From 4af624627eafdd5db9e0200bfd05c59aa60292b9 Mon Sep 17 00:00:00 2001 From: RichardHitt Date: Wed, 2 Jan 2013 14:50:26 -0800 Subject: overhauling of URL detection, including channel, nick, etc 'words' --- src/common/tree.c | 2 +- src/common/tree.h | 2 +- src/common/url.c | 459 +++++++++++++++++++++++++++++--------------------- src/common/url.h | 3 +- src/common/userlist.c | 2 +- src/common/userlist.h | 2 +- 6 files changed, 276 insertions(+), 194 deletions(-) (limited to 'src/common') diff --git a/src/common/tree.c b/src/common/tree.c index 33fe1d41..715b0e56 100644 --- a/src/common/tree.c +++ b/src/common/tree.c @@ -142,7 +142,7 @@ mybsearch (const void *key, void **array, size_t nmemb, } void * -tree_find (tree *t, void *key, tree_cmp_func *cmp, void *data, int *pos) +tree_find (tree *t, const void *key, tree_cmp_func *cmp, void *data, int *pos) { if (!t || !t->array) return NULL; diff --git a/src/common/tree.h b/src/common/tree.h index 4a158052..ced8e425 100644 --- a/src/common/tree.h +++ b/src/common/tree.h @@ -8,7 +8,7 @@ typedef int (tree_traverse_func) (const void *key, void *data); tree *tree_new (tree_cmp_func *cmp, void *data); void tree_destroy (tree *t); -void *tree_find (tree *t, void *key, tree_cmp_func *cmp, void *data, int *pos); +void *tree_find (tree *t, const void *key, tree_cmp_func *cmp, void *data, int *pos); int tree_remove (tree *t, void *key, int *pos); void *tree_remove_at_pos (tree *t, int pos); void tree_foreach (tree *t, tree_traverse_func *func, void *data); diff --git a/src/common/url.c b/src/common/url.c index c5335859..98293635 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -13,7 +13,7 @@ * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA */ #include @@ -32,6 +32,13 @@ void *url_tree = NULL; GTree *url_btree = NULL; +static int do_an_re (const char *word, int *start, int *end, int *type); +static GRegex *re_url (void); +static GRegex *re_host (void); +static GRegex *re_email (void); +static GRegex *re_nick (void); +static GRegex *re_channel (void); +static GRegex *re_path (void); static int @@ -177,158 +184,38 @@ url_add (char *urltext, int len) keep it FAST! This new version was found to be almost 3x faster than 2.4.4 release. */ +static int laststart = 0; +static int lastend = 0; +static int lasttype = 0; + int -url_check_word (const char *word, int len) +url_check_word (const char *word) { -#define D(x) (x), ((sizeof (x)) - 1) - static const struct { - const char *s; - int len; - } - prefix[] = { - { D("irc.") }, - { D("ftp.") }, - { D("www.") }, - { D("irc://") }, - { D("ftp://") }, - { D("http://") }, - { D("https://") }, - { D("file://") }, - { D("rtsp://") }, - { D("ut2004://") }, - }, - suffix[] = { - { D(".org") }, - { D(".net") }, - { D(".com") }, - { D(".edu") }, - { D(".html") }, - { D(".info") }, - { D(".name") }, - /* Some extra common suffixes. - foo.blah/baz.php etc should work now, rather than - needing http:// at the beginning. */ - { D(".php") }, - { D(".htm") }, - { D(".aero") }, - { D(".asia") }, - { D(".biz") }, - { D(".cat") }, - { D(".coop") }, - { D(".int") }, - { D(".jobs") }, - { D(".mobi") }, - { D(".museum") }, - { D(".pro") }, - { D(".tel") }, - { D(".travel") }, - { D(".xxx") }, - { D(".asp") }, - { D(".aspx") }, - { D(".shtml") }, - { D(".xml") }, - }; -#undef D - const char *at, *dot; - int i, dots; - - /* this is pretty much the same as in logmask_is_fullpath() except with length checks and .\ for portable mode */ -#ifdef WIN32 - if ((len > 1 && word[0] == '\\') || - (len > 2 && word[0] == '.' && word[1] == '\\') || - (len > 2 && (((word[0] >= 'A' && word[0] <= 'Z') || (word[0] >= 'a' && word[0] <= 'z')) && word[1] == ':'))) -#else - if (len > 1 && word[0] == '/') -#endif - { - return WORD_PATH; - } - - if (len > 1 && word[1] == '#' && strchr("@+^%*#", word[0])) - return WORD_CHANNEL; - - if ((word[0] == '#' || word[0] == '&') && word[1] != '#' && word[1] != 0) - return WORD_CHANNEL; - - for (i = 0; i < G_N_ELEMENTS(prefix); i++) - { - int l; - - l = prefix[i].len; - if (len > l) - { - int j; - - /* This is pretty much g_ascii_strncasecmp(). */ - for (j = 0; j < l; j++) - { - unsigned char c = word[j]; - if (tolower(c) != prefix[i].s[j]) - break; - } - if (j == l) - return WORD_URL; - } - } - - at = strchr (word, '@'); /* check for email addy */ - dot = strrchr (word, '.'); - if (at && dot) - { - if (at < dot) - { - if (strchr (word, '*')) - return WORD_HOST; - else - return WORD_EMAIL; - } - } - - /* check if it's an IP number */ - dots = 0; - for (i = 0; i < len; i++) - { - if (word[i] == '.' && i > 0) - dots++; /* allow 127.0.0.1:80 */ - else if (!isdigit ((unsigned char) word[i]) && word[i] != ':') - { - dots = 0; - break; - } - } - if (dots == 3) - return WORD_HOST; - - if (len > 5) + laststart = lastend = lasttype = 0; + if (do_an_re (word, &laststart, &lastend, &lasttype)) { - for (i = 0; i < G_N_ELEMENTS(suffix); i++) + switch (lasttype) { - int l; - - l = suffix[i].len; - if (len > l) - { - const unsigned char *p = &word[len - l]; - int j; - - /* This is pretty much g_ascii_strncasecmp(). */ - for (j = 0; j < l; j++) - { - if (tolower(p[j]) != suffix[i].s[j]) - break; - } - if (j == l) - return WORD_HOST; - } + case WORD_NICK: + if (!isalnum (word[laststart])) + laststart++; + if (!userlist_find (current_sess, &word[laststart])) + lasttype = 0; + return lasttype; + case WORD_EMAIL: + if (!isalnum (word[laststart])) + laststart++; + /* Fall through */ + case WORD_URL: + case WORD_HOST: + case WORD_CHANNEL: + return lasttype; + default: + return 0; /* Should not occur */ } - - if (word[len - 3] == '.' && - isalpha ((unsigned char) word[len - 2]) && - isalpha ((unsigned char) word[len - 1])) - return WORD_HOST; } - - return 0; + else + return 0; } /* List of IRC commands for which contents (and thus possible URLs) @@ -346,9 +233,10 @@ static char *commands[] = { void url_check_line (char *buf, int len) { + GRegex *re(void); + GMatchInfo *gmi; char *po = buf; - char *start; - int i, wlen; + int i; /* Skip over message prefix */ if (*po == ':') @@ -379,50 +267,243 @@ url_check_line (char *buf, int len) return; po++; - if (buf[0] == ':' && buf[1] != 0) - po++; + g_regex_match(re_url(), po, 0, &gmi); + while (g_match_info_matches(gmi)) + { + int start, end; + + g_match_info_fetch_pos(gmi, 0, &start, &end); + if (po[end - 1] == '\r') + po[--end] = 0; + if (g_strstr_len (po + start, end - start, "://")) + url_add(po + start, end - start); + g_match_info_next(gmi, NULL); + } + g_match_info_free(gmi); +} + +int +url_last (int *lstart, int *lend) +{ + *lstart = laststart; + *lend = lastend; + return lasttype; +} + +static int +do_an_re(const char *word,int *start, int *end, int *type) +{ + typedef struct func_s { + GRegex *(*fn)(void); + int type; + } func_t; + func_t funcs[] = + { + { re_email, WORD_EMAIL }, + { re_url, WORD_URL }, + { re_host, WORD_HOST }, + { re_channel, WORD_CHANNEL }, + { re_path, WORD_PATH }, + { re_nick, WORD_NICK } + }; - start = po; + GMatchInfo *gmi; + int k; - /* check each "word" (space separated) */ - while (1) + for (k = 0; k < sizeof funcs / sizeof (func_t); k++) { - switch (po[0]) + g_regex_match (funcs[k].fn(), word, 0, &gmi); + if (!g_match_info_matches (gmi)) { - case 0: - case ' ': - case '\r': - - wlen = po - start; - if (wlen > 2) - { - /* HACK! :( */ - /* This is to work around not being able to detect URLs that are at - the start of messages. */ - if (start[0] == ':') - { - start++; - wlen--; - } - if (start[0] == '+' || start[0] == '-') - { - start++; - wlen--; - } - - if (wlen > 2 && url_check_word (start, wlen) == WORD_URL) - { - url_add (start, wlen); - } - } - if (po[0] == 0) - return; - po++; - start = po; - break; - - default: - po++; + g_match_info_free (gmi); + continue; + } + while (g_match_info_matches (gmi)) + { + g_match_info_fetch_pos (gmi, 0, start, end); + g_match_info_next (gmi, NULL); } + g_match_info_free (gmi); + *type = funcs[k].type; + return TRUE; } + + return FALSE; +} + +/* Miscellaneous description --- */ +#define DOMAIN "[-a-z0-9]+(\\.[-a-z0-9]+)*\\.[a-z]+" +#define IPADDR "[0-9]+(\\.[0-9]+){3}" +#define HOST "(" DOMAIN "|" IPADDR ")" +#define OPT_PORT "(:[1-9][0-9]{0,4})?" + +GRegex * +make_re(char *grist, char *type) +{ + GRegex *ret; + GError *err = NULL; + + ret = g_regex_new (grist, G_REGEX_CASELESS + G_REGEX_OPTIMIZE, 0, &err); + g_free (grist); + return ret; +} + +/* HOST description --- */ +/* (see miscellaneous above) */ +static GRegex * +re_host (void) +{ + static GRegex *host_ret; + char *grist; + grist = g_strdup_printf ( + "(" /* HOST */ + HOST OPT_PORT + ")" + ); + host_ret = make_re (grist, "re_host"); + return host_ret; +} + +/* URL description --- */ +#define SCHEME "(%s)" +#define LPAR "\\(" +#define RPAR "\\)" +#define NOPARENS "[^() \t]*" + +char *prefix[] = { + "irc\\.", + "ftp\\.", + "www\\.", + "irc://", + "ftp://", + "http://", + "https://", + "file://", + "rtsp://", + NULL +}; + +static GRegex * +re_url (void) +{ + static GRegex *url_ret; + char *grist; + char *scheme; + + if (url_ret) return url_ret; + + scheme = g_strjoinv ("|", prefix); + grist = g_strdup_printf ( + "(" /* URL or HOST */ + SCHEME HOST OPT_PORT + "(" /* Optional "/path?query_string#fragment_id" */ + "/" /* Must start with slash */ + "(" + "(" LPAR NOPARENS RPAR ")" + "|" + "(" NOPARENS ")" + ")*" /* Zero or more occurrences of either of these */ + "(? Date: Wed, 2 Jan 2013 14:58:26 -0800 Subject: Fix FSF address to be Franklin Street --- src/common/url.c | 2 +- src/fe-gtk/xtext.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index 98293635..04fa4c22 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -13,7 +13,7 @@ * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA */ #include diff --git a/src/fe-gtk/xtext.c b/src/fe-gtk/xtext.c index 4d2a8a69..e2dc2b7c 100644 --- a/src/fe-gtk/xtext.c +++ b/src/fe-gtk/xtext.c @@ -13,7 +13,7 @@ * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA * ========================================================================= * * xtext, the text widget used by X-Chat. -- cgit 1.4.1 From 270a31535c6e134f1ee497c45b3a62bd334940eb Mon Sep 17 00:00:00 2001 From: RichardHitt Date: Wed, 2 Jan 2013 15:03:16 -0800 Subject: I forgot to add case WORD_PATH to do_an_url()'s switch statement --- src/common/url.c | 1 + 1 file changed, 1 insertion(+) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index 04fa4c22..4d9d9f33 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -209,6 +209,7 @@ url_check_word (const char *word) case WORD_URL: case WORD_HOST: case WORD_CHANNEL: + case WORD_PATH: return lasttype; default: return 0; /* Should not occur */ -- cgit 1.4.1 From f4c2c08ccbd12e8fbd07666effada7f7470c1dc2 Mon Sep 17 00:00:00 2001 From: RichardHitt Date: Wed, 2 Jan 2013 23:47:58 -0800 Subject: Also allow URLs of of the form \.[a-z]+\/ --- src/common/url.c | 35 +++++++++++++++++++++++------------ 1 file changed, 23 insertions(+), 12 deletions(-) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index 4d9d9f33..e3834a70 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -396,19 +396,30 @@ re_url (void) scheme = g_strjoinv ("|", prefix); grist = g_strdup_printf ( "(" /* URL or HOST */ - SCHEME HOST OPT_PORT - "(" /* Optional "/path?query_string#fragment_id" */ - "/" /* Must start with slash */ - "(" - "(" LPAR NOPARENS RPAR ")" - "|" - "(" NOPARENS ")" - ")*" /* Zero or more occurrences of either of these */ - "(? Date: Thu, 3 Jan 2013 14:13:20 -0800 Subject: For NICK0 use NICKSPE, not NICKDIG --- src/common/url.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index e3834a70..c21870a0 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -453,7 +453,7 @@ re_email (void) #define NICKDIG "0-9" /* Note for NICKSPE: \\\\ boils down to a single \ */ #define NICKSPE "\\[\\]\\\\`_^{|}" -#define NICK0 "[" NICKPRE "]?[" NICKLET NICKDIG "]" +#define NICK0 "[" NICKPRE "]?[" NICKLET NICKSPE "]" #define NICK1 "[" NICKHYP NICKLET NICKDIG NICKSPE "]+" #define NICK NICK0 NICK1 -- cgit 1.4.1 From 4bf2fbe092fbe791b5ea0268b07ae3aaaa7a5870 Mon Sep 17 00:00:00 2001 From: RichardHitt Date: Thu, 3 Jan 2013 14:24:04 -0800 Subject: Ywes, nicks can be as short as one character long! --- src/common/url.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index c21870a0..b5af5d3a 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -454,7 +454,7 @@ re_email (void) /* Note for NICKSPE: \\\\ boils down to a single \ */ #define NICKSPE "\\[\\]\\\\`_^{|}" #define NICK0 "[" NICKPRE "]?[" NICKLET NICKSPE "]" -#define NICK1 "[" NICKHYP NICKLET NICKDIG NICKSPE "]+" +#define NICK1 "[" NICKHYP NICKLET NICKDIG NICKSPE "]*" #define NICK NICK0 NICK1 static GRegex * -- cgit 1.4.1 From 5ec9d8534e43bb6c64c2e7fc7a7f6ee6247bdc0c Mon Sep 17 00:00:00 2001 From: Richard Hitt Date: Tue, 8 Jan 2013 12:55:20 -0800 Subject: Add "ircs://" to the list of URL schemes --- src/common/url.c | 1 + 1 file changed, 1 insertion(+) (limited to 'src/common') diff --git a/src/common/url.c b/src/common/url.c index e3834a70..c93bc0f6 100644 --- a/src/common/url.c +++ b/src/common/url.c @@ -376,6 +376,7 @@ char *prefix[] = { "ftp\\.", "www\\.", "irc://", + "ircs://", "ftp://", "http://", "https://", -- cgit 1.4.1