#include <time.h>
#define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file"
-#define VERSION "0.9.5"
+#define VERSION "0.9.6"
#define MAX_NB_SEARCH_CONDITIONS 32
/* Global variables! */
-int paranoid;
-int quiet;
-int ignore_dot_files;
+int global_quiet;
+int global_use_leading_time;
-regex_t leading_from_line_regexp;
+regex_t global_leading_from_line_regexp;
/********************************************************************/
ID_DATE,
ID_PARTICIPANT,
ID_BODY,
- ID_INTERVAL,
+ ID_TIME_INTERVAL,
MAX_ID
};
-static char *field_names[] = {
+static char *field_keys[] = {
"mail",
"lead",
"from",
fprintf(out, " print the version number\n");
fprintf(out, " -q, --quiet\n");
fprintf(out, " do not print information during search\n");
+ fprintf(out, " -t, --use-leading-time\n");
+ fprintf(out, " use the time stamp from the leading line of each mail and not the Date:\n");
+ fprintf(out, " field\n");
fprintf(out, " -p <db filename pattern>, --db-pattern <db filename pattern>\n");
fprintf(out, " set the db filename pattern for recursive search\n");
fprintf(out, " -r <db root path>, --db-root <db root path>\n");
return
strcmp(name, ".") == 0 ||
strcmp(name, "..") == 0 ||
- (ignore_dot_files && name[0] == '.' && name[1] != '/');
+ (name[0] == '.' && name[1] != '/');
}
int is_a_leading_from_line(char *mbox_line) {
return
strncmp(mbox_line, "From ", 5) == 0 &&
- regexec(&leading_from_line_regexp, mbox_line, 0, 0, 0) == 0;
+ regexec(&global_leading_from_line_regexp, mbox_line, 0, 0, 0) == 0;
}
int mbox_line_match_search(struct search_condition *condition,
int mbox_id, const char *mbox_value) {
- if(condition->field_id == ID_INTERVAL) {
- if(mbox_id == ID_LEADING_LINE) {
- const char *c;
- time_t t;
- struct tm tm;
-
- c = mbox_value;
- while(*c && *c != ' ') c++; while(*c && *c == ' ') c++;
- strptime(c, "%a %b %e %k:%M:%S %Y", &tm);
- t = mktime(&tm);
-
- return (t >= condition->interval_start &&
- (condition->interval_stop == 0 ||
- t <= condition->interval_stop));
- } else {
- return 0;
- }
- } else {
- return
- (
-
- (condition->field_id == mbox_id)
+ return
+ (
+ (condition->field_id == mbox_id)
- ||
+ ||
- (condition->field_id == ID_PARTICIPANT && (mbox_id == ID_LEADING_LINE ||
- mbox_id == ID_FROM ||
- mbox_id == ID_TO))
- ||
+ (condition->field_id == ID_PARTICIPANT && (mbox_id == ID_LEADING_LINE ||
+ mbox_id == ID_FROM ||
+ mbox_id == ID_TO))
+ ||
- (condition->field_id == ID_FROM && mbox_id == ID_LEADING_LINE)
+ (condition->field_id == ID_FROM && mbox_id == ID_LEADING_LINE)
- )
+ )
- &&
+ &&
- regexec(&condition->regexp, mbox_value, 0, 0, 0) == 0;
- }
+ regexec(&condition->regexp, mbox_value, 0, 0, 0) == 0;
}
void update_body_hits(char *mail_filename, int position_in_mail,
fclose(mail_file);
}
+int check_full_mail_match(char *current_mail_filename,
+ time_t mail_time,
+ int nb_search_conditions,
+ struct search_condition *search_conditions,
+ int nb_body_conditions,
+ int *hits,
+ int current_position_in_mail) {
+ int n, nb_fulfilled_body_conditions;
+
+ for(n = 0; n < nb_search_conditions; n++) {
+ if(search_conditions[n].field_id == ID_TIME_INTERVAL) {
+ hits[n] = (mail_time >= search_conditions[n].interval_start &&
+ (search_conditions[n].interval_stop == 0 ||
+ mail_time <= search_conditions[n].interval_stop));
+ }
+ }
+
+ /* We first check all conditions but the body ones */
+
+ for(n = 0; n < nb_search_conditions &&
+ ((search_conditions[n].field_id == ID_BODY) ||
+ xor(hits[n], search_conditions[n].negation)); n++);
+
+ if(n == nb_search_conditions) {
+
+ /* Now check the body ones */
+
+ nb_fulfilled_body_conditions = 0;
+
+ if(nb_body_conditions > 0) {
+ update_body_hits(current_mail_filename, current_position_in_mail,
+ nb_search_conditions, search_conditions,
+ nb_body_conditions,
+ hits);
+
+ 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++;
+ }
+ }
+ }
+ return nb_body_conditions == nb_fulfilled_body_conditions;
+ } else {
+ return 0;
+ }
+}
+
+/* We use the mail leading line time by default, and if we should and
+ can, we update with the Date: field */
+
+void update_time(int mbox_id, const char *mbox_value, time_t *t) {
+ const char *c;
+ struct tm tm;
+
+ if(mbox_id == ID_LEADING_LINE) {
+ c = mbox_value;
+ while(*c && *c != ' ') c++; while(*c && *c == ' ') c++;
+ strptime(c, "%a %b %e %k:%M:%S %Y", &tm);
+ *t = mktime(&tm);
+ } else {
+ if(!global_use_leading_time) {
+ if(mbox_id == ID_DATE) {
+ if(strptime(mbox_value, "%a, %d %b %Y %k:%M:%S", &tm) ||
+ strptime(mbox_value, "%d %b %Y %k:%M:%S", &tm)) {
+ *t = mktime(&tm);
+ }
+ }
+ }
+ }
+}
+
int search_in_db(const char *db_filename,
int nb_search_conditions,
struct search_condition *search_conditions,
char raw_db_line[BUFFER_SIZE];
char current_mail_filename[PATH_MAX + 1];
unsigned long int current_position_in_mail;
- char mbox_name[TOKEN_BUFFER_SIZE];
+ char mbox_key[TOKEN_BUFFER_SIZE];
+ char position_in_file_string[TOKEN_BUFFER_SIZE];
const char *mbox_value;
int mbox_id;
int already_written, m, n;
- int nb_body_conditions, nb_fulfilled_body_conditions;
+ int nb_body_conditions, need_time;
FILE *db_file;
int nb_extracted_mails;
+ time_t mail_time;
nb_extracted_mails = 0;
- if(!quiet) {
+ if(!global_quiet) {
printf("Searching in '%s' ... ", db_filename);
fflush(stdout);
}
for(n = 0; n < nb_search_conditions; n++) { hits[n] = 0; }
nb_body_conditions = 0;
+ need_time = 0;
+ mail_time = 0;
+
for(n = 0; n < nb_search_conditions; n++) {
if(search_conditions[n].field_id == ID_BODY) {
nb_body_conditions++;
}
+ else if(search_conditions[n].field_id == ID_TIME_INTERVAL) {
+ need_time = 1;
+ }
}
strcpy(current_mail_filename, "");
while(fgets(raw_db_line, BUFFER_SIZE, db_file)) {
- mbox_value = parse_token(mbox_name, TOKEN_BUFFER_SIZE, ' ', raw_db_line);
-
- if(strcmp("mail", mbox_name) == 0) {
- char position_in_file_string[TOKEN_BUFFER_SIZE];
+ mbox_value = parse_token(mbox_key, TOKEN_BUFFER_SIZE, ' ', raw_db_line);
+ if(strcmp("mail", mbox_key) == 0) {
if(current_mail_filename[0]) {
-
- /* We first check all conditions but the body ones */
-
- for(n = 0; n < nb_search_conditions &&
- ((search_conditions[n].field_id == ID_BODY) ||
- xor(hits[n], search_conditions[n].negation)); n++);
-
- if(n == nb_search_conditions) {
-
- /* Now check the body ones */
-
- nb_fulfilled_body_conditions = 0;
-
- if(nb_body_conditions > 0) {
- update_body_hits(current_mail_filename, current_position_in_mail,
- nb_search_conditions, search_conditions,
- nb_body_conditions,
- hits);
-
- 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++;
- }
- }
- }
-
- if(nb_body_conditions == nb_fulfilled_body_conditions) {
- nb_extracted_mails++;
- extract_mail(current_mail_filename, current_position_in_mail, output_file);
- }
+ if(check_full_mail_match(current_mail_filename,
+ mail_time,
+ nb_search_conditions, search_conditions,
+ nb_body_conditions, hits, current_position_in_mail)) {
+ extract_mail(current_mail_filename, current_position_in_mail, output_file);
+ nb_extracted_mails++;
}
}
for(n = 0; n < nb_search_conditions; n++) { hits[n] = 0; }
-
mbox_value = parse_token(position_in_file_string, TOKEN_BUFFER_SIZE, ' ', mbox_value);
- mbox_value = parse_token(current_mail_filename, TOKEN_BUFFER_SIZE, '\n', mbox_value);
+ mbox_value = parse_token(current_mail_filename, PATH_MAX+1, '\n', mbox_value);
current_position_in_mail = atol(position_in_file_string);
already_written = 0;
}
else {
mbox_id = -1;
for(m = 0; (m < MAX_ID) && mbox_id == -1; m++) {
- if(strncmp(field_names[m], mbox_name, strlen(mbox_name)) == 0) {
+ if(strncmp(field_keys[m], mbox_key, strlen(mbox_key)) == 0) {
mbox_id = m;
}
}
+
for(n = 0; n < nb_search_conditions; n++) {
hits[n] |= mbox_line_match_search(&search_conditions[n],
mbox_id, mbox_value);
}
+
+ if(need_time) {
+ update_time(mbox_id, mbox_value, &mail_time);
+ }
+ }
+ }
+
+ if(current_mail_filename[0]) {
+ if(check_full_mail_match(current_mail_filename,
+ mail_time,
+ nb_search_conditions, search_conditions,
+ nb_body_conditions, hits, current_position_in_mail)) {
+ extract_mail(current_mail_filename, current_position_in_mail, output_file);
+ nb_extracted_mails++;
}
}
fclose(db_file);
- if(!quiet) {
+ if(!global_quiet) {
printf("done.\n");
fflush(stdout);
}
for(f = 0; f < nb_fields_to_parse; f++) {
if(regexec(&fields_to_parse[f].regexp, raw_mbox_line, 1, &matches, 0) == 0) {
fprintf(db_file, "%s %s\n",
- field_names[fields_to_parse[f].id],
+ field_keys[fields_to_parse[f].id],
raw_mbox_line + matches.rm_eo);
}
}
while(fgets(raw_mbox_line, BUFFER_SIZE, file)) {
if(is_a_leading_from_line(raw_mbox_line)) {
+ /* This starts a new mail */
if(in_header) {
fprintf(stderr,
"Got a ^\"From \" in the header in %s:%lu.\n",
mbox_filename, position_in_file);
fprintf(stderr, "%s", raw_mbox_line);
- if(paranoid) { exit(EXIT_FAILURE); }
}
+
+ /* printf("LEADING_LINE %s", raw_mbox_line); */
+
in_header = 1;
new_header = 1;
} else if(raw_mbox_line[0] == '\n') {
- if(in_header) { in_header = 0; }
+ if(in_header) {
+ in_header = 0;
+ /* We leave the header, index the current line */
+ if(full_line[0]) {
+ /* printf("INDEX %s\n", full_line); */
+ index_one_mbox_line(nb_fields_to_parse, fields_to_parse, full_line, db_file);
+ }
+ end_of_full_line = full_line;
+ *end_of_full_line = '\0';
+ }
}
if(in_header) {
}
if(raw_mbox_line[0] == ' ' || raw_mbox_line[0] == '\t') {
+ /* Continuation of a line */
char *start = raw_mbox_line;
while(*start == ' ' || *start == '\t') start++;
*(end_of_full_line++) = ' ';
}
else {
- /*
- if(!((raw_mbox_line[0] >= 'a' && raw_mbox_line[0] <= 'z') ||
- (raw_mbox_line[0] >= 'A' && raw_mbox_line[0] <= 'Z'))) {
- fprintf(stderr,
- "Header line syntax error %s:%lu.\n",
- mbox_filename, position_in_file);
- fprintf(stderr, "%s", raw_mbox_line);
- }
- */
+ /* Start a new header line, not a continuation */
if(full_line[0]) {
+ /* printf("INDEX %s\n", full_line); */
index_one_mbox_line(nb_fields_to_parse, fields_to_parse, full_line, db_file);
}
{ "help", no_argument, 0, 'h' },
{ "version", no_argument, 0, 'v' },
{ "quiet", no_argument, 0, 'q' },
+ { "use-leading-time", no_argument, 0, 't' },
{ "db-file-generate", 1, 0, 'd' },
{ "db-pattern", 1, 0, 'p' },
{ "db-root", 1, 0, 'r' },
struct time_criterion {
char *label;
+ int day_criterion;
int start_hour, end_hour;
int past_week_day;
};
static struct time_criterion time_criteria[] = {
- { "8h", 8, -1, -1 },
- { "today", 24, -1, -1 },
- { "24h", 24, -1, -1 },
- { "48h", 48, -1, -1 },
- { "week", 24 * 7, -1, -1 },
- { "month", 24 * 31, -1, -1 },
- { "year", 24 * 365, -1, -1 },
+ { "8h", 0, 8, -1, -1 },
+ { "24h", 0, 24, -1, -1 },
+ { "48h", 0, 48, -1, -1 },
+ { "week", 0, 24 * 7, -1, -1 },
+ { "month", 0, 24 * 31, -1, -1 },
+ { "year", 0, 24 * 365, -1, -1 },
- { "yesterday", 48, 24, -1 },
+ { "yesterday", 1, -1, -1, -1 },
+ { "today", 1, -1, -1, 0 },
- { "monday", -1, -1, 1 },
- { "tuesday", -1, -1, 2 },
- { "wednesday", -1, -1, 3 },
- { "thursday", -1, -1, 4 },
- { "friday", -1, -1, 5 },
- { "saturday", -1, -1, 6 },
- { "sunday", -1, -1, 7 },
+ { "monday", 1, -1, -1, 1 },
+ { "tuesday", 1, -1, -1, 2 },
+ { "wednesday", 1, -1, -1, 3 },
+ { "thursday", 1, -1, -1, 4 },
+ { "friday", 1, -1, -1, 5 },
+ { "saturday", 1, -1, -1, 6 },
+ { "sunday", 1, -1, -1, 7 },
};
int delta_day;
t = time(0);
tm = localtime(&t);
- delta_day = (7 + tm->tm_wday - day) % 7;
- if(delta_day == 0) { delta_day = 7; }
+ if(day > 0) {
+ delta_day = (7 + tm->tm_wday - day) % 7;
+ } else {
+ delta_day = - day;
+ }
return t - (delta_day * 3600 * 24 + tm->tm_sec + 60 * tm->tm_min + 3600 * tm->tm_hour);
}
for(k = 0; k < sizeof(time_criteria) / sizeof(struct time_criterion); k++) {
if(strcmp(time_criteria[k].label, search_field) == 0) {
- condition->field_id = ID_INTERVAL;
- if(time_criteria[k].past_week_day < 0) {
+ condition->field_id = ID_TIME_INTERVAL;
+ if(time_criteria[k].day_criterion) {
+ condition->interval_start = time_for_past_day(time_criteria[k].past_week_day);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ } else {
condition->interval_start = time(0) - 3600 * time_criteria[k].start_hour;
if(time_criteria[k].end_hour >= 0) {
condition->interval_stop = time(0) - 3600 * time_criteria[k].end_hour;
} else {
condition->interval_stop = 0;
}
- } else {
- condition->interval_start = time_for_past_day(time_criteria[k].past_week_day);
- condition->interval_stop = condition->interval_start + 3600 * 24;
}
break;
}
/* No time condition matched, look for the search fields */
for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) {
- if(strncmp(field_names[m], search_field, strlen(search_field)) == 0) {
+ if(strncmp(field_keys[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],
+ if(strncmp(field_keys[m],
default_search_field, strlen(default_search_field)) == 0) {
condition->field_id = m;
}
}
string = full_string;
+ if(string[0] == '!') { string++; }
}
}
if(condition->field_id == -1) {
fprintf(stderr,
- "mymail: Syntax error in field name \"%s\".\n",
+ "mymail: Syntax error in field key \"%s\".\n",
search_field);
exit(EXIT_FAILURE);
}
fprintf(stderr,
"mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
string,
- field_names[condition->field_id]);
+ field_keys[condition->field_id]);
exit(EXIT_FAILURE);
}
}
}
void free_condition(struct search_condition *condition) {
- if(condition->field_id != ID_INTERVAL) {
+ if(condition->field_id != ID_TIME_INTERVAL) {
regfree(&condition->regexp);
}
}
unsigned int nb_search_conditions;
struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS];
- if(regcomp(&leading_from_line_regexp, LEADING_FROM_LINE_REGEXP_STRING, 0)) {
+ if(regcomp(&global_leading_from_line_regexp, LEADING_FROM_LINE_REGEXP_STRING, 0)) {
fprintf(stderr,
"mymail: Cannot compile leading \"from\" line regexp. That is strange.\n");
exit(EXIT_FAILURE);
}
- paranoid = 0;
- quiet = 0;
+ global_quiet = 0;
+ global_use_leading_time = 0;
default_search_field = 0;
- ignore_dot_files = 1;
strncpy(output_filename, "", PATH_MAX);
setlocale(LC_ALL, "");
break;
case 'q':
- quiet = 1;
+ global_quiet = 1;
+ break;
+
+ case 't':
+ global_use_leading_time = 1;
break;
case 'i':
fprintf(stderr,
"mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
fields_to_parse[f].regexp_string,
- field_names[fields_to_parse[f].id]);
+ field_keys[fields_to_parse[f].id]);
exit(EXIT_FAILURE);
}
}
output_file = safe_fopen(output_filename, "w", "result mbox");
} else {
output_file = stdout;
- quiet = 1;
+ global_quiet = 1;
}
if(nb_search_conditions > 0) {
}
}
- if(!quiet) {
+ if(!global_quiet) {
if(nb_extracted_mails > 0) {
printf("Found %d matching mails.\n", nb_extracted_mails);
} else {
free(db_filename_list);
free(mbox_filename_regexp_string);
- regfree(&leading_from_line_regexp);
+ regfree(&global_leading_from_line_regexp);
exit(EXIT_SUCCESS);
}
projects / mymail.git / blobdiff