#define MAX_NB_SEARCH_CONDITIONS 10
#define BUFFER_SIZE 65536
+#define TOKEN_BUFFER_SIZE 1024
regex_t leading_from_line_regexp;
-char *db_filename;
-char *db_filename_regexp_string;
-char *db_root_path;
-char *db_filename_list;
-char output_filename[PATH_MAX + 1];
+/* Global variables! */
int paranoid;
-int action_index;
int quiet;
-
-time_t being_today;
+char *default_search_field;
/********************************************************************/
return (a && !b) || (!a && b);
}
-char *segment_next_field(char *current) {
- while(*current && *current != ' ') current++;
- *current = '\0'; current++;
- while(*current && *current == ' ') current++;
- return current;
+char *parse_token(char *token_buffer, size_t token_buffer_size,
+ char separator, char *string) {
+ char *u = token_buffer;
+ while(u < token_buffer + token_buffer_size - 1 && *string &&
+ *string != separator) {
+ *(u++) = *(string++);
+ }
+ while(*string == separator) string++;
+ *u = '\0';
+ return string;
}
void remove_eof(char *c) {
fprintf(out, " index mails\n");
fprintf(out, " -o <output filename>, --output <output filename>\n");
fprintf(out, " set the result file, use stdout if unset\n");
+ fprintf(out, " -a <search field>, --default-search <search field>\n");
+ fprintf(out, " set the default search field\n");
+}
+
+/*********************************************************************/
+
+time_t time_for_past_day(int day) {
+ time_t t;
+ struct tm *tm;
+ int delta_day;
+ t = time(0);
+ tm = localtime(&t);
+ delta_day = (7 + tm->tm_wday - day) % 7 + 1;
+ return t - delta_day * 3600 * 24 + tm->tm_sec + 60 * tm->tm_min + 3600 * tm->tm_hour;
}
/*********************************************************************/
(name[0] == '.' && name[1] != '/');
}
-int is_a_leading_from_line(char *s) {
- return strncmp(s, "From ", 5) == 0 &&
- regexec(&leading_from_line_regexp, s, 0, 0, 0) == 0;
+int is_a_leading_from_line(int last_mbox_line_was_empty, char *mbox_line) {
+ return
+
+ /*
+
+ The mbox man page in qmail documentation states:
+
+ > The reader should not attempt to take advantage of the fact
+ > that every From_ line (past the beginning of the file) is
+ > preceded by a blank line.
+
+ */
+
+ /* last_mbox_line_was_empty && */
+ strncmp(mbox_line, "From ", 5) == 0 &&
+ regexec(&leading_from_line_regexp, mbox_line, 0, 0, 0) == 0;
}
int mbox_line_match_search(struct search_condition *condition,
}
}
-void search_in_db(FILE *db_file,
+void update_body_hits(char *mail_filename, int position_in_mail,
+ int nb_search_conditions, struct search_condition *search_conditions,
+ int nb_body_conditions,
+ int *hits) {
+ FILE *mail_file;
+ int header, n;
+ int last_mbox_line_was_empty;
+ char raw_mbox_line[BUFFER_SIZE];
+ int nb_body_hits;
+
+ nb_body_hits = 0;
+
+ header = 1;
+ mail_file = fopen(mail_filename, "r");
+
+ if(!mail_file) {
+ fprintf(stderr,
+ "mymail: Cannot open mbox '%s' for body scan.\n",
+ mail_filename);
+ exit(EXIT_FAILURE);
+ }
+
+ fseek(mail_file, position_in_mail, SEEK_SET);
+
+ if(fgets(raw_mbox_line, BUFFER_SIZE, mail_file)) {
+ while(nb_body_hits < nb_body_conditions) {
+ last_mbox_line_was_empty = (raw_mbox_line[0] == '\n');
+
+ if(last_mbox_line_was_empty) { header = 0; }
+
+ if(!header) {
+ for(n = 0; n < nb_search_conditions; n++) {
+ if(search_conditions[n].field_id == ID_BODY && !hits[n]) {
+ hits[n] =
+ (regexec(&search_conditions[n].regexp, raw_mbox_line, 0, 0, 0) == 0);
+ if(hits[n]) {
+ nb_body_hits++;
+ }
+ }
+ }
+ }
+
+ if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) ||
+ (is_a_leading_from_line(last_mbox_line_was_empty, raw_mbox_line)))
+ break;
+ }
+ }
+
+ fclose(mail_file);
+}
+
+void search_in_db(const char *db_filename,
int nb_search_conditions,
struct search_condition *search_conditions,
FILE *output_file) {
char raw_mbox_line[BUFFER_SIZE];
char current_mail_filename[PATH_MAX + 1];
unsigned long int current_position_in_mail;
- char *mbox_name, *mbox_value;
+ char mbox_name[TOKEN_BUFFER_SIZE], *mbox_value;
int mbox_id;
int already_written, m, n;
int last_mbox_line_was_empty;
int nb_body_conditions, nb_fulfilled_body_conditions;
+ FILE *db_file;
+
+ if(!quiet) {
+ printf("Searching in '%s' ... ", db_filename);
+ fflush(stdout);
+ }
+
+ db_file = fopen(db_filename, "r");
+
+ if(!db_file) {
+ fprintf(stderr,
+ "mymail: Cannot open \"%s\" for reading: %s\n",
+ db_filename,
+ strerror(errno));
+ exit(EXIT_FAILURE);
+ }
+
+ if(fgets(raw_db_line, BUFFER_SIZE, db_file)) {
+ if(strncmp(raw_db_line, MYMAIL_DB_MAGIC_TOKEN, strlen(MYMAIL_DB_MAGIC_TOKEN))) {
+ fprintf(stderr,
+ "mymail: Header line in '%s' does not match the mymail db format.\n",
+ db_filename);
+ exit(EXIT_FAILURE);
+ }
+ } else {
+ fprintf(stderr,
+ "mymail: Cannot read the header line in '%s'.\n",
+ db_filename);
+ exit(EXIT_FAILURE);
+ }
current_position_in_mail = 0;
already_written = 0;
strcpy(current_mail_filename, "");
while(fgets(raw_db_line, BUFFER_SIZE, db_file)) {
- mbox_name = raw_db_line;
- mbox_value = segment_next_field(raw_db_line);
+ mbox_value = parse_token(mbox_name, TOKEN_BUFFER_SIZE, ' ', raw_db_line);
if(strcmp("mail", mbox_name) == 0) {
- char *position_in_file_string;
- char *mail_filename;
+ char position_in_file_string[TOKEN_BUFFER_SIZE];
if(current_mail_filename[0]) {
/* all conditions but the body ones are fine, check the body
ones */
- nb_fulfilled_body_conditions = 0;
-
if(nb_body_conditions > 0) {
- FILE *mail_file;
- int header;
-
- header = 1;
- mail_file = fopen(current_mail_filename, "r");
-
- if(!mail_file) {
- fprintf(stderr,
- "mymail: Cannot open mbox '%s' for body scan.\n",
- current_mail_filename);
- exit(EXIT_FAILURE);
- }
-
- fseek(mail_file, current_position_in_mail, SEEK_SET);
-
- if(fgets(raw_mbox_line, BUFFER_SIZE, mail_file)) {
- while(nb_fulfilled_body_conditions < nb_body_conditions) {
- last_mbox_line_was_empty = (raw_mbox_line[0] == '\n');
-
- if(last_mbox_line_was_empty) { header = 0; }
+ update_body_hits(current_mail_filename, current_position_in_mail,
+ nb_search_conditions, search_conditions,
+ nb_body_conditions,
+ hits);
+ }
- if(!header) {
- for(n = 0; n < nb_search_conditions; n++) {
- if(search_conditions[n].field_id == ID_BODY && !hits[n]) {
- hits[n] =
- (regexec(&search_conditions[n].regexp, raw_mbox_line, 0, 0, 0) == 0);
- if(hits[n]) {
- nb_fulfilled_body_conditions++;
- }
- }
- }
- }
+ nb_fulfilled_body_conditions = 0;
- if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) ||
- (last_mbox_line_was_empty &&
- is_a_leading_from_line(raw_mbox_line)))
- break;
- }
+ for(n = 0; n < nb_search_conditions; n++) {
+ if(search_conditions[n].field_id == ID_BODY &&
+ xor(hits[n], search_conditions[n].negation)) {
+ nb_fulfilled_body_conditions++;
}
-
- fclose(mail_file);
}
if(nb_body_conditions == nb_fulfilled_body_conditions) {
fseek(mail_file, current_position_in_mail, SEEK_SET);
if(fgets(raw_mbox_line, BUFFER_SIZE, mail_file)) {
- last_mbox_line_was_empty = 1;
+ last_mbox_line_was_empty = 0;
fprintf(output_file, "%s", raw_mbox_line);
while(1) {
if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) ||
- (last_mbox_line_was_empty &&
- is_a_leading_from_line(raw_mbox_line))
+ (is_a_leading_from_line(last_mbox_line_was_empty, raw_mbox_line))
)
break;
last_mbox_line_was_empty = (raw_mbox_line[0] == '\n');
for(n = 0; n < nb_search_conditions; n++) { hits[n] = 0; }
- position_in_file_string = mbox_value;
- mail_filename = segment_next_field(mbox_value);
+ mbox_value = parse_token(position_in_file_string, TOKEN_BUFFER_SIZE, ' ', mbox_value);
+ mbox_value = parse_token(current_mail_filename, TOKEN_BUFFER_SIZE, ' ', mbox_value);
current_position_in_mail = atol(position_in_file_string);
- strcpy(current_mail_filename, mail_filename);
-
remove_eof(current_mail_filename);
already_written = 0;
}
}
}
}
+
+ fclose(db_file);
+
+ if(!quiet) {
+ printf("done.\n");
+ fflush(stdout);
+ }
}
void recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp,
DIR *dir;
struct dirent *dir_e;
struct stat sb;
- char raw_db_line[BUFFER_SIZE];
char subname[PATH_MAX + 1];
if(lstat(entry_name, &sb) != 0) {
while(*s) { if(*s == '/') { filename = s+1; } s++; }
if(regexec(db_filename_regexp, filename, 0, 0, 0) == 0) {
- FILE *db_file = fopen(entry_name, "r");
-
- if(!quiet) {
- printf("Searching in '%s' ... ", entry_name);
- fflush(stdout);
- }
-
- if(!db_file) {
- fprintf(stderr,
- "mymail: Cannot open \"%s\" for reading: %s\n",
- db_filename,
- strerror(errno));
- exit(EXIT_FAILURE);
- }
-
- if(fgets(raw_db_line, BUFFER_SIZE, db_file)) {
- if(strncmp(raw_db_line, MYMAIL_DB_MAGIC_TOKEN, strlen(MYMAIL_DB_MAGIC_TOKEN))) {
- fprintf(stderr,
- "mymail: Header line in '%s' does not match the mymail db format.\n",
- entry_name);
- exit(EXIT_FAILURE);
- }
- } else {
- fprintf(stderr,
- "mymail: Cannot read the header line in '%s'.\n",
- entry_name);
- exit(EXIT_FAILURE);
- }
-
- search_in_db(db_file, nb_search_conditions, search_conditions, output_file);
-
- fclose(db_file);
-
- if(!quiet) {
- printf("done.\n");
- fflush(stdout);
- }
+ search_in_db(entry_name, nb_search_conditions, search_conditions, output_file);
}
}
}
last_mbox_line_was_empty = 1;
while(fgets(raw_mbox_line, BUFFER_SIZE, file)) {
- if(last_mbox_line_was_empty &&
- is_a_leading_from_line(raw_mbox_line)) {
+ if(is_a_leading_from_line(last_mbox_line_was_empty, raw_mbox_line)) {
if(in_header) {
fprintf(stderr,
"Got a ^\"From \" in the header in %s:%lu.\n",
{ "search", 1, 0, 's' },
{ "index", 0, 0, 'i' },
{ "output", 1, 0, 'o' },
+ { "default-search", 1, 0, 'a' },
{ 0, 0, 0, 0 }
};
/*********************************************************************/
+void init_condition(struct search_condition *condition, char *full_string) {
+ char full_search_field[TOKEN_BUFFER_SIZE], *search_field;
+ int m;
+ char *string;
+
+ string = parse_token(full_search_field, TOKEN_BUFFER_SIZE, ' ', full_string);
+ search_field = full_search_field;
+
+ if(search_field[0] == '!') {
+ search_field++;
+ condition->negation = 1;
+ } else {
+ condition->negation = 0;
+ }
+
+ /* Recently */
+
+ if(strcmp(search_field, "8h") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time(0) - 3600 * 8;
+ condition->interval_stop = 0;
+ }
+
+ else if(strcmp(search_field, "week") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time(0) - 3600 * 24 * 7;
+ condition->interval_stop = 0;
+ }
+
+ else if(strcmp(search_field, "month") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time(0) - 3600 * 24 * 31;
+ condition->interval_stop = 0;
+ }
+
+ else if(strcmp(search_field, "24h") == 0 ||
+ strcmp(search_field, "today") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time(0) - 3600 * 24;
+ condition->interval_stop = 0;
+ }
+
+ /* Yesterday */
+
+ else if(strcmp(search_field, "yesterday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time(0) - 2 * 3600 * 24;
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ /* Week days */
+
+ else if(strcmp(search_field, "monday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(1);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "tuesday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(2);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "wednesday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(3);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "thursday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(4);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "friday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(5);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "saturday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(6);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else if(strcmp(search_field, "sunday") == 0) {
+ condition->field_id = ID_INTERVAL;
+ condition->interval_start = time_for_past_day(7);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+
+ else {
+
+ /* header-related conditions */
+
+ condition->field_id = -1;
+
+ for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) {
+ if(strncmp(field_names[m], search_field, strlen(search_field)) == 0) {
+ condition->field_id = m;
+ }
+ }
+
+ if(condition->field_id == -1) {
+ if(default_search_field) {
+ for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) {
+ if(strncmp(field_names[m],
+ default_search_field, strlen(default_search_field)) == 0) {
+ condition->field_id = m;
+ }
+ }
+ string = full_string;
+ }
+ }
+
+ if(condition->field_id == -1) {
+ fprintf(stderr,
+ "mymail: Syntax error in field name \"%s\".\n",
+ search_field);
+ exit(EXIT_FAILURE);
+ }
+
+ if(regcomp(&condition->regexp,
+ string,
+ REG_ICASE)) {
+ fprintf(stderr,
+ "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
+ string,
+ field_names[condition->field_id]);
+ exit(EXIT_FAILURE);
+ }
+ }
+}
+
+void free_condition(struct search_condition *condition) {
+ if(condition->field_id != ID_INTERVAL) {
+ regfree(&condition->regexp);
+ }
+}
+
+/*********************************************************************/
+/*********************************************************************/
+/*********************************************************************/
+
int main(int argc, char **argv) {
+ char *db_filename;
+ char *db_filename_regexp_string;
+ char *db_root_path;
+ char *db_filename_list;
+ char output_filename[PATH_MAX + 1];
+ int action_index;
int error = 0, show_help = 0;
const int nb_fields_to_parse = sizeof(fields_to_parse) / sizeof(struct parsable_field);
char c;
- int f;
+ int f, n;
int nb_search_conditions;
- char *search_condition_strings[MAX_NB_SEARCH_CONDITIONS];
FILE *output_file;
+ struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS];
if(regcomp(&leading_from_line_regexp,
- "^From [^ ]*@[^ ]* \\(Mon\\|Tue\\|Wed\\|Thu\\|Fri\\|Sat\\|Sun\\) \\(Jan\\|Feb\\|Mar\\|Apr\\|Jun\\|Jul\\|Aug\\|Sep\\|Oct\\|Nov\\|Dec\\) [ 123][0-9] [0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9][0-9][0-9][0-9]\n$",
+ "^From [^ ]* \\(Mon\\|Tue\\|Wed\\|Thu\\|Fri\\|Sat\\|Sun\\) \\(Jan\\|Feb\\|Mar\\|Apr\\|May\\|Jun\\|Jul\\|Aug\\|Sep\\|Oct\\|Nov\\|Dec\\) [ 123][0-9] [0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9][0-9][0-9][0-9]\n$",
0)) {
fprintf(stderr,
- "mymail: Cannot compile leading from line regexp. That is strange.\n");
+ "mymail: Cannot compile leading \"from\" line regexp. That is strange.\n");
exit(EXIT_FAILURE);
}
- /* printf("%d\n", regexec(&leading_from_line_regexp, "From root@idiap.ch Mon Apr 18 08:25:06 2011\n", 0, 0, 0)); */
- /* exit(EXIT_SUCCESS); */
-
paranoid = 0;
action_index = 0;
db_filename = 0;
+ db_filename_regexp_string = 0;
db_root_path = 0;
db_filename_list = 0;
quiet = 0;
+ default_search_field = 0;
setlocale(LC_ALL, "");
nb_search_conditions = 0;
- while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:",
+ while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:a:",
long_options, NULL)) != -1) {
switch(c) {
fprintf(stderr, "mymail: Too many search patterns.\n");
exit(EXIT_FAILURE);
}
- search_condition_strings[nb_search_conditions++] = strdup(optarg);
+ init_condition(&search_conditions[nb_search_conditions], optarg);
+ nb_search_conditions++;
+ break;
+
+ case 'a':
+ default_search_field = optarg;
break;
default:
else {
if(nb_search_conditions > 0) {
- struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS];
- char *search_field, *search_regexp_string;
- int m, n;
-
- for(n = 0; n < nb_search_conditions; n++) {
- search_field = search_condition_strings[n];
-
- if(search_field[0] == '!') {
- search_field++;
- search_conditions[n].negation = 1;
- } else {
- search_conditions[n].negation = 0;
- }
-
- if(strcmp(search_field, "today") == 0) {
- search_conditions[n].field_id = ID_INTERVAL;
- search_conditions[n].interval_start = time(0) - 3600 * 24;
- search_conditions[n].interval_stop = 0;
- }
-
- else if(strcmp(search_field, "yesterday") == 0) {
- search_conditions[n].field_id = ID_INTERVAL;
- search_conditions[n].interval_start = time(0) - 2 * 3600 * 24;
- search_conditions[n].interval_stop = time(0) - 3600 * 24;
- }
-
- else {
- search_regexp_string = segment_next_field(search_condition_strings[n]);
-
- search_conditions[n].field_id = -1;
-
- for(m = 0; (m < MAX_ID) && search_conditions[n].field_id == -1; m++) {
- if(strncmp(field_names[m], search_field, strlen(search_field)) == 0) {
- search_conditions[n].field_id = m;
- }
- }
-
- if(search_conditions[n].field_id == -1) {
- fprintf(stderr,
- "mymail: Syntax error in field name \"%s\".\n",
- search_field);
- exit(EXIT_FAILURE);
- }
-
- if(regcomp(&search_conditions[n].regexp,
- search_regexp_string,
- REG_ICASE)) {
- fprintf(stderr,
- "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
- search_regexp_string,
- field_names[search_conditions[n].field_id]);
- exit(EXIT_FAILURE);
- }
- }
- }
/* Recursive search if db_root_path is set */
if(db_filename_list) {
char db_filename[PATH_MAX + 1];
- char *s, *t;
- FILE *db_file;
+ char *s;
s = db_filename_list;
while(*s) {
- t = db_filename;
- while(*s == ';') { s++; }
- while(*s && *s != ';') { *t++ = *s++; }
- *t++ = '\0';
+ s = parse_token(db_filename, PATH_MAX + 1, ';', s);
if(db_filename[0]) {
- db_file = fopen(db_filename, "r");
-
- if(!db_file) {
- fprintf(stderr,
- "mymail: Cannot open \"%s\" for reading: %s\n",
- argv[optind],
- strerror(errno));
- exit(EXIT_FAILURE);
- }
-
- search_in_db(db_file, nb_search_conditions, search_conditions, output_file);
-
- fclose(db_file);
+ search_in_db(db_filename, nb_search_conditions, search_conditions, output_file);
}
}
}
/* Search in all db files listed in the command arguments */
while(optind < argc) {
- FILE *db_file = fopen(argv[optind], "r");
-
- if(!db_file) {
- fprintf(stderr,
- "mymail: Cannot open \"%s\" for reading: %s\n",
- argv[optind],
- strerror(errno));
- exit(EXIT_FAILURE);
- }
-
- search_in_db(db_file, nb_search_conditions, search_conditions, output_file);
-
- fclose(db_file);
+ search_in_db(argv[optind], nb_search_conditions, search_conditions, output_file);
optind++;
}
-
- for(n = 0; n < nb_search_conditions; n++) {
- regfree(&search_conditions[n].regexp);
- free(search_condition_strings[n]);
- }
}
}
+ for(n = 0; n < nb_search_conditions; n++) {
+ free_condition(&search_conditions[n]);
+ }
+
if(output_file != stdout) {
fflush(output_file);
fclose(output_file);
projects / mymail.git / blobdiff