aboutsummaryrefslogtreecommitdiffstats
path: root/util-src/encodings.c
diff options
context:
space:
mode:
authorMatthew Wild <mwild1@gmail.com>2015-03-27 22:11:58 +0000
committerMatthew Wild <mwild1@gmail.com>2015-03-27 22:11:58 +0000
commit29f906c30d10ce983b6c034d88bf7a3c8b576bb0 (patch)
tree05c689e53d3416f52f0e9c782cf05887e5fe14da /util-src/encodings.c
parent13edf3e093fd656d5110b277eca4c9805f3d925c (diff)
parent51bef3673c57c5a2e323a7eb1424c85498905f53 (diff)
downloadprosody-29f906c30d10ce983b6c034d88bf7a3c8b576bb0.tar.gz
prosody-29f906c30d10ce983b6c034d88bf7a3c8b576bb0.zip
Merge 0.9->0.10 (third time lucky)
Diffstat (limited to 'util-src/encodings.c')
-rw-r--r--util-src/encodings.c98
1 files changed, 95 insertions, 3 deletions
diff --git a/util-src/encodings.c b/util-src/encodings.c
index 2d5d49d4..529a6c22 100644
--- a/util-src/encodings.c
+++ b/util-src/encodings.c
@@ -1,6 +1,7 @@
/* Prosody IM
-- Copyright (C) 2008-2010 Matthew Wild
-- Copyright (C) 2008-2010 Waqas Hussain
+-- Copyright (C) 1994-2015 Lua.org, PUC-Rio.
--
-- This project is MIT/X11 licensed. Please see the
-- COPYING file in the source package for more information.
@@ -120,6 +121,88 @@ static const luaL_Reg Reg_base64[] =
{ NULL, NULL }
};
+/******************* UTF-8 ********************/
+
+/*
+ * Adapted from Lua 5.3
+ * Needed because libidn does not validate that input is valid UTF-8
+ */
+
+#define MAXUNICODE 0x10FFFF
+
+/*
+ * Decode one UTF-8 sequence, returning NULL if byte sequence is invalid.
+ */
+static const char *utf8_decode (const char *o, int *val) {
+ static unsigned int limits[] = {0xFF, 0x7F, 0x7FF, 0xFFFF};
+ const unsigned char *s = (const unsigned char *)o;
+ unsigned int c = s[0];
+ unsigned int res = 0; /* final result */
+ if (c < 0x80) /* ascii? */
+ res = c;
+ else {
+ int count = 0; /* to count number of continuation bytes */
+ while (c & 0x40) { /* still have continuation bytes? */
+ int cc = s[++count]; /* read next byte */
+ if ((cc & 0xC0) != 0x80) /* not a continuation byte? */
+ return NULL; /* invalid byte sequence */
+ res = (res << 6) | (cc & 0x3F); /* add lower 6 bits from cont. byte */
+ c <<= 1; /* to test next bit */
+ }
+ res |= ((c & 0x7F) << (count * 5)); /* add first byte */
+ if (count > 3 || res > MAXUNICODE || res <= limits[count] || (0xd800 <= res && res <= 0xdfff) )
+ return NULL; /* invalid byte sequence */
+ s += count; /* skip continuation bytes read */
+ }
+ if (val) *val = res;
+ return (const char *)s + 1; /* +1 to include first byte */
+}
+
+/*
+ * Check that a string is valid UTF-8
+ * Returns NULL if not
+ */
+const char* check_utf8 (lua_State *L, int idx, size_t *l) {
+ size_t pos, len;
+ const char *s = luaL_checklstring(L, 1, &len);
+ pos = 0;
+ while (pos <= len) {
+ const char *s1 = utf8_decode(s + pos, NULL);
+ if (s1 == NULL) { /* conversion error? */
+ return NULL;
+ }
+ pos = s1 - s;
+ }
+ if(l != NULL) {
+ *l = len;
+ }
+ return s;
+}
+
+static int Lutf8_valid(lua_State *L) {
+ lua_pushboolean(L, check_utf8(L, 1, NULL) != NULL);
+ return 1;
+}
+
+static int Lutf8_length(lua_State *L) {
+ size_t len;
+ if(!check_utf8(L, 1, &len)) {
+ lua_pushnil(L);
+ lua_pushliteral(L, "invalid utf8");
+ return 2;
+ }
+ lua_pushinteger(L, len);
+ return 1;
+}
+
+static const luaL_Reg Reg_utf8[] =
+{
+ { "valid", Lutf8_valid },
+ { "length", Lutf8_length },
+ { NULL, NULL }
+};
+
+
/***************** STRINGPREP *****************/
#ifdef USE_STRINGPREP_ICU
@@ -216,8 +299,8 @@ static int stringprep_prep(lua_State *L, const Stringprep_profile *profile)
lua_pushnil(L);
return 1;
}
- s = lua_tolstring(L, 1, &len);
- if (len >= 1024) {
+ s = check_utf8(L, 1, &len);
+ if (s == NULL || len >= 1024 || len != strlen(s)) {
lua_pushnil(L);
return 1; /* TODO return error message */
}
@@ -324,7 +407,11 @@ static int Lidna_to_unicode(lua_State *L) /** idna.to_unicode(s) */
static int Lidna_to_ascii(lua_State *L) /** idna.to_ascii(s) */
{
size_t len;
- const char *s = luaL_checklstring(L, 1, &len);
+ const char *s = check_utf8(L, 1, &len);
+ if (s == NULL || len != strlen(s)) {
+ lua_pushnil(L);
+ return 1; /* TODO return error message */
+ }
char* output = NULL;
int ret = idna_to_ascii_8z(s, &output, IDNA_USE_STD3_ASCII_RULES);
if (ret == IDNA_SUCCESS) {
@@ -384,6 +471,11 @@ LUALIB_API int luaopen_util_encodings(lua_State *L)
luaL_register(L, NULL, Reg_idna);
lua_setfield(L, -2, "idna");
+ lua_newtable(L);
+ luaL_register(L, NULL, Reg_utf8);
+ lua_setfield(L, -2, "utf8");
+
+ lua_pushliteral(L, "version"); /** version */
lua_pushliteral(L, "-3.14");
lua_setfield(L, -2, "version");
return 1;