CodingGuidelines: add a note to help contributors catch some errors earlier.
[git/ydirson.git] / url.c
blob6a5495960f03ded65f0f5f8b8bd5c7cd98c0b05e
1 #include "cache.h"
3 int is_urlschemechar(int first_flag, int ch)
5 /*
6 * The set of valid URL schemes, as per STD66 (RFC3986) is
7 * '[A-Za-z][A-Za-z0-9+.-]*'. But use sightly looser check
8 * of '[A-Za-z0-9][A-Za-z0-9+.-]*' because earlier version
9 * of check used '[A-Za-z0-9]+' so not to break any remote
10 * helpers.
12 int alphanumeric, special;
13 alphanumeric = ch > 0 && isalnum(ch);
14 special = ch == '+' || ch == '-' || ch == '.';
15 return alphanumeric || (!first_flag && special);
18 int is_url(const char *url)
20 const char *url2, *first_slash;
22 if (!url)
23 return 0;
24 url2 = url;
25 first_slash = strchr(url, '/');
27 /* Input with no slash at all or slash first can't be URL. */
28 if (!first_slash || first_slash == url)
29 return 0;
30 /* Character before must be : and next must be /. */
31 if (first_slash[-1] != ':' || first_slash[1] != '/')
32 return 0;
33 /* There must be something before the :// */
34 if (first_slash == url + 1)
35 return 0;
37 * Check all characters up to first slash - 1. Only alphanum
38 * is allowed.
40 url2 = url;
41 while (url2 < first_slash - 1) {
42 if (!is_urlschemechar(url2 == url, (unsigned char)*url2))
43 return 0;
44 url2++;
47 /* Valid enough. */
48 return 1;
51 static int url_decode_char(const char *q)
53 int i;
54 unsigned char val = 0;
55 for (i = 0; i < 2; i++) {
56 unsigned char c = *q++;
57 val <<= 4;
58 if (c >= '0' && c <= '9')
59 val += c - '0';
60 else if (c >= 'a' && c <= 'f')
61 val += c - 'a' + 10;
62 else if (c >= 'A' && c <= 'F')
63 val += c - 'A' + 10;
64 else
65 return -1;
67 return val;
70 static char *url_decode_internal(const char **query, const char *stop_at,
71 struct strbuf *out, int decode_plus)
73 const char *q = *query;
75 do {
76 unsigned char c = *q;
78 if (!c)
79 break;
80 if (stop_at && strchr(stop_at, c)) {
81 q++;
82 break;
85 if (c == '%') {
86 int val = url_decode_char(q + 1);
87 if (0 <= val) {
88 strbuf_addch(out, val);
89 q += 3;
90 continue;
94 if (decode_plus && c == '+')
95 strbuf_addch(out, ' ');
96 else
97 strbuf_addch(out, c);
98 q++;
99 } while (1);
100 *query = q;
101 return strbuf_detach(out, NULL);
104 char *url_decode(const char *url)
106 struct strbuf out = STRBUF_INIT;
107 const char *colon = strchr(url, ':');
109 /* Skip protocol part if present */
110 if (colon && url < colon) {
111 strbuf_add(&out, url, colon - url);
112 url = colon;
114 return url_decode_internal(&url, NULL, &out, 0);
117 char *url_decode_parameter_name(const char **query)
119 struct strbuf out = STRBUF_INIT;
120 return url_decode_internal(query, "&=", &out, 1);
123 char *url_decode_parameter_value(const char **query)
125 struct strbuf out = STRBUF_INIT;
126 return url_decode_internal(query, "&", &out, 1);
129 void end_url_with_slash(struct strbuf *buf, const char *url)
131 strbuf_addstr(buf, url);
132 if (buf->len && buf->buf[buf->len - 1] != '/')
133 strbuf_addstr(buf, "/");
136 void str_end_url_with_slash(const char *url, char **dest) {
137 struct strbuf buf = STRBUF_INIT;
138 end_url_with_slash(&buf, url);
139 free(*dest);
140 *dest = strbuf_detach(&buf, NULL);