+void
+rspamd_url_unescape (gchar *s)
 {
        gchar *t = s;                           /* t - tortoise */
        gchar *h = s;                           /* h - hare     */
 
        for (; *h; h++, t++) {
                if (*h != '%') {
-copychar:
                        *t = *h;
                }
                else {
                        gchar c;
-                       /* Do nothing if '%' is not followed by two hex digits. */
                        if (!h[1] || !h[2] ||
-                               !(g_ascii_isxdigit (h[1]) && g_ascii_isxdigit (h[2])))
-                               goto copychar;
-                       c = X2DIGITS_TO_NUM (h[1], h[2]);
-                       /* Don't unescape %00 because there is no way to insert it
-                        * into a C string without effectively truncating it. */
-                       if (c == '\0')
-                               goto copychar;
-                       *t = c;
-                       h += 2;
+                               !(g_ascii_isxdigit (h[1]) && g_ascii_isxdigit (h[2]))) {
+                               *t = *h;
+                       }
+                       else {
+                               c = X2DIGITS_TO_NUM (h[1], h[2]);
+                               if (c != '\0') {
+                                       *t = c;
+                                       h += 2;
+                               }
+                               else {
+                                       *t = *h;
+                               }
+                       }
                }
        }
        *t = '\0';
@@ -1082,12 +1075,8 @@ char_needs_escaping (const gchar *p)
        return FALSE;
 }
 
-/* Translate a %-escaped (but possibly non-conformant) input string S
-   into a %-escaped (and conformant) output string.
- */
-
 static gchar *
-reencode_escapes (gchar *s, rspamd_mempool_t * pool)
+rspamd_url_reencode_escapes (gchar *s, rspamd_mempool_t * pool)
 {
        const gchar *p1;
        gchar *newstr, *p2;
@@ -1131,31 +1120,6 @@ reencode_escapes (gchar *s, rspamd_mempool_t * pool)
        return newstr;
 }
 
-/* Unescape CHR in an otherwise escaped STR.  Used to selectively
-   escaping of certain characters, such as "/" and ":".  Returns a
-   count of unescaped chars.  */
-
-static void
-unescape_single_char (gchar *str, gchar chr)
-{
-       const gchar c1 = XNUM_TO_DIGIT (chr >> 4);
-       const gchar c2 = XNUM_TO_DIGIT (chr & 0xf);
-       gchar *h = str;                             /* hare */
-       gchar *t = str;                             /* tortoise */
-
-       for (; *h; h++, t++) {
-               if (h[0] == '%' && h[1] == c1 && h[2] == c2) {
-                       *t = chr;
-                       h += 2;
-               }
-               else {
-                       *t = *h;
-               }
-       }
-       *t = '\0';
-}
-
-
 /*
  * Resolve "." and ".." elements of PATH by destructively modifying
  * PATH and return non-zero if PATH has been modified, zero otherwise.
@@ -1234,7 +1198,7 @@ parse_uri (struct uri *uri, gchar *uristring, rspamd_mempool_t * pool)
        if (!*uristring)
                return URI_ERRNO_EMPTY;
 
-       uri->string = reencode_escapes (uristring, pool);
+       uri->string = rspamd_url_reencode_escapes (uristring, pool);
        msg_debug ("reencoding escapes in original url: '%s'", struri (uri));
        uri->protocollen = get_protocol_length (struri (uri));
 
@@ -1456,7 +1420,7 @@ parse_uri (struct uri *uri, gchar *uristring, rspamd_mempool_t * pool)
        }
 
        url_strip (struri (uri));
-       url_unescape (uri->host);
+       rspamd_url_unescape (uri->host);
 
        path_simplify (uri->data);
 

diff --git a/src/libserver/url.h b/src/libserver/url.h
index d8877f2798011deaaf20145e22c0b8becf63c17c..5b508e056d2acf453f78bc53139986cd4b16c82f 100644 (file)
--- a/src/libserver/url.h
+++ b/src/libserver/url.h
@@ -119,4 +119,9 @@ gboolean url_try_text (rspamd_mempool_t *pool,
  */
 const gchar * url_strerror (enum uri_errno err);
 
+/*
+ * URL unescape characters in the specified string
+ */
+void rspamd_url_unescape (gchar *s);
+
 #endif