X-Git-Url: http://sjero.net/git/?a=blobdiff_plain;f=src%2Furl.c;h=f97a31801ea5a6861fccc47b4a1d4d60632bf6b0;hb=aa07e689f2c03dd25342859e7e527a13467ad219;hp=0a23df32d38ba2806cb8ae0f20a0587f802d6016;hpb=cfec85440d40f4ea3a71040ae35c050c71327529;p=wget diff --git a/src/url.c b/src/url.c index 0a23df32..f97a3180 100644 --- a/src/url.c +++ b/src/url.c @@ -508,7 +508,8 @@ parse_credentials (const char *beg, const char *end, char **user, char **passwd) } /* Used by main.c: detect URLs written using the "shorthand" URL forms - popularized by Netscape and NcFTP. HTTP shorthands look like this: + originally popularized by Netscape and NcFTP. HTTP shorthands look + like this: www.foo.com[:port]/dir/file -> http://www.foo.com[:port]/dir/file www.foo.com[:port] -> http://www.foo.com[:port] @@ -524,52 +525,42 @@ char * rewrite_shorthand_url (const char *url) { const char *p; + char *ret; if (url_scheme (url) != SCHEME_INVALID) return NULL; /* Look for a ':' or '/'. The former signifies NcFTP syntax, the latter Netscape. */ - for (p = url; *p && *p != ':' && *p != '/'; p++) - ; - + p = strpbrk (url, ":/"); if (p == url) return NULL; /* If we're looking at "://", it means the URL uses a scheme we don't support, which may include "https" when compiled without SSL support. Don't bogusly rewrite such URLs. */ - if (p[0] == ':' && p[1] == '/' && p[2] == '/') + if (p && p[0] == ':' && p[1] == '/' && p[2] == '/') return NULL; - if (*p == ':') + if (p && *p == ':') { - const char *pp; - char *res; - /* If the characters after the colon and before the next slash - or end of string are all digits, it's HTTP. */ - int digits = 0; - for (pp = p + 1; ISDIGIT (*pp); pp++) - ++digits; - if (digits > 0 && (*pp == '/' || *pp == '\0')) + /* Colon indicates ftp, as in foo.bar.com:path. Check for + special case of http port number ("localhost:10000"). */ + int digits = strspn (p + 1, "0123456789"); + if (digits && (p[1 + digits] == '/' || p[1 + digits] == '\0')) goto http; - /* Prepend "ftp://" to the entire URL... */ - res = xmalloc (6 + strlen (url) + 1); - sprintf (res, "ftp://%s", url); - /* ...and replace ':' with '/'. */ - res[6 + (p - url)] = '/'; - return res; + /* Turn "foo.bar.com:path" to "ftp://foo.bar.com/path". */ + ret = aprintf ("ftp://%s", url); + ret[6 + (p - url)] = '/'; } else { - char *res; http: - /* Just prepend "http://" to what we have. */ - res = xmalloc (7 + strlen (url) + 1); - sprintf (res, "http://%s", url); - return res; + /* Just prepend "http://" to URL. */ + ret = aprintf ("http://%s", url); } + return ret; } static void split_path (const char *, char **, char **); @@ -1520,8 +1511,7 @@ path_simplify (char *path) { char *h = path; /* hare */ char *t = path; /* tortoise */ - char *beg = path; /* boundary for backing the tortoise */ - char *end = path + strlen (path); + char *end = strchr (path, '\0'); while (h < end) { @@ -1536,26 +1526,17 @@ path_simplify (char *path) { /* Handle "../" by retreating the tortoise by one path element -- but not past beggining. */ - if (t > beg) + if (t > path) { /* Move backwards until T hits the beginning of the previous path element or the beginning of path. */ - for (--t; t > beg && t[-1] != '/'; t--) + for (--t; t > path && t[-1] != '/'; t--) ; } - else - { - /* If we're at the beginning, copy the "../" literally - move the beginning so a later ".." doesn't remove - it. */ - beg = t + 3; - goto regular; - } h += 3; } else { - regular: /* A regular path element. If H hasn't advanced past T, simply skip to the next path element. Otherwise, copy the path element until the next slash. */ @@ -1981,8 +1962,8 @@ test_path_simplify (void) { "", "", false }, { ".", "", true }, { "./", "", true }, - { "..", "..", false }, - { "../", "../", false }, + { "..", "", true }, + { "../", "", true }, { "foo", "foo", false }, { "foo/bar", "foo/bar", false }, { "foo///bar", "foo///bar", false }, @@ -1995,9 +1976,9 @@ test_path_simplify (void) { "foo/bar/../x", "foo/x", true }, { "foo/bar/../x/", "foo/x/", true }, { "foo/..", "", true }, - { "foo/../..", "..", true }, - { "foo/../../..", "../..", true }, - { "foo/../../bar/../../baz", "../../baz", true }, + { "foo/../..", "", true }, + { "foo/../../..", "", true }, + { "foo/../../bar/../../baz", "baz", true }, { "a/b/../../c", "c", true }, { "./a/../b", "b", true } };