diff options
author | Bart Polot <bart@net.in.tum.de> | 2012-12-17 12:35:19 +0000 |
---|---|---|
committer | Bart Polot <bart@net.in.tum.de> | 2012-12-17 12:35:19 +0000 |
commit | 11cc044427e734c83af65ee340dd10ec70c97b1c (patch) | |
tree | 2c807dc234644519ff2f341cf6ca2f1cb2a084f2 /src/regex/regex_test_lib.c | |
parent | 07ce0aaf09b4bcdf37b1524a9a497d4a185aba91 (diff) | |
download | gnunet-11cc044427e734c83af65ee340dd10ec70c97b1c.tar.gz gnunet-11cc044427e734c83af65ee340dd10ec70c97b1c.zip |
- move regex_from_file and combine to separate lib file
Diffstat (limited to 'src/regex/regex_test_lib.c')
-rw-r--r-- | src/regex/regex_test_lib.c | 248 |
1 files changed, 248 insertions, 0 deletions
diff --git a/src/regex/regex_test_lib.c b/src/regex/regex_test_lib.c new file mode 100644 index 000000000..d02afa942 --- /dev/null +++ b/src/regex/regex_test_lib.c | |||
@@ -0,0 +1,248 @@ | |||
1 | /* | ||
2 | * This file is part of GNUnet | ||
3 | * (C) 2012 Christian Grothoff (and other contributing authors) | ||
4 | * | ||
5 | * GNUnet is free software; you can redistribute it and/or modify | ||
6 | * it under the terms of the GNU General Public License as published | ||
7 | * by the Free Software Foundation; either version 3, or (at your | ||
8 | * option) any later version. | ||
9 | * | ||
10 | * GNUnet is distributed in the hope that it will be useful, but | ||
11 | * WITHOUT ANY WARRANTY; without even the implied warranty of | ||
12 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | ||
13 | * General Public License for more details. | ||
14 | * | ||
15 | * You should have received a copy of the GNU General Public License | ||
16 | * along with GNUnet; see the file COPYING. If not, write to the | ||
17 | * Free Software Foundation, Inc., 59 Temple Place - Suite 330, | ||
18 | * Boston, MA 02111-1307, USA. | ||
19 | */ | ||
20 | /** | ||
21 | * @file src/regex/regex_test_lib.c | ||
22 | * @brief library to read regexes representing IP networks from a file. | ||
23 | * and simplyfinying the into one big regex, in order to run | ||
24 | * tests (regex performance, mesh profiler). | ||
25 | * @author Bartlomiej Polot | ||
26 | */ | ||
27 | |||
28 | #include "platform.h" | ||
29 | #include "gnunet_util_lib.h" | ||
30 | |||
31 | struct RegexCombineCtx { | ||
32 | struct RegexCombineCtx *next; | ||
33 | struct RegexCombineCtx *prev; | ||
34 | |||
35 | struct RegexCombineCtx *head; | ||
36 | struct RegexCombineCtx *tail; | ||
37 | |||
38 | char *s; | ||
39 | }; | ||
40 | |||
41 | |||
42 | static char * | ||
43 | regex_combine (struct RegexCombineCtx *ctx) | ||
44 | { | ||
45 | struct RegexCombineCtx *p; | ||
46 | size_t len; | ||
47 | char *regex; | ||
48 | char *tmp; | ||
49 | char *s; | ||
50 | |||
51 | if (NULL != ctx->s) | ||
52 | GNUNET_asprintf (®ex, "%s(", ctx->s); | ||
53 | else | ||
54 | regex = GNUNET_strdup ("("); | ||
55 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "prefix: %s\n", regex); | ||
56 | |||
57 | for (p = ctx->head; NULL != p; p = p->next) | ||
58 | { | ||
59 | s = regex_combine (p); | ||
60 | GNUNET_asprintf (&tmp, "%s%s|", regex, s); | ||
61 | GNUNET_free_non_null (s); | ||
62 | GNUNET_free_non_null (regex); | ||
63 | regex = tmp; | ||
64 | } | ||
65 | len = strlen (regex); | ||
66 | if (1 == len) | ||
67 | return GNUNET_strdup (""); | ||
68 | |||
69 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "pre-partial: %s\n", regex); | ||
70 | if ('|' == regex[len - 1]) | ||
71 | regex[len - 1] = ')'; | ||
72 | if ('(' == regex[len - 1]) | ||
73 | regex[len - 1] = '\0'; | ||
74 | |||
75 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "partial: %s\n", regex); | ||
76 | return regex; | ||
77 | } | ||
78 | |||
79 | static void | ||
80 | regex_add (struct RegexCombineCtx *ctx, const char *regex) | ||
81 | { | ||
82 | struct RegexCombineCtx *p; | ||
83 | const char *rest; | ||
84 | |||
85 | rest = ®ex[1]; | ||
86 | for (p = ctx->head; NULL != p; p = p->next) | ||
87 | { | ||
88 | if (p->s[0] == regex[0]) | ||
89 | { | ||
90 | if (1 == strlen(p->s)) | ||
91 | { | ||
92 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "common char %s\n", p->s); | ||
93 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "adding %s\n", rest); | ||
94 | regex_add (p, rest); | ||
95 | } | ||
96 | else | ||
97 | { | ||
98 | struct RegexCombineCtx *new; | ||
99 | new = GNUNET_malloc (sizeof (struct RegexCombineCtx)); | ||
100 | new->s = GNUNET_strdup (&p->s[1]); | ||
101 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " p has now %s\n", p->s); | ||
102 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " p will have %.1s\n", p->s); | ||
103 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " regex is %s\n", regex); | ||
104 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " new has now %s\n", new->s); | ||
105 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " rest is now %s\n", rest); | ||
106 | p->s[1] = '\0'; /* dont realloc */ | ||
107 | GNUNET_CONTAINER_DLL_insert (p->head, p->tail, new); | ||
108 | regex_add (p, rest); | ||
109 | } | ||
110 | return; | ||
111 | } | ||
112 | } | ||
113 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " no match\n"); | ||
114 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " new state %s\n", regex); | ||
115 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, " under %s\n", ctx->s); | ||
116 | p = GNUNET_malloc (sizeof (struct RegexCombineCtx)); | ||
117 | p->s = GNUNET_strdup (regex); | ||
118 | GNUNET_CONTAINER_DLL_insert (ctx->head, ctx->tail, p); | ||
119 | } | ||
120 | |||
121 | |||
122 | /** | ||
123 | * Return a prefix-combine regex that matches the same strings as | ||
124 | * any of the original regexes. | ||
125 | * | ||
126 | * WARNING: only useful for reading specific regexes for specific applications, | ||
127 | * namely the gnunet-regex-profiler / gnunet-regex-daemon. | ||
128 | * This function DOES NOT support arbitrary regex combining. | ||
129 | */ | ||
130 | char * | ||
131 | GNUNET_REGEX_combine (char * const regexes[]) | ||
132 | { | ||
133 | unsigned int i; | ||
134 | char *combined; | ||
135 | const char *current; | ||
136 | struct RegexCombineCtx ctx; | ||
137 | |||
138 | memset (&ctx, 0, sizeof (struct RegexCombineCtx)); | ||
139 | for (i = 0; regexes[i]; i++) | ||
140 | { | ||
141 | current = regexes[i]; | ||
142 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "Regex %u: %s\n", i, current); | ||
143 | regex_add (&ctx, current); | ||
144 | } | ||
145 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, "\nCombining...\n"); | ||
146 | |||
147 | combined = regex_combine (&ctx); | ||
148 | |||
149 | return combined; | ||
150 | } | ||
151 | |||
152 | |||
153 | /** | ||
154 | * Read a set of regexes from a file, one per line and return them in an array | ||
155 | * suitable for GNUNET_REGEX_combine. | ||
156 | * The array must be free'd using GNUNET_REGEX_free_from_file. | ||
157 | * | ||
158 | * @param filename Name of the file containing the regexes. | ||
159 | * | ||
160 | * @return A newly allocated, NULL terminated array of regexes. | ||
161 | */ | ||
162 | char ** | ||
163 | GNUNET_REGEX_read_from_file (const char *filename) | ||
164 | { | ||
165 | struct GNUNET_DISK_FileHandle *f; | ||
166 | unsigned int nr; | ||
167 | unsigned int offset; | ||
168 | off_t size; | ||
169 | size_t len; | ||
170 | char *buffer; | ||
171 | char *regex; | ||
172 | char **regexes; | ||
173 | |||
174 | f = GNUNET_DISK_file_open (filename, | ||
175 | GNUNET_DISK_OPEN_READ, | ||
176 | GNUNET_DISK_PERM_NONE); | ||
177 | if (NULL == f) | ||
178 | { | ||
179 | GNUNET_log (GNUNET_ERROR_TYPE_ERROR, | ||
180 | "Can't open file %s for reading\n", filename); | ||
181 | return NULL; | ||
182 | } | ||
183 | if (GNUNET_OK != GNUNET_DISK_file_handle_size (f, &size)) | ||
184 | { | ||
185 | GNUNET_log (GNUNET_ERROR_TYPE_ERROR, | ||
186 | "Can't get size of file %s\n", filename); | ||
187 | return NULL; | ||
188 | } | ||
189 | GNUNET_log (GNUNET_ERROR_TYPE_DEBUG, | ||
190 | "using file %s, size %llu\n", | ||
191 | filename, (unsigned long long) size); | ||
192 | |||
193 | buffer = GNUNET_malloc (size + 1); | ||
194 | GNUNET_DISK_file_read (f, buffer, size); | ||
195 | GNUNET_DISK_file_close (f); | ||
196 | regexes = GNUNET_malloc (sizeof (char *)); | ||
197 | nr = 1; | ||
198 | offset = 0; | ||
199 | regex = NULL; | ||
200 | do | ||
201 | { | ||
202 | if (NULL == regex) | ||
203 | regex = GNUNET_malloc (size + 1); | ||
204 | len = (size_t) sscanf (&buffer[offset], "%s", regex); | ||
205 | if (0 == len) | ||
206 | break; | ||
207 | len = strlen (regex); | ||
208 | offset += len + 1; | ||
209 | if (len < 1) | ||
210 | continue; | ||
211 | if (len < 6 || strncmp (®ex[len - 6], "(0|1)*", 6) != 0) | ||
212 | { | ||
213 | GNUNET_log (GNUNET_ERROR_TYPE_WARNING, | ||
214 | "%s (line %u) does not end in \"(0|1)*\"\n", | ||
215 | buffer, nr); | ||
216 | } | ||
217 | else | ||
218 | { | ||
219 | len -= 6; | ||
220 | buffer[len] = '\0'; | ||
221 | } | ||
222 | regex = GNUNET_realloc (regex, len + 1); | ||
223 | GNUNET_array_grow (regexes, nr, nr + 1); | ||
224 | regexes[nr - 2] = regex; | ||
225 | regexes[nr - 1] = NULL; | ||
226 | } while (offset < size); | ||
227 | GNUNET_free (buffer); | ||
228 | |||
229 | return regexes; | ||
230 | } | ||
231 | |||
232 | |||
233 | /** | ||
234 | * Free all memory reserved for a set of regexes created by read_from_file. | ||
235 | * | ||
236 | * @param regexes NULL-terminated array of regexes. | ||
237 | */ | ||
238 | void | ||
239 | GNUNET_REGEX_free_from_file (char **regexes) | ||
240 | { | ||
241 | unsigned int i; | ||
242 | |||
243 | for (i = 0; regexes[i]; i++) | ||
244 | GNUNET_free (regexes[i]); | ||
245 | GNUNET_free (regexes); | ||
246 | } | ||
247 | |||
248 | /* end of regex_test_lib.c */ \ No newline at end of file | ||