/src/curl/lib/curlx/strparse.c
Line | Count | Source |
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 | | #include "strparse.h" |
25 | | |
26 | | void curlx_str_init(struct Curl_str *out) |
27 | 732k | { |
28 | 732k | out->str = NULL; |
29 | 732k | out->len = 0; |
30 | 732k | } |
31 | | |
32 | | void curlx_str_assign(struct Curl_str *out, const char *str, size_t len) |
33 | 13.7k | { |
34 | 13.7k | out->str = str; |
35 | 13.7k | out->len = len; |
36 | 13.7k | } |
37 | | |
38 | | /* Get a word until the first DELIM or end of string. At least one byte long. |
39 | | return non-zero on error */ |
40 | | int curlx_str_until(const char **linep, struct Curl_str *out, |
41 | | const size_t max, char delim) |
42 | 413k | { |
43 | 413k | const char *s = *linep; |
44 | 413k | size_t len = 0; |
45 | 413k | DEBUGASSERT(linep && *linep && out && max && delim); |
46 | | |
47 | 413k | curlx_str_init(out); |
48 | 9.56M | while(*s && (*s != delim)) { |
49 | 9.14M | s++; |
50 | 9.14M | if(++len > max) { |
51 | 572 | return STRE_BIG; |
52 | 572 | } |
53 | 9.14M | } |
54 | 412k | if(!len) |
55 | 22.6k | return STRE_SHORT; |
56 | 389k | out->str = *linep; |
57 | 389k | out->len = len; |
58 | 389k | *linep = s; /* point to the first byte after the word */ |
59 | 389k | return STRE_OK; |
60 | 412k | } |
61 | | |
62 | | /* Get a word until the first space or end of string. At least one byte long. |
63 | | return non-zero on error */ |
64 | | int curlx_str_word(const char **linep, struct Curl_str *out, const size_t max) |
65 | 171k | { |
66 | 171k | return curlx_str_until(linep, out, max, ' '); |
67 | 171k | } |
68 | | |
69 | | /* Get a word until a newline byte or end of string. At least one byte long. |
70 | | return non-zero on error */ |
71 | | int curlx_str_untilnl(const char **linep, struct Curl_str *out, |
72 | | const size_t max) |
73 | 95.2k | { |
74 | 95.2k | const char *s = *linep; |
75 | 95.2k | size_t len = 0; |
76 | 95.2k | DEBUGASSERT(linep && *linep && out && max); |
77 | | |
78 | 95.2k | curlx_str_init(out); |
79 | 9.70M | while(*s && !ISNEWLINE(*s)) { |
80 | 9.61M | s++; |
81 | 9.61M | if(++len > max) |
82 | 2 | return STRE_BIG; |
83 | 9.61M | } |
84 | 95.2k | if(!len) |
85 | 6.31k | return STRE_SHORT; |
86 | 88.9k | out->str = *linep; |
87 | 88.9k | out->len = len; |
88 | 88.9k | *linep = s; /* point to the first byte after the word */ |
89 | 88.9k | return STRE_OK; |
90 | 95.2k | } |
91 | | |
92 | | /* Get a "quoted" word. Escaped quotes are supported. |
93 | | return non-zero on error */ |
94 | | int curlx_str_quotedword(const char **linep, struct Curl_str *out, |
95 | | const size_t max) |
96 | 6.30k | { |
97 | 6.30k | const char *s = *linep; |
98 | 6.30k | size_t len = 0; |
99 | 6.30k | DEBUGASSERT(linep && *linep && out && max); |
100 | | |
101 | 6.30k | curlx_str_init(out); |
102 | 6.30k | if(*s != '\"') |
103 | 13 | return STRE_BEGQUOTE; |
104 | 6.28k | s++; |
105 | 137k | while(*s && (*s != '\"')) { |
106 | 131k | if(*s == '\\' && s[1]) { |
107 | 996 | s++; |
108 | 996 | if(++len > max) |
109 | 0 | return STRE_BIG; |
110 | 996 | } |
111 | 131k | s++; |
112 | 131k | if(++len > max) |
113 | 0 | return STRE_BIG; |
114 | 131k | } |
115 | 6.28k | if(*s != '\"') |
116 | 1.61k | return STRE_ENDQUOTE; |
117 | 4.67k | out->str = (*linep) + 1; |
118 | 4.67k | out->len = len; |
119 | 4.67k | *linep = s + 1; |
120 | 4.67k | return STRE_OK; |
121 | 6.28k | } |
122 | | |
123 | | /* Advance over a single character. |
124 | | return non-zero on error */ |
125 | | int curlx_str_single(const char **linep, char byte) |
126 | 4.66M | { |
127 | 4.66M | DEBUGASSERT(linep && *linep); |
128 | 4.66M | if(**linep != byte) |
129 | 3.71M | return STRE_BYTE; |
130 | 949k | (*linep)++; /* move over it */ |
131 | 949k | return STRE_OK; |
132 | 4.66M | } |
133 | | |
134 | | /* Advance over a single space. |
135 | | return non-zero on error */ |
136 | | int curlx_str_singlespace(const char **linep) |
137 | 171k | { |
138 | 171k | return curlx_str_single(linep, ' '); |
139 | 171k | } |
140 | | |
141 | | /* given an ASCII character and max ascii, return TRUE if valid */ |
142 | | #define valid_digit(x, m) \ |
143 | 28.1M | (((x) >= '0') && ((x) <= m) && curlx_hexasciitable[(x) - '0']) |
144 | | |
145 | | /* We use 16 for the zero index (and the necessary bitwise AND in the loop) |
146 | | to be able to have a non-zero value there to make valid_digit() able to |
147 | | use the info */ |
148 | | const unsigned char curlx_hexasciitable[] = { |
149 | | 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, /* 0x30: 0 - 9 */ |
150 | | 0, 0, 0, 0, 0, 0, 0, |
151 | | 10, 11, 12, 13, 14, 15, /* 0x41: A - F */ |
152 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
153 | | 10, 11, 12, 13, 14, 15 /* 0x61: a - f */ |
154 | | }; |
155 | | |
156 | | /* no support for 0x prefix nor leading spaces */ |
157 | | static int str_num_base(const char **linep, curl_off_t *nump, curl_off_t max, |
158 | | int base) /* 8, 10 or 16, nothing else */ |
159 | 14.4M | { |
160 | 14.4M | curl_off_t num = 0; |
161 | 14.4M | const char *p; |
162 | 14.4M | int m = (base == 10) ? '9' : /* the largest digit possible */ |
163 | 14.4M | (base == 16) ? 'f' : '7'; |
164 | 14.4M | DEBUGASSERT(linep && *linep && nump); |
165 | 14.4M | DEBUGASSERT((base == 8) || (base == 10) || (base == 16)); |
166 | 14.4M | DEBUGASSERT(max >= 0); /* mostly to catch SIZE_MAX, which is too large */ |
167 | 14.4M | *nump = 0; |
168 | 14.4M | p = *linep; |
169 | 14.4M | if(!valid_digit(*p, m)) |
170 | 7.55M | return STRE_NO_NUM; |
171 | 6.87M | if(max < base) { |
172 | | /* special-case low max scenario because check needs to be different */ |
173 | 0 | do { |
174 | 0 | int n = curlx_hexval(*p++); |
175 | 0 | num = num * base + n; |
176 | 0 | if(num > max) |
177 | 0 | return STRE_OVERFLOW; |
178 | 0 | } while(valid_digit(*p, m)); |
179 | 0 | } |
180 | 6.87M | else { |
181 | 13.7M | do { |
182 | 13.7M | int n = curlx_hexval(*p++); |
183 | 13.7M | if(num > ((max - n) / base)) |
184 | 7.23k | return STRE_OVERFLOW; |
185 | 13.7M | num = num * base + n; |
186 | 13.7M | } while(valid_digit(*p, m)); |
187 | 6.87M | } |
188 | 6.86M | *nump = num; |
189 | 6.86M | *linep = p; |
190 | 6.86M | return STRE_OK; |
191 | 6.87M | } |
192 | | |
193 | | /* Get an unsigned decimal number with no leading space or minus. Leading |
194 | | zeroes are accepted. return non-zero on error */ |
195 | | int curlx_str_number(const char **linep, curl_off_t *nump, curl_off_t max) |
196 | 14.3M | { |
197 | 14.3M | return str_num_base(linep, nump, max, 10); |
198 | 14.3M | } |
199 | | |
200 | | /* Get an unsigned hexadecimal number with no leading space or minus and no |
201 | | "0x" support. Leading zeroes are accepted. return non-zero on error */ |
202 | | int curlx_str_hex(const char **linep, curl_off_t *nump, curl_off_t max) |
203 | 2.55k | { |
204 | 2.55k | return str_num_base(linep, nump, max, 16); |
205 | 2.55k | } |
206 | | |
207 | | /* Get an unsigned octal number with no leading space or minus and no "0" |
208 | | prefix support. Leading zeroes are accepted. return non-zero on error */ |
209 | | int curlx_str_octal(const char **linep, curl_off_t *nump, curl_off_t max) |
210 | 69.1k | { |
211 | 69.1k | return str_num_base(linep, nump, max, 8); |
212 | 69.1k | } |
213 | | |
214 | | /* |
215 | | * Parse a positive number up to 63-bit number written in ASCII. Skip leading |
216 | | * blanks. No support for prefixes. |
217 | | */ |
218 | | int curlx_str_numblanks(const char **str, curl_off_t *num) |
219 | 6.87k | { |
220 | 6.87k | curlx_str_passblanks(str); |
221 | 6.87k | return curlx_str_number(str, num, CURL_OFF_T_MAX); |
222 | 6.87k | } |
223 | | |
224 | | /* CR or LF |
225 | | return non-zero on error */ |
226 | | int curlx_str_newline(const char **linep) |
227 | 2.74k | { |
228 | 2.74k | DEBUGASSERT(linep && *linep); |
229 | 2.74k | if(ISNEWLINE(**linep)) { |
230 | 2.71k | (*linep)++; |
231 | 2.71k | return STRE_OK; /* yessir */ |
232 | 2.71k | } |
233 | 28 | return STRE_NEWLINE; |
234 | 2.74k | } |
235 | | |
236 | | #ifndef WITHOUT_LIBCURL |
237 | | /* case insensitive compare that the parsed string matches the given string. |
238 | | Returns non-zero on match. */ |
239 | | int curlx_str_casecompare(struct Curl_str *str, const char *check) |
240 | 1.52M | { |
241 | 1.52M | size_t clen = check ? strlen(check) : 0; |
242 | 1.52M | return ((str->len == clen) && curl_strnequal(str->str, check, clen)); |
243 | 1.52M | } |
244 | | #endif |
245 | | |
246 | | /* case sensitive string compare. Returns non-zero on match. */ |
247 | | int curlx_str_cmp(struct Curl_str *str, const char *check) |
248 | 47.6k | { |
249 | 47.6k | if(check) { |
250 | 47.6k | size_t clen = strlen(check); |
251 | 47.6k | return ((str->len == clen) && !strncmp(str->str, check, clen)); |
252 | 47.6k | } |
253 | 0 | return !!(str->len); |
254 | 47.6k | } |
255 | | |
256 | | /* Trim off 'num' number of bytes from the beginning (left side) of the |
257 | | string. If 'num' is larger than the string, return error. */ |
258 | | int curlx_str_nudge(struct Curl_str *str, size_t num) |
259 | 114k | { |
260 | 114k | if(num <= str->len) { |
261 | 114k | str->str += num; |
262 | 114k | str->len -= num; |
263 | 114k | return STRE_OK; |
264 | 114k | } |
265 | 0 | return STRE_OVERFLOW; |
266 | 114k | } |
267 | | |
268 | | /* Get the following character sequence that consists only of bytes not |
269 | | present in the 'reject' string. Like strcspn(). */ |
270 | | int curlx_str_cspn(const char **linep, struct Curl_str *out, |
271 | | const char *reject) |
272 | 849k | { |
273 | 849k | const char *s = *linep; |
274 | 849k | size_t len; |
275 | 849k | DEBUGASSERT(linep && *linep); |
276 | | |
277 | 849k | len = strcspn(s, reject); |
278 | 849k | if(len) { |
279 | 757k | out->str = s; |
280 | 757k | out->len = len; |
281 | 757k | *linep = &s[len]; |
282 | 757k | return STRE_OK; |
283 | 757k | } |
284 | 92.1k | curlx_str_init(out); |
285 | 92.1k | return STRE_SHORT; |
286 | 849k | } |
287 | | |
288 | | /* remove ISBLANK()s from both ends of the string */ |
289 | | void curlx_str_trimblanks(struct Curl_str *out) |
290 | 781k | { |
291 | 895k | while(out->len && ISBLANK(*out->str)) |
292 | 113k | curlx_str_nudge(out, 1); |
293 | | |
294 | | /* trim trailing spaces and tabs */ |
295 | 812k | while(out->len && ISBLANK(out->str[out->len - 1])) |
296 | 31.2k | out->len--; |
297 | 781k | } |
298 | | |
299 | | /* increase the pointer until it has moved over all blanks */ |
300 | | void curlx_str_passblanks(const char **linep) |
301 | 942k | { |
302 | 1.06M | while(ISBLANK(**linep)) |
303 | 122k | (*linep)++; /* move over it */ |
304 | 942k | } |