/work/workdir/UnpackedTarball/libexttextcat/src/common.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ |
2 | | /* |
3 | | * common.c -- miscellanious helper functions. |
4 | | * |
5 | | * Copyright (c) 2003, WiseGuys Internet B.V. |
6 | | * All rights reserved. |
7 | | * |
8 | | * Redistribution and use in source and binary forms, with or without |
9 | | * modification, are permitted provided that the following conditions |
10 | | * are met: |
11 | | * |
12 | | * - Redistributions of source code must retain the above copyright |
13 | | * notice, this list of conditions and the following disclaimer. |
14 | | * |
15 | | * - Redistributions in binary form must reproduce the above copyright |
16 | | * notice, this list of conditions and the following disclaimer in the |
17 | | * documentation and/or other materials provided with the |
18 | | * distribution. |
19 | | * |
20 | | * - Neither the name of the WiseGuys Internet B.V. nor the names of |
21 | | * its contributors may be used to endorse or promote products derived |
22 | | * from this software without specific prior written permission. |
23 | | * |
24 | | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
25 | | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
26 | | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
27 | | * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
28 | | * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
29 | | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
30 | | * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
31 | | * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
32 | | * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
33 | | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
34 | | * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
35 | | */ |
36 | | #ifdef HAVE_CONFIG_H |
37 | | #include "config.h" |
38 | | #endif |
39 | | #include <stdio.h> |
40 | | #include <string.h> |
41 | | #include <stdarg.h> |
42 | | #include <ctype.h> |
43 | | #include "common_impl.h" |
44 | | |
45 | | extern char *wg_getline(char *line, int size, FILE * fp) |
46 | 0 | { |
47 | 0 | char *p; |
48 | |
|
49 | 0 | if (fgets(line, size, fp) == NULL) |
50 | 0 | { |
51 | 0 | return NULL; |
52 | 0 | } |
53 | | |
54 | | /** kill term null **/ |
55 | 0 | if ((p = strpbrk(line, "\n\r"))) |
56 | 0 | { |
57 | 0 | *p = '\0'; |
58 | 0 | } |
59 | |
|
60 | 0 | return line; |
61 | 0 | } |
62 | | |
63 | | |
64 | | |
65 | | /* |
66 | | * wg_split: split a line into segments, using whitespace-sequences as separators. |
67 | | * |
68 | | * ARGUMENTS: |
69 | | * - result: |
70 | | * |
71 | | * After the split, this array contains pointers to the start of each |
72 | | * detected segment. Must be preallocated and at least as large as |
73 | | * maxsegments. The pointers point into the dest buffer. |
74 | | * |
75 | | * - dest: |
76 | | * |
77 | | * String into which result points as an index. Must be preallocated, and |
78 | | * at least as big as src. You can use src as dest, but in that case src |
79 | | * is overwritten! |
80 | | * |
81 | | * - src: |
82 | | * |
83 | | * The string to split. Sequences of whitespace are treated as separators, unless |
84 | | * escaped. There are two ways to escape: by using single quotes (anything |
85 | | * between single quotes is treated as one segment), or by using a backslash |
86 | | * to escape the next character. The backslash escape works inside quotation |
87 | | * as well. |
88 | | * |
89 | | * Example: |
90 | | * |
91 | | * "It\'s very\ easy 'to use WiseGuys\' wg_split()' function" is split into: |
92 | | * |
93 | | * "It's" |
94 | | * "very easy" |
95 | | * "to use WiseGuys' wg_split()" |
96 | | * "function" |
97 | | * |
98 | | * - maxsegments: |
99 | | * |
100 | | * The maximum number of segments. If the splitter runs out of segments, |
101 | | * the remainder of the string is stored in the last segment. |
102 | | * |
103 | | * RETURN VALUE: |
104 | | * The number of segments found. |
105 | | */ |
106 | | unsigned int wg_split(char **result, char *dest, char *src, int maxsegments) |
107 | 0 | { |
108 | 0 | char *p = src; |
109 | 0 | char *w = dest; |
110 | 0 | int cnt = 0; |
111 | 0 | int state = 0; |
112 | |
|
113 | 0 | if (maxsegments == 0) |
114 | 0 | { |
115 | 0 | return 0; |
116 | 0 | } |
117 | | |
118 | 0 | maxsegments--; |
119 | |
|
120 | 0 | while (cnt < maxsegments && *p) |
121 | 0 | { |
122 | |
|
123 | 0 | switch (state) |
124 | 0 | { |
125 | 0 | case 0: |
126 | | /*** Skip spaces ***/ |
127 | 0 | while (isspace((unsigned char)*p)) |
128 | 0 | { |
129 | 0 | p++; |
130 | 0 | } |
131 | | |
132 | | /*** Start segment ***/ |
133 | 0 | result[cnt] = w; |
134 | 0 | cnt++; |
135 | 0 | state = 1; |
136 | | |
137 | | /*** fall through **/ |
138 | |
|
139 | 0 | case 1: |
140 | | /*** Unquoted segment ***/ |
141 | 0 | while (*p) |
142 | 0 | { |
143 | 0 | if (isspace((unsigned char)*p)) |
144 | 0 | { |
145 | 0 | *w++ = '\0'; |
146 | 0 | p++; |
147 | 0 | state = 0; |
148 | 0 | break; |
149 | 0 | } |
150 | 0 | else if (*p == '\'') |
151 | 0 | { |
152 | | /*** Start quotation ***/ |
153 | 0 | p++; |
154 | 0 | state = 2; |
155 | 0 | break; |
156 | 0 | } |
157 | 0 | else if (*p == '\\' && p[1]) |
158 | 0 | { |
159 | | /*** Escape ***/ |
160 | 0 | p++; |
161 | 0 | *w++ = *p++; |
162 | 0 | } |
163 | 0 | else |
164 | 0 | { |
165 | 0 | *w++ = *p++; |
166 | 0 | } |
167 | 0 | } |
168 | 0 | break; |
169 | | |
170 | 0 | case 2: |
171 | | /*** Inside quotes ***/ |
172 | 0 | while (*p) |
173 | 0 | { |
174 | 0 | if (*p == '\'') |
175 | 0 | { |
176 | 0 | p++; |
177 | 0 | break; |
178 | 0 | } |
179 | 0 | else if (*p == '\\' && p[1]) |
180 | 0 | { |
181 | | /*** Escape ***/ |
182 | 0 | p++; |
183 | 0 | *w++ = *p++; |
184 | 0 | } |
185 | 0 | else |
186 | 0 | { |
187 | 0 | *w++ = *p++; |
188 | 0 | } |
189 | 0 | } |
190 | 0 | state = 1; |
191 | 0 | break; |
192 | |
|
193 | 0 | } |
194 | 0 | } |
195 | | |
196 | 0 | if (!*p) |
197 | 0 | { |
198 | 0 | *w = '\0'; |
199 | 0 | return cnt; |
200 | 0 | } |
201 | | |
202 | | /*** We ran out of segments; copy the remainder of the string into last segment ***/ |
203 | 0 | result[cnt++] = w; |
204 | 0 | while (*p) |
205 | 0 | { |
206 | 0 | *w++ = *p++; |
207 | 0 | } |
208 | 0 | *w = '\0'; |
209 | 0 | return cnt; |
210 | 0 | } |
211 | | |
212 | | /** |
213 | | * wg_strgmov -- a guarded strcpy() variation |
214 | | * |
215 | | * copies src to dest (including terminating zero), and returns |
216 | | * pointer to position of terminating zero in dest. The function is |
217 | | * guaranteed not to write past destlimit. If the copy couldn't be |
218 | | * finished, the function returns NULL after restoring the first |
219 | | * character in dest for your convenience (since this is usually a zero). |
220 | | */ |
221 | | char *wg_strgmov(char *dest, const char *src, const char *destlimit) |
222 | 0 | { |
223 | 0 | char tmp, *w; |
224 | |
|
225 | 0 | if (!dest || dest >= destlimit) |
226 | 0 | { |
227 | 0 | return NULL; |
228 | 0 | } |
229 | | |
230 | 0 | tmp = *dest; |
231 | 0 | w = dest; |
232 | |
|
233 | 0 | while (*src) |
234 | 0 | { |
235 | |
|
236 | 0 | *w++ = *src++; |
237 | |
|
238 | 0 | if (w == destlimit) |
239 | 0 | { |
240 | | /*** restore old situation ***/ |
241 | 0 | *dest = tmp; |
242 | 0 | return NULL; |
243 | 0 | } |
244 | 0 | } |
245 | | |
246 | 0 | *w = '\0'; |
247 | 0 | return w; |
248 | |
|
249 | 0 | } |
250 | | |
251 | | /* |
252 | | * wg_trim() -- remove whitespace surrounding a string. |
253 | | * |
254 | | * Example: " bla bla bla " becomes "bla bla bla" after trimming. |
255 | | * |
256 | | * ARGUMENTS |
257 | | * - dest : After the trim, this will point to the trimmed string. |
258 | | * Must be preallocated and at least as big as src. You can |
259 | | * use src as dest. |
260 | | * - src : Points to the string to be trimmed. |
261 | | * |
262 | | * RETURNS: |
263 | | * dest |
264 | | */ |
265 | | char *wg_trim(char *dest, const char *src) |
266 | 0 | { |
267 | 0 | char *lastnonspace = &dest[-1]; |
268 | 0 | const char *p = src; |
269 | 0 | char *w = dest; |
270 | |
|
271 | 0 | while (isspace((unsigned char)*p)) |
272 | 0 | { |
273 | 0 | p++; |
274 | 0 | } |
275 | 0 | while (*p) |
276 | 0 | { |
277 | 0 | if (!isspace((unsigned char)*p)) |
278 | 0 | { |
279 | 0 | lastnonspace = w; |
280 | 0 | } |
281 | 0 | *w++ = *p++; |
282 | 0 | } |
283 | 0 | lastnonspace[1] = '\0'; |
284 | |
|
285 | 0 | return dest; |
286 | 0 | } |
287 | | |
288 | | /* vim:set shiftwidth=4 softtabstop=4 expandtab: */ |