"branch -d" can remove more than one branches
[git/gitweb.git] / url.c
blob3e06fd34c46a00b41f4843ea87a151314ce0789c
1 #include "cache.h"
2 #include "url.h"
4 int is_urlschemechar(int first_flag, int ch)
6 /*
7 * The set of valid URL schemes, as per STD66 (RFC3986) is
8 * '[A-Za-z][A-Za-z0-9+.-]*'. But use sightly looser check
9 * of '[A-Za-z0-9][A-Za-z0-9+.-]*' because earlier version
10 * of check used '[A-Za-z0-9]+' so not to break any remote
11 * helpers.
13 int alphanumeric, special;
14 alphanumeric = ch > 0 && isalnum(ch);
15 special = ch == '+' || ch == '-' || ch == '.';
16 return alphanumeric || (!first_flag && special);
19 int is_url(const char *url)
21 const char *url2, *first_slash;
23 if (!url)
24 return 0;
25 url2 = url;
26 first_slash = strchr(url, '/');
28 /* Input with no slash at all or slash first can't be URL. */
29 if (!first_slash || first_slash == url)
30 return 0;
31 /* Character before must be : and next must be /. */
32 if (first_slash[-1] != ':' || first_slash[1] != '/')
33 return 0;
34 /* There must be something before the :// */
35 if (first_slash == url + 1)
36 return 0;
38 * Check all characters up to first slash - 1. Only alphanum
39 * is allowed.
41 url2 = url;
42 while (url2 < first_slash - 1) {
43 if (!is_urlschemechar(url2 == url, (unsigned char)*url2))
44 return 0;
45 url2++;
48 /* Valid enough. */
49 return 1;
52 static int url_decode_char(const char *q)
54 int i;
55 unsigned char val = 0;
56 for (i = 0; i < 2; i++) {
57 unsigned char c = *q++;
58 val <<= 4;
59 if (c >= '0' && c <= '9')
60 val += c - '0';
61 else if (c >= 'a' && c <= 'f')
62 val += c - 'a' + 10;
63 else if (c >= 'A' && c <= 'F')
64 val += c - 'A' + 10;
65 else
66 return -1;
68 return val;
71 static char *url_decode_internal(const char **query, const char *stop_at,
72 struct strbuf *out, int decode_plus)
74 const char *q = *query;
76 do {
77 unsigned char c = *q;
79 if (!c)
80 break;
81 if (stop_at && strchr(stop_at, c)) {
82 q++;
83 break;
86 if (c == '%') {
87 int val = url_decode_char(q + 1);
88 if (0 <= val) {
89 strbuf_addch(out, val);
90 q += 3;
91 continue;
95 if (decode_plus && c == '+')
96 strbuf_addch(out, ' ');
97 else
98 strbuf_addch(out, c);
99 q++;
100 } while (1);
101 *query = q;
102 return strbuf_detach(out, NULL);
105 char *url_decode(const char *url)
107 struct strbuf out = STRBUF_INIT;
108 const char *colon = strchr(url, ':');
110 /* Skip protocol part if present */
111 if (colon && url < colon) {
112 strbuf_add(&out, url, colon - url);
113 url = colon;
115 return url_decode_internal(&url, NULL, &out, 0);
118 char *url_decode_parameter_name(const char **query)
120 struct strbuf out = STRBUF_INIT;
121 return url_decode_internal(query, "&=", &out, 1);
124 char *url_decode_parameter_value(const char **query)
126 struct strbuf out = STRBUF_INIT;
127 return url_decode_internal(query, "&", &out, 1);
130 void end_url_with_slash(struct strbuf *buf, const char *url)
132 strbuf_addstr(buf, url);
133 if (buf->len && buf->buf[buf->len - 1] != '/')
134 strbuf_addstr(buf, "/");
137 void str_end_url_with_slash(const char *url, char **dest) {
138 struct strbuf buf = STRBUF_INIT;
139 end_url_with_slash(&buf, url);
140 free(*dest);
141 *dest = strbuf_detach(&buf, NULL);