1 /***************************************************************************
2 * _ _ ____ _
3 * Project ___| | | | _ \| |
4 * / __| | | | |_) | |
5 * | (__| |_| | _ <| |___
6 * \___|\___/|_| \_\_____|
7 *
8 * Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al.
9 *
10 * This software is licensed as described in the file COPYING, which
11 * you should have received as part of this distribution. The terms
12 * are also available at https://curl.se/docs/copyright.html.
13 *
14 * You may opt to use, copy, modify, merge, publish, distribute and/or sell
15 * copies of the Software, and permit persons to whom the Software is
16 * furnished to do so, under the terms of the COPYING file.
17 *
18 * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
19 * KIND, either express or implied.
20 *
21 * SPDX-License-Identifier: curl
22 *
23 ***************************************************************************/
24
25 /* Escape and unescape URL encoding in strings. The functions return a new
26 * allocated string or NULL if an error occurred. */
27
28 #include "curl_setup.h"
29
30 #include <curl/curl.h>
31
32 struct Curl_easy;
33
34 #include "urldata.h"
35 #include "warnless.h"
36 #include "escape.h"
37 #include "strdup.h"
38 /* The last 3 #include files should be in this order */
39 #include "curl_printf.h"
40 #include "curl_memory.h"
41 #include "memdebug.h"
42
43 /* for ABI-compatibility with previous versions */
curl_escape(const char * string,int inlength)44 char *curl_escape(const char *string, int inlength)
45 {
46 return curl_easy_escape(NULL, string, inlength);
47 }
48
49 /* for ABI-compatibility with previous versions */
curl_unescape(const char * string,int length)50 char *curl_unescape(const char *string, int length)
51 {
52 return curl_easy_unescape(NULL, string, length, NULL);
53 }
54
55 /* Escapes for URL the given unescaped string of given length.
56 * 'data' is ignored since 7.82.0.
57 */
curl_easy_escape(CURL * data,const char * string,int inlength)58 char *curl_easy_escape(CURL *data, const char *string,
59 int inlength)
60 {
61 size_t length;
62 struct dynbuf d;
63 (void)data;
64
65 if(!string || (inlength < 0))
66 return NULL;
67
68 length = (inlength ? (size_t)inlength : strlen(string));
69 if(!length)
70 return strdup("");
71
72 Curl_dyn_init(&d, length * 3 + 1);
73
74 while(length--) {
75 /* treat the characters unsigned */
76 unsigned char in = (unsigned char)*string++;
77
78 if(ISUNRESERVED(in)) {
79 /* append this */
80 if(Curl_dyn_addn(&d, &in, 1))
81 return NULL;
82 }
83 else {
84 /* encode it */
85 const char hex[] = "0123456789ABCDEF";
86 char out[3]={'%'};
87 out[1] = hex[in >> 4];
88 out[2] = hex[in & 0xf];
89 if(Curl_dyn_addn(&d, out, 3))
90 return NULL;
91 }
92 }
93
94 return Curl_dyn_ptr(&d);
95 }
96
97 static const unsigned char hextable[] = {
98 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 0, 0, 0, 0, 0, /* 0x30 - 0x3f */
99 0, 10, 11, 12, 13, 14, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 0x40 - 0x4f */
100 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 0x50 - 0x5f */
101 0, 10, 11, 12, 13, 14, 15 /* 0x60 - 0x66 */
102 };
103
104 /* the input is a single hex digit */
105 #define onehex2dec(x) hextable[x - '0']
106
107 /*
108 * Curl_urldecode() URL decodes the given string.
109 *
110 * Returns a pointer to a malloced string in *ostring with length given in
111 * *olen. If length == 0, the length is assumed to be strlen(string).
112 *
113 * ctrl options:
114 * - REJECT_NADA: accept everything
115 * - REJECT_CTRL: rejects control characters (byte codes lower than 32) in
116 * the data
117 * - REJECT_ZERO: rejects decoded zero bytes
118 *
119 * The values for the enum starts at 2, to make the assert detect legacy
120 * invokes that used TRUE/FALSE (0 and 1).
121 */
122
Curl_urldecode(const char * string,size_t length,char ** ostring,size_t * olen,enum urlreject ctrl)123 CURLcode Curl_urldecode(const char *string, size_t length,
124 char **ostring, size_t *olen,
125 enum urlreject ctrl)
126 {
127 size_t alloc;
128 char *ns;
129
130 DEBUGASSERT(string);
131 DEBUGASSERT(ctrl >= REJECT_NADA); /* crash on TRUE/FALSE */
132
133 alloc = (length ? length : strlen(string));
134 ns = malloc(alloc + 1);
135
136 if(!ns)
137 return CURLE_OUT_OF_MEMORY;
138
139 /* store output string */
140 *ostring = ns;
141
142 while(alloc) {
143 unsigned char in = (unsigned char)*string;
144 if(('%' == in) && (alloc > 2) &&
145 ISXDIGIT(string[1]) && ISXDIGIT(string[2])) {
146 /* this is two hexadecimal digits following a '%' */
147 in = (unsigned char)(onehex2dec(string[1]) << 4) | onehex2dec(string[2]);
148
149 string += 3;
150 alloc -= 3;
151 }
152 else {
153 string++;
154 alloc--;
155 }
156
157 if(((ctrl == REJECT_CTRL) && (in < 0x20)) ||
158 ((ctrl == REJECT_ZERO) && (in == 0))) {
159 Curl_safefree(*ostring);
160 return CURLE_URL_MALFORMAT;
161 }
162
163 *ns++ = (char)in;
164 }
165 *ns = 0; /* terminate it */
166
167 if(olen)
168 /* store output size */
169 *olen = ns - *ostring;
170
171 return CURLE_OK;
172 }
173
174 /*
175 * Unescapes the given URL escaped string of given length. Returns a
176 * pointer to a malloced string with length given in *olen.
177 * If length == 0, the length is assumed to be strlen(string).
178 * If olen == NULL, no output length is stored.
179 * 'data' is ignored since 7.82.0.
180 */
curl_easy_unescape(CURL * data,const char * string,int length,int * olen)181 char *curl_easy_unescape(CURL *data, const char *string,
182 int length, int *olen)
183 {
184 char *str = NULL;
185 (void)data;
186 if(string && (length >= 0)) {
187 size_t inputlen = (size_t)length;
188 size_t outputlen;
189 CURLcode res = Curl_urldecode(string, inputlen, &str, &outputlen,
190 REJECT_NADA);
191 if(res)
192 return NULL;
193
194 if(olen) {
195 if(outputlen <= (size_t) INT_MAX)
196 *olen = curlx_uztosi(outputlen);
197 else
198 /* too large to return in an int, fail! */
199 Curl_safefree(str);
200 }
201 }
202 return str;
203 }
204
205 /* For operating systems/environments that use different malloc/free
206 systems for the app and for this library, we provide a free that uses
207 the library's memory system */
curl_free(void * p)208 void curl_free(void *p)
209 {
210 free(p);
211 }
212
213 /*
214 * Curl_hexencode()
215 *
216 * Converts binary input to lowercase hex-encoded ASCII output.
217 * Null-terminated.
218 */
Curl_hexencode(const unsigned char * src,size_t len,unsigned char * out,size_t olen)219 void Curl_hexencode(const unsigned char *src, size_t len, /* input length */
220 unsigned char *out, size_t olen) /* output buffer size */
221 {
222 const char *hex = "0123456789abcdef";
223 DEBUGASSERT(src && len && (olen >= 3));
224 if(src && len && (olen >= 3)) {
225 while(len-- && (olen >= 3)) {
226 /* clang-tidy warns on this line without this comment: */
227 /* NOLINTNEXTLINE(clang-analyzer-core.UndefinedBinaryOperatorResult) */
228 *out++ = (unsigned char)hex[(*src & 0xF0) >> 4];
229 *out++ = (unsigned char)hex[*src & 0x0F];
230 ++src;
231 olen -= 2;
232 }
233 *out = 0;
234 }
235 else if(olen)
236 *out = 0;
237 }
238
