3 * Copyright (c) 2013 Francois Fleuret
4 * Written by Francois Fleuret <francois@fleuret.org>
6 * This file is part of mymail.
8 * mymail is free software: you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License version 3 as
10 * published by the Free Software Foundation.
12 * mymail is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with mymail. If not, see <http://www.gnu.org/licenses/>.
24 This command is a dumb mail indexer. It can either (1) scan
25 directories containing mbox files, and create a db file containing
26 for each mail a list of fields computed from the header, or (2)
27 read such a db file and get all the mails matching regexp-defined
28 conditions on the fields.
30 It is low-tech, simple, light and fast.
49 #define BUFFER_SIZE 16384
51 struct parsable_field {
63 char *segment_next_field(char *current) {
64 while(*current && *current != ' ') current++;
65 *current = '\0'; current++;
66 while(*current && *current == ' ') current++;
70 void remove_eof(char *c) {
71 while(*c && *c != '\n' && *c != '\r') c++;
75 /********************************************************************/
77 /* malloc with error checking. */
79 void *safe_malloc(size_t n) {
83 "mymail: can not allocate memory: %s\n", strerror(errno));
89 /*********************************************************************/
91 void usage(FILE *out) {
92 fprintf(out, "mymail version %s (%s)\n", VERSION, UNAME);
93 fprintf(out, "Written by Francois Fleuret <francois@fleuret.org>.\n");
95 fprintf(out, "Usage: mymail [options] [<filename1> [<filename2> ...]]\n");
99 /*********************************************************************/
101 void search_in_db(const char *search_name, const char *search_regexp_string,
103 char raw_line[BUFFER_SIZE];
104 char current_mail_filename[BUFFER_SIZE];
105 unsigned long int current_position_in_mail;
111 search_regexp_string,
114 "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
115 search_regexp_string,
120 current_position_in_mail = 0;
123 while(fgets(raw_line, BUFFER_SIZE, db_file)) {
125 value = segment_next_field(raw_line);
127 if(strcmp("mail", name) == 0) {
128 char *position_in_file_string = value;
129 char *mail_filename = segment_next_field(value);
130 current_position_in_mail = atol(position_in_file_string);
131 strcpy(current_mail_filename, mail_filename);
132 remove_eof(current_mail_filename);
136 else if(!already_written) {
137 if(strcmp(search_name, name) == 0 && regexec(®exp, value, 0, 0, 0) == 0) {
139 mail_file = fopen(current_mail_filename, "r");
141 fprintf(stderr, "mymail: Can not open `%s'.\n", current_mail_filename);
144 fseek(mail_file, current_position_in_mail, SEEK_SET);
145 if(fgets(raw_line, BUFFER_SIZE, mail_file)) {
146 printf("%s", raw_line);
147 while(fgets(raw_line, BUFFER_SIZE, mail_file) &&
148 strncmp(raw_line, "From ", 5)) {
149 printf("%s", raw_line);
161 /*********************************************************************/
163 void read_file(const char *input_filename,
164 int nb_fields_to_parse, struct parsable_field *fields_to_parse,
166 char raw_line[BUFFER_SIZE];
168 int in_header, new_header;
169 unsigned long int position_in_file;
171 file = fopen(input_filename, "r");
174 fprintf(stderr, "mymail: Can not open `%s'.\n", input_filename);
175 if(paranoid) { exit(EXIT_FAILURE); }
182 position_in_file = 0;
184 while(fgets(raw_line, BUFFER_SIZE, file)) {
185 if(strncmp(raw_line, "From ", 5) == 0) {
188 "Got a ^\"From \" in the header in %s:%lu.\n",
189 input_filename, position_in_file);
190 fprintf(stderr, "%s", raw_line);
191 if(paranoid) { exit(EXIT_FAILURE); }
195 } else if(strncmp(raw_line, "\n", 1) == 0) {
196 if(in_header) { in_header = 0; }
203 fprintf(db_file, "mail %lu %s\n", position_in_file, input_filename);
206 for(f = 0; f < nb_fields_to_parse; f++) {
207 if(regexec(&fields_to_parse[f].regexp, raw_line, 1, &matches, 0) == 0) {
208 fprintf(db_file, "%s %s",
209 fields_to_parse[f].name,
210 raw_line + matches.rm_eo);
215 position_in_file += strlen(raw_line);
221 int ignore_entry(const char *name) {
223 /* strcmp(name, ".") == 0 || */
224 /* strcmp(name, "..") == 0 || */
225 (name[0] == '.' && name[1] != '/');
228 void process_entry(const char *dir_name,
229 int nb_fields_to_parse, struct parsable_field *fields_to_parse,
232 struct dirent *dir_e;
234 char subname[PATH_MAX + 1];
236 if(lstat(dir_name, &sb) != 0) {
238 "mymail: Can not stat \"%s\": %s\n",
244 dir = opendir(dir_name);
247 printf("Processing directory '%s'.\n", dir_name);
248 while((dir_e = readdir(dir))) {
249 if(!ignore_entry(dir_e->d_name)) {
250 snprintf(subname, PATH_MAX, "%s/%s", dir_name, dir_e->d_name);
251 process_entry(subname, nb_fields_to_parse, fields_to_parse, db_file);
256 read_file(dir_name, nb_fields_to_parse, fields_to_parse, db_file);
260 /*********************************************************************/
262 /* For long options that have no equivalent short option, use a
263 non-character as a pseudo short option, starting with CHAR_MAX + 1. */
266 OPT_BASH_MODE = CHAR_MAX + 1
269 static struct option long_options[] = {
270 { "help", no_argument, 0, 'h' },
271 { "db-prefix", 1, 0, 'p' },
272 { "search-pattern", 1, 0, 's' },
273 { "index", 0, 0, 'i' },
277 static struct parsable_field fields_to_parse[] = {
280 "^\\([Ff][Rr][Oo][Mm]:\\|From\\) *",
281 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
286 "^\\([Tt][Oo]\\|[Cc][Cc]\\|[Bb][Cc][Cc]\\): *",
287 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
291 int main(int argc, char **argv) {
292 int error = 0, show_help = 0;
293 const int nb_fields_to_parse = sizeof(fields_to_parse) / sizeof(struct parsable_field);
301 setlocale(LC_ALL, "");
303 while ((c = getopt_long(argc, argv, "hip:s:",
304 long_options, NULL)) != -1) {
317 db_filename = strdup(optarg);
322 fprintf(stderr, "mymail: Search pattern already defined.\n");
325 search_pattern = strdup(optarg);
335 db_filename = strdup("/tmp/mymail");
349 FILE *db_file = fopen(db_filename, "w");
352 "mymail: Can not open \"%s\" for writing: %s\n",
358 for(f = 0; f < nb_fields_to_parse; f++) {
359 if(regcomp(&fields_to_parse[f].regexp,
360 fields_to_parse[f].regexp_string,
363 "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
364 fields_to_parse[f].regexp_string,
365 fields_to_parse[f].name);
370 while(optind < argc) {
371 process_entry(argv[optind],
372 nb_fields_to_parse, fields_to_parse, db_file);
378 for(f = 0; f < nb_fields_to_parse; f++) {
379 regfree(&fields_to_parse[f].regexp);
387 char *search_regexp_string;
388 search_name = search_pattern;
389 search_regexp_string = segment_next_field(search_pattern);
390 if(!*search_regexp_string) {
392 "Syntax error in the search pattern.\n");
396 /* printf("Starting search in %s for field \"%s\" matching \"%s\".\n", */
399 /* search_regexp_string); */
401 db_file = fopen(db_filename, "r");
405 "mymail: Can not open \"%s\" for reading: %s\n",
411 search_in_db(search_name, search_regexp_string, db_file);
414 free(search_pattern);