X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=blobdiff_plain;f=mymail.c;h=b3f76d11182cc8448b7e8f45682f77d0f72b81e0;hb=26322035eb062c640bd5eef21ce1d706a55ee3cd;hp=e078bf6f7f603cc884bb7182198d952f7ed93108;hpb=cf24e85e7e8dad2d7e002982bca9124e0ae9832f;p=mymail.git diff --git a/mymail.c b/mymail.c index e078bf6..b3f76d1 100644 --- a/mymail.c +++ b/mymail.c @@ -46,7 +46,7 @@ #include #define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file" -#define VERSION "0.9.5" +#define VERSION "0.9.6" #define MAX_NB_SEARCH_CONDITIONS 32 @@ -73,7 +73,7 @@ enum { ID_DATE, ID_PARTICIPANT, ID_BODY, - ID_INTERVAL, + ID_TIME_INTERVAL, MAX_ID }; @@ -92,10 +92,10 @@ static char *field_keys[] = { /********************************************************************/ struct search_condition { - int field_id; + int db_key; + regex_t db_value_regexp; int negation; - regex_t regexp; - time_t interval_start, interval_stop; + time_t time_start, time_stop; }; /********************************************************************/ @@ -249,57 +249,27 @@ int is_a_leading_from_line(char *mbox_line) { 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) { - const char *c; - time_t t; - struct tm tm; - if(global_use_leading_time) { - 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); - return (t >= condition->interval_start && - (condition->interval_stop == 0 || - t <= condition->interval_stop)); - } else { - return 0; - } - } else { - if(mbox_id == ID_DATE) { - strptime(mbox_value, "%a, %d %b %Y %k:%M:%S", &tm); - t = mktime(&tm); - return (t >= condition->interval_start && - (condition->interval_stop == 0 || - t <= condition->interval_stop)); - } else { - return 0; - } - } - } else { - return - ( +int db_line_match_search(struct search_condition *condition, + int db_key, const char *db_value) { - (condition->field_id == mbox_id) + return + ( + (condition->db_key == db_key) - || + || - (condition->field_id == ID_PARTICIPANT && (mbox_id == ID_LEADING_LINE || - mbox_id == ID_FROM || - mbox_id == ID_TO)) - || + (condition->db_key == ID_PARTICIPANT && (db_key == ID_LEADING_LINE || + db_key == ID_FROM || + db_key == ID_TO)) + || - (condition->field_id == ID_FROM && mbox_id == ID_LEADING_LINE) + (condition->db_key == ID_FROM && db_key == ID_LEADING_LINE) - ) + ) - && + && - regexec(&condition->regexp, mbox_value, 0, 0, 0) == 0; - } + regexec(&condition->db_value_regexp, db_value, 0, 0, 0) == 0; } void update_body_hits(char *mail_filename, int position_in_mail, @@ -327,9 +297,9 @@ void update_body_hits(char *mail_filename, int position_in_mail, if(!header) { for(n = 0; n < nb_search_conditions; n++) { - if(search_conditions[n].field_id == ID_BODY && !hits[n]) { + if(search_conditions[n].db_key == ID_BODY && !hits[n]) { hits[n] = - (regexec(&search_conditions[n].regexp, raw_mbox_line, 0, 0, 0) == 0); + (regexec(&search_conditions[n].db_value_regexp, raw_mbox_line, 0, 0, 0) == 0); if(hits[n]) { nb_body_hits++; } @@ -347,7 +317,7 @@ void update_body_hits(char *mail_filename, int position_in_mail, } void extract_mail(const char *mail_filename, unsigned long int position_in_mail, - FILE *output_file) { + FILE *output_file) { char raw_mbox_line[BUFFER_SIZE]; FILE *mail_file; @@ -368,22 +338,97 @@ void extract_mail(const char *mail_filename, unsigned long 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].db_key == ID_TIME_INTERVAL) { + hits[n] = (mail_time >= search_conditions[n].time_start && + (search_conditions[n].time_stop == 0 || + mail_time <= search_conditions[n].time_stop)); + } + } + + /* We first check all conditions but the body ones */ + + for(n = 0; n < nb_search_conditions && + ((search_conditions[n].db_key == 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].db_key == 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 db_key, const char *db_value, time_t *t) { + const char *c; + struct tm tm; + + if(db_key == ID_LEADING_LINE) { + c = db_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(db_key == ID_DATE) { + if(strptime(db_value, "%a, %d %b %Y %k:%M:%S", &tm) || + strptime(db_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, FILE *output_file) { - int hits[MAX_NB_SEARCH_CONDITIONS]; + FILE *db_file; char raw_db_line[BUFFER_SIZE]; char current_mail_filename[PATH_MAX + 1]; + char db_key_string[TOKEN_BUFFER_SIZE]; + char position_in_file_string[TOKEN_BUFFER_SIZE]; unsigned long int current_position_in_mail; - char mbox_key[TOKEN_BUFFER_SIZE]; - const char *mbox_value; - int mbox_id; - int already_written, m, n; - int nb_body_conditions, nb_fulfilled_body_conditions; - FILE *db_file; + const char *db_value; + int db_key; + int hits[MAX_NB_SEARCH_CONDITIONS]; + int nb_body_conditions, need_time; int nb_extracted_mails; + time_t mail_time; + + int m, n; nb_extracted_mails = 0; @@ -413,82 +458,73 @@ int search_in_db(const char *db_filename, /* Then parse the said db file */ current_position_in_mail = 0; - already_written = 0; 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) { + if(search_conditions[n].db_key == ID_BODY) { nb_body_conditions++; } + else if(search_conditions[n].db_key == ID_TIME_INTERVAL) { + need_time = 1; + } } strcpy(current_mail_filename, ""); while(fgets(raw_db_line, BUFFER_SIZE, db_file)) { - mbox_value = parse_token(mbox_key, TOKEN_BUFFER_SIZE, ' ', raw_db_line); - - if(strcmp("mail", mbox_key) == 0) { - char position_in_file_string[TOKEN_BUFFER_SIZE]; + db_value = parse_token(db_key_string, TOKEN_BUFFER_SIZE, ' ', raw_db_line); + if(strcmp("mail", db_key_string) == 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); + db_value = parse_token(position_in_file_string, TOKEN_BUFFER_SIZE, ' ', db_value); + db_value = parse_token(current_mail_filename, PATH_MAX+1, '\n', db_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_keys[m], mbox_key, strlen(mbox_key)) == 0) { - mbox_id = m; + db_key = -1; + for(m = 0; (m < MAX_ID) && db_key == -1; m++) { + if(strncmp(field_keys[m], db_key_string, strlen(db_key_string)) == 0) { + db_key = m; } } + for(n = 0; n < nb_search_conditions; n++) { - hits[n] |= mbox_line_match_search(&search_conditions[n], - mbox_id, mbox_value); + hits[n] |= db_line_match_search(&search_conditions[n], + db_key, db_value); + } + + if(need_time) { + update_time(db_key, db_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(!global_quiet) { @@ -712,6 +748,7 @@ static struct option long_options[] = { struct time_criterion { char *label; + int day_criterion; int start_hour, end_hour; int past_week_day; }; @@ -720,23 +757,23 @@ struct time_criterion { 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 }, }; @@ -748,8 +785,11 @@ time_t time_for_past_day(int day) { 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); } @@ -769,46 +809,46 @@ void init_condition(struct search_condition *condition, const char *full_string, condition->negation = 0; } - condition->field_id = -1; + condition->db_key = -1; - /* Recently */ + /* Time condition */ 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->interval_start = time(0) - 3600 * time_criteria[k].start_hour; + condition->db_key = ID_TIME_INTERVAL; + if(time_criteria[k].day_criterion) { + condition->time_start = time_for_past_day(time_criteria[k].past_week_day); + condition->time_stop = condition->time_start + 3600 * 24; + } else { + condition->time_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; + condition->time_stop = time(0) - 3600 * time_criteria[k].end_hour; } else { - condition->interval_stop = 0; + condition->time_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; } } - if(condition->field_id == -1) { + if(condition->db_key == -1) { /* No time condition matched, look for the search fields */ - for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) { + for(m = 0; (m < MAX_ID) && condition->db_key == -1; m++) { if(strncmp(field_keys[m], search_field, strlen(search_field)) == 0) { - condition->field_id = m; + condition->db_key = m; } } /* None match, if there is a default search field, re-run the search with it */ - if(condition->field_id == -1) { + if(condition->db_key == -1) { if(default_search_field) { - for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) { + for(m = 0; (m < MAX_ID) && condition->db_key == -1; m++) { if(strncmp(field_keys[m], default_search_field, strlen(default_search_field)) == 0) { - condition->field_id = m; + condition->db_key = m; } } string = full_string; @@ -816,28 +856,28 @@ void init_condition(struct search_condition *condition, const char *full_string, } } - if(condition->field_id == -1) { + if(condition->db_key == -1) { fprintf(stderr, "mymail: Syntax error in field key \"%s\".\n", search_field); exit(EXIT_FAILURE); } - if(regcomp(&condition->regexp, + if(regcomp(&condition->db_value_regexp, string, REG_ICASE)) { fprintf(stderr, "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n", string, - field_keys[condition->field_id]); + field_keys[condition->db_key]); exit(EXIT_FAILURE); } } } void free_condition(struct search_condition *condition) { - if(condition->field_id != ID_INTERVAL) { - regfree(&condition->regexp); + if(condition->db_key != ID_TIME_INTERVAL) { + regfree(&condition->db_value_regexp); } }