Attachment 211791 Details for Bug 313121 – Backport from php 4.4.7. It is adds php_url_parse_ex function with it's dependency php_replace_controlchars_ex. No other functions are modified.

[patch] Backport from php 4.4.7. It is adds php_url_parse_ex function with it's dependency php_replace_controlchars_ex. No other functions are modified.

php-4.3.9-php_url_parse_ex.patch (text/plain), 9.76 KB, created by Arturas Moskvinas on 2007-09-30 15:41:08 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Arturas Moskvinas

Created: 2007-09-30 15:41:08 UTC

Size: 9.76 KB

patch

obsolete

>--- php-4.3.9/ext/standard/url.c.php_url_parse_ex	2004-05-28 14:33:42.000000000 +0300
>+++ php-4.3.9/ext/standard/url.c	2007-09-30 00:10:36.000000000 +0300
>@@ -79,7 +79,29 @@
> 	return (str);
> } 
> /* }}} */
>- 
>+
>+/*backport from 4.4.7
>+*/ 
>+PHPAPI char *php_replace_controlchars_ex(char *str, int len)
>+{
>+        unsigned char *s = (unsigned char *)str;
>+        unsigned char *e = (unsigned char *)str + len;
>+
>+        if (!str) {
>+                return (NULL);
>+        }
>+
>+        while (s < e) {
>+            
>+                if (iscntrl(*s)) {
>+                        *s='_';
>+                }
>+                s++;
>+        }
>+
>+        return (str);
>+} 
>+/* }}} */
> 
> /* {{{ php_url_parse
>  */
>@@ -291,6 +313,246 @@
> }
> /* }}} */
> 
>+/* backport from php 4.4.7
>+*/
>+PHPAPI php_url *php_url_parse_ex(char const *str, int length)
>+{
>+        char port_buf[6];
>+        php_url *ret = ecalloc(1, sizeof(php_url));
>+        const char *s, *e, *p, *pp, *ue;
>+
>+        s = str;
>+        ue = s + length;
>+
>+        /* parse scheme */
>+        if ((e = memchr(s, ':', length)) && (e - s)) {
>+                /* validate scheme */
>+                p = s;
>+                while (p < e) {
>+                        /* scheme = 1*[ lowalpha | digit | "+" | "-" | "." ] */
>+                        if (!isalpha(*p) && !isdigit(*p) && *p != '+' && *p != '.' && *p != '-') {
>+                                if (e + 1 < ue) {
>+                                        goto parse_port;
>+                                } else {
>+                                        goto just_path;
>+                                }
>+                        }
>+                        p++;
>+                }
>+
>+                if (*(e + 1) == '\0') { /* only scheme is available */
>+                        ret->scheme = estrndup(s, (e - s));
>+                        php_replace_controlchars_ex(ret->scheme, (e - s));
>+                        goto end;
>+                }
>+
>+                /* 
>+                 * certain schemas like mailto: and zlib: may not have any / after them
>+                 * this check ensures we support those.
>+                 */
>+                if (*(e+1) != '/') {
>+                        /* check if the data we get is a port this allows us to 
>+                         * correctly parse things like a.com:80
>+                         */
>+                        p = e + 1;
>+                        while (isdigit(*p)) {
>+                                p++;
>+                        }
>+
>+                        if ((*p == '\0' || *p == '/') && (p - e) < 7) {
>+                                goto parse_port;
>+                        }
>+
>+                        ret->scheme = estrndup(s, (e-s));
>+                        php_replace_controlchars_ex(ret->scheme, (e - s));
>+
>+                        length -= ++e - s;
>+                        s = e;
>+                        goto just_path;
>+                } else {
>+                        ret->scheme = estrndup(s, (e-s));
>+                        php_replace_controlchars_ex(ret->scheme, (e - s));
>+
>+                        if (*(e+2) == '/') {
>+                                s = e + 3;
>+                                if (!strncasecmp("file", ret->scheme, sizeof("file"))) {
>+                                        if (*(e + 3) == '/') {
>+                                                goto nohost;
>+                                        }
>+                                }
>+                        } else {
>+                                if (!strncasecmp("file", ret->scheme, sizeof("file"))) {
>+                                        s = e + 1;
>+                                        goto nohost;
>+                                } else {
>+                                        length -= ++e - s;
>+                                        s = e;
>+                                        goto just_path;
>+                                }
>+                        }
>+                }
>+        } else if (e) { /* no scheme, look for port */
>+                parse_port:
>+                p = e + 1;
>+                pp = p;
>+
>+                while (pp-p < 6 && isdigit(*pp)) {
>+                        pp++;
>+                }
>+
>+                if (pp-p < 6 && (*pp == '/' || *pp == '\0')) {
>+                        memcpy(port_buf, p, (pp-p));
>+                        port_buf[pp-p] = '\0';
>+                        ret->port = atoi(port_buf);
>+                } else {
>+                        goto just_path;
>+                }
>+        } else {
>+                just_path:
>+                ue = s + length;
>+                goto nohost;
>+        }
>+
>+        e = ue;
>+
>+        if (!(p = memchr(s, '/', (ue - s)))) {
>+                if ((p = memchr(s, '?', (ue - s)))) {
>+                        e = p;
>+                } else if ((p = memchr(s, '#', (ue - s)))) {
>+                        e = p;
>+                }
>+        } else {
>+                e = p;
>+        }
>+
>+        {
>+                const char *t = s;
>+                p = NULL;
>+                while (e > t && (t = memchr(t, '@', (e-t)))) {
>+                        p = t++;
>+                }
>+        }
>+
>+        /* check for login and password */
>+        if (p) {
>+                if ((pp = memchr(s, ':', (p-s)))) {
>+                        if ((pp-s) > 0) {
>+                                ret->user = estrndup(s, (pp-s));
>+                                php_replace_controlchars_ex(ret->user, (pp - s));
>+                        }
>+
>+                        pp++;
>+                        if (p-pp > 0) {
>+                                ret->pass = estrndup(pp, (p-pp));
>+                                php_replace_controlchars_ex(ret->pass, (p-pp));
>+                        }
>+                } else {
>+                        ret->user = estrndup(s, (p-s));
>+                        php_replace_controlchars_ex(ret->user, (p-s));
>+                }
>+
>+                s = p + 1;
>+        }
>+
>+        /* check for port */
>+        if (*s == '[' && *(e-1) == ']') {
>+                /* Short circuit portscan
>+                   we're dealing with an
>+                   IPv6 embedded address */
>+                p = s;
>+        } else {
>+                /* memchr is a GNU specific extension
>+                   Emulate for wide compatability */
>+                for(p = e; *p != ':' && p >= s; p--);
>+        }
>+
>+        if (p >= s && *p == ':') {
>+                if (!ret->port) {
>+                        p++;
>+                        if (e-p > 5) { /* port cannot be longer then 5 characters */
>+                                STR_FREE(ret->scheme);
>+                                STR_FREE(ret->user);
>+                                STR_FREE(ret->pass);
>+                                efree(ret);
>+                                return NULL;
>+                        } else if (e - p > 0) {
>+                                memcpy(port_buf, p, (e-p));
>+                                port_buf[e-p] = '\0';
>+                                ret->port = atoi(port_buf);
>+                        }
>+                        p--;
>+                }
>+        } else {
>+                p = e;
>+        }
>+
>+        /* check if we have a valid host, if we don't reject the string as url */
>+        if ((p-s) < 1) {
>+                STR_FREE(ret->scheme);
>+                STR_FREE(ret->user);
>+                STR_FREE(ret->pass);
>+                efree(ret);
>+                return NULL;
>+        }
>+
>+        ret->host = estrndup(s, (p-s));
>+        php_replace_controlchars_ex(ret->host, (p - s));
>+
>+        if (e == ue) {
>+                return ret;
>+        }
>+
>+        s = e;
>+
>+        nohost:
>+
>+        if ((p = memchr(s, '?', (ue - s)))) {
>+                pp = strchr(s, '#');
>+
>+                if (pp && pp < p) {
>+                        p = pp;
>+                        pp = strchr(pp+2, '#');
>+                }
>+
>+                if (p - s) {
>+                        ret->path = estrndup(s, (p-s));
>+                        php_replace_controlchars_ex(ret->path, (p - s));
>+                }
>+
>+                if (pp) {
>+                        if (pp - ++p) { 
>+                                ret->query = estrndup(p, (pp-p));
>+                                php_replace_controlchars_ex(ret->query, (pp - p));
>+                        }
>+                        p = pp;
>+                        goto label_parse;
>+                } else if (++p - ue) {
>+                        ret->query = estrndup(p, (ue-p));
>+                        php_replace_controlchars_ex(ret->query, (ue - p));
>+                }
>+        } else if ((p = memchr(s, '#', (ue - s)))) {
>+                if (p - s) {
>+                        ret->path = estrndup(s, (p-s));
>+                        php_replace_controlchars_ex(ret->path, (p - s));
>+                }
>+
>+                label_parse:
>+                p++;
>+
>+                if (ue - p) {
>+                        ret->fragment = estrndup(p, (ue-p));
>+                        php_replace_controlchars_ex(ret->fragment, (ue - p));
>+                }
>+        } else {
>+                ret->path = estrndup(s, (ue-s));
>+                php_replace_controlchars_ex(ret->path, (ue - s));
>+        }
>+end:
>+        return ret;
>+}
>+/* }}} */
>+
>+
> /* {{{ proto array parse_url(string url)
>    Parse a URL and return its components */
> PHP_FUNCTION(parse_url)
>--- php-4.3.9/ext/standard/url.h.php_url_parse_ex	2002-12-31 18:35:35.000000000 +0200
>+++ php-4.3.9/ext/standard/url.h	2007-09-30 00:01:01.000000000 +0300
>@@ -33,6 +33,7 @@
> 
> PHPAPI void php_url_free(php_url *theurl);
> PHPAPI php_url *php_url_parse(char *str);
>+PHPAPI php_url *php_url_parse_ex(char const *str, int length);
> PHPAPI int php_url_decode(char *str, int len); /* return value: length of decoded string */
> PHPAPI int php_raw_url_decode(char *str, int len); /* return value: length of decoded string */
> PHPAPI char *php_url_encode(char *s, int len, int *new_length);

Actions: View | Diff

Attachments on bug 313121: 211791