Some Perl modules have moved to UCW namespace.

[libucw.git] / lib / url.c

diff --git a/lib/url.c b/lib/url.c
index e38238ec3973718b0badf8f56580eac6baac5ebd..8f1a435bb241fa2c4592f5f6d93e545baa10d2ef 100644 (file)
--- a/lib/url.c
+++ b/lib/url.c
@@ -2,7 +2,7 @@
  *     UCW Library -- URL Functions
  *
  *     (c) 1997--2004 Martin Mares <mj@ucw.cz>
- *     (c) 2001 Robert Spalek <robert@ucw.cz>
+ *     (c) 2001--2005 Robert Spalek <robert@ucw.cz>
  *
  *     This software may be freely distributed and used according to the terms
  *     of the GNU Lesser General Public License.
@@ -29,35 +29,36 @@
 
 static uns url_ignore_spaces;
 static uns url_ignore_underflow;
-static byte *url_component_separators = "";
+static char *url_component_separators = "";
 static uns url_min_repeat_count = 0x7fffffff;
 static uns url_max_repeat_length = 0;
 
-static struct cfitem url_config[] = {
-  { "URL",                             CT_SECTION,     NULL },
-  { "IgnoreSpaces",                    CT_INT,         &url_ignore_spaces },
-  { "IgnoreUnderflow",                 CT_INT,         &url_ignore_underflow },
-  { "ComponentSeparators",             CT_STRING,      &url_component_separators },
-  { "MinRepeatCount",                  CT_INT,         &url_min_repeat_count },
-  { "MaxRepeatLength",                 CT_INT,         &url_max_repeat_length },
-  { NULL,                              CT_STOP,        NULL }
+static struct cf_section url_config = {
+  CF_ITEMS {
+    CF_UNS("IgnoreSpaces", &url_ignore_spaces),
+    CF_UNS("IgnoreUnderflow", &url_ignore_underflow),
+    CF_STRING("ComponentSeparators", &url_component_separators),
+    CF_UNS("MinRepeatCount", &url_min_repeat_count),
+    CF_UNS("MaxRepeatLength", &url_max_repeat_length),
+    CF_END
+  }
 };
 
 static void CONSTRUCTOR url_init_config(void)
 {
-  cf_register(url_config);
+  cf_declare_section("URL", &url_config, 0);
 }
 
 /* Escaping and de-escaping */
 
-uns
+static uns
 enhex(uns x)
 {
   return (x<10) ? (x + '0') : (x - 10 + 'A');
 }
 
 int
-url_deescape(byte *s, byte *d)
+url_deescape(const byte *s, byte *d)
 {
   byte *dstart = d;
   byte *end = d + MAX_URL_SIZE - 10;
@@ -99,7 +100,7 @@ url_deescape(byte *s, byte *d)
        *d++ = *s++;
       else if (Cspace(*s))
        {
-         byte *s0 = s;
+         const byte *s0 = s;
          while (Cspace(*s))
            s++;
          if (!url_ignore_spaces || !(!*s || d == dstart))
@@ -120,7 +121,7 @@ url_deescape(byte *s, byte *d)
 }
 
 int
-url_enescape(byte *s, byte *d)
+url_enescape(const byte *s, byte *d)
 {
   byte *end = d + MAX_URL_SIZE - 10;
   unsigned int c;
@@ -149,13 +150,36 @@ url_enescape(byte *s, byte *d)
   return 0;
 }
 
+int
+url_enescape_friendly(const byte *src, byte *dest)
+{
+  byte *end = dest + MAX_URL_SIZE - 10;
+  while (*src)
+    {
+      if (dest >= end)
+       return URL_ERR_TOO_LONG;
+      if (*src < NCC_MAX)
+       *dest++ = NCC_CHARS[*src++];
+      else if (*src >= 0x20 && *src < 0x7f)
+       *dest++ = *src++;
+      else
+       {
+         *dest++ = '%';
+         *dest++ = enhex(*src >> 4);
+         *dest++ = enhex(*src++ & 0x0f);
+       }
+    }
+  *dest = 0;
+  return 0;
+}
+
 /* Split an URL (several parts may be copied to the destination buffer) */
 
 byte *url_proto_names[URL_PROTO_MAX] = URL_PNAMES;
 static int url_proto_path_flags[URL_PROTO_MAX] = URL_PATH_FLAGS;
 
 uns
-identify_protocol(byte *p)
+identify_protocol(const byte *p)
 {
   uns i;
 
@@ -201,18 +225,28 @@ url_split(byte *s, struct url *u, byte *d)
     {
       if (s[1] == '/')                 /* Host spec */
        {
-         byte *q, *w, *e;
+         byte *q, *e;
+         byte *at = NULL;
          char *ep;
 
          s += 2;
          q = d;
          while (*s && *s != '/' && *s != '?')  /* Copy user:passwd@host:port */
-           *d++ = *s++;
+           {
+             if (*s != '@')
+               *d++ = *s;
+             else if (!at)
+               {
+                 *d++ = 0;
+                 at = d;
+               }
+             else                      /* This shouldn't happen with sane URL's, but we need to be sure */
+               *d++ = NCC_AT;
+             s++;
+           }
          *d++ = 0;
-         w = strchr(q, '@');
-         if (w)                        /* user:passwd present */
+         if (at)                       /* user:passwd present */
            {
-             *w++ = 0;
              u->user = q;
              if (e = strchr(q, ':'))
                {
@@ -221,8 +255,8 @@ url_split(byte *s, struct url *u, byte *d)
                }
            }
          else
-           w = q;
-         e = strchr(w, ':');
+           at = q;
+         e = strchr(at, ':');
          if (e)                        /* host:port present */
            {
              uns p;
@@ -233,7 +267,7 @@ url_split(byte *s, struct url *u, byte *d)
              else if (p)               /* Port 0 (e.g. in :/) is treated as default port */
                u->port = p;
            }
-         u->host = w;
+         u->host = at;
        }
     }
 
@@ -458,7 +492,7 @@ url_canonicalize(struct url *u)
 /* Pack a broken-down URL */
 
 static byte *
-append(byte *d, byte *s, byte *e)
+append(byte *d, const byte *s, byte *e)
 {
   if (d)
     while (*s)
@@ -537,7 +571,7 @@ url_error(uns err)
 /* Standard cookbook recipes */
 
 int
-url_canon_split_rel(byte *u, byte *buf1, byte *buf2, struct url *url, struct url *base)
+url_canon_split_rel(const byte *u, byte *buf1, byte *buf2, struct url *url, struct url *base)
 {
   int err;
 
@@ -551,7 +585,7 @@ url_canon_split_rel(byte *u, byte *buf1, byte *buf2, struct url *url, struct url
 }
 
 int
-url_auto_canonicalize_rel(byte *src, byte *dst, struct url *base)
+url_auto_canonicalize_rel(const byte *src, byte *dst, struct url *base)
 {
   byte buf1[MAX_URL_SIZE], buf2[MAX_URL_SIZE], buf3[MAX_URL_SIZE];
   int err;
@@ -631,13 +665,13 @@ int main(int argc, char **argv)
 #endif
 
 struct component {
-       byte *start;
+       const byte *start;
        int length;
        u32 hash;
 };
 
 static inline u32
-hashf(byte *start, int length)
+hashf(const byte *start, int length)
 {
        u32 hf = length;
        while (length-- > 0)
@@ -667,11 +701,11 @@ repeat_count(struct component *comp, uns count, uns len)
 }
 
 int
-url_has_repeated_component(byte *url)
+url_has_repeated_component(const byte *url)
 {
        struct component *comp;
        uns comps, comp_len, rep_prefix;
-       byte *c;
+       const byte *c;
        uns i;
 
        for (comps=0, c=url; c; comps++)