1 #include "cache.h"
3 int is_urlschemechar(int first_flag, int ch)
4 {
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.
11 */
12 int alphanumeric, special;
13 alphanumeric = ch > 0 && isalnum(ch);
14 special = ch == '+' || ch == '-' || ch == '.';
15 return alphanumeric || (!first_flag && special);
16 }
18 int is_url(const char *url)
19 {
20 /* Is "scheme" part reasonable? */
21 if (!url || !is_urlschemechar(1, *url++))
22 return 0;
23 while (*url && *url != ':') {
24 if (!is_urlschemechar(0, *url++))
25 return 0;
26 }
27 /* We've seen "scheme"; we want colon-slash-slash */
28 return (url[0] == ':' && url[1] == '/' && url[2] == '/');
29 }
31 static int url_decode_char(const char *q)
32 {
33 int i;
34 unsigned char val = 0;
35 for (i = 0; i < 2; i++) {
36 unsigned char c = *q++;
37 val <<= 4;
38 if (c >= '0' && c <= '9')
39 val += c - '0';
40 else if (c >= 'a' && c <= 'f')
41 val += c - 'a' + 10;
42 else if (c >= 'A' && c <= 'F')
43 val += c - 'A' + 10;
44 else
45 return -1;
46 }
47 return val;
48 }
50 static char *url_decode_internal(const char **query, const char *stop_at,
51 struct strbuf *out, int decode_plus)
52 {
53 const char *q = *query;
55 do {
56 unsigned char c = *q;
58 if (!c)
59 break;
60 if (stop_at && strchr(stop_at, c)) {
61 q++;
62 break;
63 }
65 if (c == '%') {
66 int val = url_decode_char(q + 1);
67 if (0 <= val) {
68 strbuf_addch(out, val);
69 q += 3;
70 continue;
71 }
72 }
74 if (decode_plus && c == '+')
75 strbuf_addch(out, ' ');
76 else
77 strbuf_addch(out, c);
78 q++;
79 } while (1);
80 *query = q;
81 return strbuf_detach(out, NULL);
82 }
84 char *url_decode(const char *url)
85 {
86 struct strbuf out = STRBUF_INIT;
87 const char *colon = strchr(url, ':');
89 /* Skip protocol part if present */
90 if (colon && url < colon) {
91 strbuf_add(&out, url, colon - url);
92 url = colon;
93 }
94 return url_decode_internal(&url, NULL, &out, 0);
95 }
97 char *url_decode_parameter_name(const char **query)
98 {
99 struct strbuf out = STRBUF_INIT;
100 return url_decode_internal(query, "&=", &out, 1);
101 }
103 char *url_decode_parameter_value(const char **query)
104 {
105 struct strbuf out = STRBUF_INIT;
106 return url_decode_internal(query, "&", &out, 1);
107 }
109 void end_url_with_slash(struct strbuf *buf, const char *url)
110 {
111 strbuf_addstr(buf, url);
112 if (buf->len && buf->buf[buf->len - 1] != '/')
113 strbuf_addstr(buf, "/");
114 }
116 void str_end_url_with_slash(const char *url, char **dest) {
117 struct strbuf buf = STRBUF_INIT;
118 end_url_with_slash(&buf, url);
119 free(*dest);
120 *dest = strbuf_detach(&buf, NULL);
121 }