X-Git-Url: https://www.fleuret.org/cgi-bin/gitweb/gitweb.cgi?p=mymail.git;a=blobdiff_plain;f=mymail.c;h=bac2f45ebb0ad3fa1bd7b71451933c624a756fd4;hp=dc7c58c2d4429bb7b31aeded25a99ddaf8bffa15;hb=cbe29e579b66cf31ec64c142286e7e2b865d84d7;hpb=8dea757944e71980e7b76ede4553eb1fb457c1ec diff --git a/mymail.c b/mymail.c index dc7c58c..bac2f45 100644 --- a/mymail.c +++ b/mymail.c @@ -19,6 +19,18 @@ * */ +/* + + This command is a dumb mail indexer. It can either (1) scan + directories containing mbox files, and create a db file containing + for each mail a list of fields computed from the header, or (2) + read such a db file and get all the mails matching regexp-defined + conditions on the fields. + + It is low-tech, simple, light and fast. + +*/ + #define _GNU_SOURCE #include @@ -31,19 +43,122 @@ #include #include #include +#include + +#define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file" +#define VERSION "0.9.1" + +#define MAX_NB_SEARCH_CONDITIONS 10 + +#define BUFFER_SIZE 65536 +#define TOKEN_BUFFER_SIZE 1024 + +regex_t leading_from_line_regexp; + +/* Global variables! */ + +int paranoid; +int quiet; -#define VERSION "0.1" +/********************************************************************/ + +enum { + ID_MAIL = 0, + ID_LEADING_LINE, + ID_FROM, + ID_TO, + ID_SUBJECT, + ID_DATE, + ID_PARTICIPANT, + ID_BODY, + ID_INTERVAL, + MAX_ID +}; -#define BUFFER_SIZE 16384 +static char *field_names[] = { + "mail", + "lead", + "from", + "to", + "subject", + "date", + "part", + "body", + "interval" +}; + +/********************************************************************/ + +struct search_condition { + int field_id; + int negation; + regex_t regexp; + time_t interval_start, interval_stop; +}; + +/********************************************************************/ struct parsable_field { - char *name; + int id; char *regexp_string; regex_t regexp; - FILE *db_file; }; -char *db_filename_prefix; +static struct parsable_field fields_to_parse[] = { + { + ID_LEADING_LINE, + "^From ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + + { + ID_FROM, + "^\\([Ff][Rr][Oo][Mm]:\\|[Rr][Ee][Pp][Ll][Yy]-[Tt][Oo]:\\|[Ss][Ee][Nn][Dd][Ee][Rr]:\\)", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + + { + ID_TO, + "^\\([Tt][Oo]\\|[Cc][Cc]\\|[Bb][Cc][Cc]\\): ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + + { + ID_SUBJECT, + "^[Ss][Uu][Bb][Jj][Ee][Cc][Tt]: ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + + { + ID_DATE, + "^[Dd][Aa][Tt][Ee]: ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + +}; + +/********************************************************************/ + +int xor(int a, int b) { + return (a && !b) || (!a && b); +} + +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) { + while(*c && *c != '\n' && *c != '\r') c++; + *c = '\0'; +} /********************************************************************/ @@ -53,7 +168,7 @@ void *safe_malloc(size_t n) { void *p = malloc(n); if(!p && n != 0) { fprintf(stderr, - "mymail: can not allocate memory: %s\n", strerror(errno)); + "mymail: cannot allocate memory: %s\n", strerror(errno)); exit(EXIT_FAILURE); } return p; @@ -61,113 +176,491 @@ void *safe_malloc(size_t n) { /*********************************************************************/ -void usage(FILE *out) { +void print_version(FILE *out) { fprintf(out, "mymail version %s (%s)\n", VERSION, UNAME); +} + +void print_usage(FILE *out) { + print_version(out); fprintf(out, "Written by Francois Fleuret .\n"); fprintf(out, "\n"); - fprintf(out, "Usage: mymail [options] [ [ ...]]\n"); + fprintf(out, "Usage: mymail [options] [ [ ...]| [ ...]]\n"); fprintf(out, "\n"); + fprintf(out, " -h, --help\n"); + fprintf(out, " show this help\n"); + fprintf(out, " -v, --version\n"); + fprintf(out, " print the version number\n"); + fprintf(out, " -q, --quiet\n"); + fprintf(out, " do not print information during search\n"); + fprintf(out, " -p , --db-pattern \n"); + fprintf(out, " set the db filename pattern for recursive search\n"); + fprintf(out, " -r , --db-root \n"); + fprintf(out, " set the db root path for recursive search\n"); + fprintf(out, " -l , --db-list \n"); + fprintf(out, " set the semicolon-separated list of db files for search\n"); + fprintf(out, " -s , --search \n"); + fprintf(out, " search for matching mails in the db file\n"); + fprintf(out, " -d , --db-file \n"); + fprintf(out, " set the db filename for indexing\n"); + fprintf(out, " -i, --index\n"); + fprintf(out, " index mails\n"); + fprintf(out, " -o , --output \n"); + fprintf(out, " set the result file, use stdout if unset\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; +} + +/*********************************************************************/ + +int ignore_entry(const char *name) { + return + /* strcmp(name, ".") == 0 || */ + /* strcmp(name, "..") == 0 || */ + (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 mbox_line_match_search(struct search_condition *condition, + int mbox_id, char *mbox_value) { + + if(condition->field_id == ID_INTERVAL) { + if(mbox_id == ID_LEADING_LINE) { + 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) + + || + + (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) + + ) + && + regexec(&condition->regexp, mbox_value, 0, 0, 0) == 0; + } +} + +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) || + (last_mbox_line_was_empty && + is_a_leading_from_line(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) { + + int hits[MAX_NB_SEARCH_CONDITIONS]; + char raw_db_line[BUFFER_SIZE]; + char raw_mbox_line[BUFFER_SIZE]; + char current_mail_filename[PATH_MAX + 1]; + unsigned long int current_position_in_mail; + 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; + + for(n = 0; n < nb_search_conditions; n++) { hits[n] = 0; } + + nb_body_conditions = 0; + for(n = 0; n < nb_search_conditions; n++) { + if(search_conditions[n].field_id == ID_BODY) { + nb_body_conditions++; + } + } + + 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]; + + 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) { + + /* all conditions but the body ones are fine, check the body + ones */ + + if(nb_body_conditions > 0) { + update_body_hits(current_mail_filename, current_position_in_mail, + nb_search_conditions, search_conditions, + nb_body_conditions, + hits); + } + + nb_fulfilled_body_conditions = 0; + + 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) { + FILE *mail_file; + + mail_file = fopen(current_mail_filename, "r"); + + if(!mail_file) { + fprintf(stderr, + "mymail: Cannot open mbox '%s' for mail extraction.\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)) { + 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)) + ) + break; + last_mbox_line_was_empty = (raw_mbox_line[0] == '\n'); + fprintf(output_file, "%s", raw_mbox_line); + } + } + + fclose(mail_file); + } + } + } + + 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, ' ', mbox_value); + current_position_in_mail = atol(position_in_file_string); + remove_eof(current_mail_filename); + 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) { + mbox_id = m; + } + } + for(n = 0; n < nb_search_conditions; n++) { + hits[n] |= mbox_line_match_search(&search_conditions[n], + mbox_id, mbox_value); + } + } + } + + fclose(db_file); + + if(!quiet) { + printf("done.\n"); + fflush(stdout); + } +} + +void recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp, + int nb_search_conditions, + struct search_condition *search_conditions, + FILE *output_file) { + DIR *dir; + struct dirent *dir_e; + struct stat sb; + char subname[PATH_MAX + 1]; + + if(lstat(entry_name, &sb) != 0) { + fprintf(stderr, + "mymail: Cannot stat \"%s\": %s\n", + entry_name, + strerror(errno)); + exit(EXIT_FAILURE); + } + + dir = opendir(entry_name); + + if(dir) { + while((dir_e = readdir(dir))) { + if(!ignore_entry(dir_e->d_name)) { + snprintf(subname, PATH_MAX, "%s/%s", entry_name, dir_e->d_name); + recursive_search_in_db(subname, db_filename_regexp, + nb_search_conditions, search_conditions, + output_file); + } + } + closedir(dir); + } + + else { + const char *s = entry_name, *filename = entry_name; + while(*s) { if(*s == '/') { filename = s+1; } s++; } + + if(regexec(db_filename_regexp, filename, 0, 0, 0) == 0) { + search_in_db(entry_name, nb_search_conditions, search_conditions, output_file); + } + } } -void read_file(const char *input_filename, - int nb_fields_to_parse, struct parsable_field *fields_to_parse) { - char raw_line[BUFFER_SIZE]; +/*********************************************************************/ + +void index_one_mbox_line(int nb_fields_to_parse, struct parsable_field *fields_to_parse, + char *raw_mbox_line, FILE *db_file) { + regmatch_t matches; + int f; + 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], + raw_mbox_line + matches.rm_eo); + } + } +} + +void index_mbox(const char *mbox_filename, + int nb_fields_to_parse, struct parsable_field *fields_to_parse, + FILE *db_file) { + char raw_mbox_line[BUFFER_SIZE], full_line[BUFFER_SIZE]; + char *end_of_full_line; FILE *file; - int in_header; - unsigned int position_in_file; + int in_header, new_header, last_mbox_line_was_empty; + unsigned long int position_in_file; - file = fopen(input_filename, "r"); + file = fopen(mbox_filename, "r"); if(!file) { - fprintf(stderr, "mymail: Can not open `%s'.\n", input_filename); - exit(EXIT_FAILURE); + fprintf(stderr, "mymail: Cannot open '%s'.\n", mbox_filename); + if(paranoid) { exit(EXIT_FAILURE); } + return; } in_header = 0; + new_header = 0; position_in_file = 0; + end_of_full_line = 0; + full_line[0] = '\0'; + last_mbox_line_was_empty = 1; - while(fgets(raw_line, BUFFER_SIZE, file)) { - if(strncmp(raw_line, "From ", 5) == 0) { + while(fgets(raw_mbox_line, BUFFER_SIZE, file)) { + if(last_mbox_line_was_empty && + is_a_leading_from_line(raw_mbox_line)) { if(in_header) { fprintf(stderr, - "Got a 'From ' in the header in %s:%u.\n", - input_filename, position_in_file); - fprintf(stderr, "%s", raw_line); - exit(EXIT_FAILURE); + "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); } } in_header = 1; - } else if(strncmp(raw_line, "\n", 1) == 0) { + new_header = 1; + } else if(raw_mbox_line[0] == '\n') { if(in_header) { in_header = 0; } } - /* if(in_header) { */ - /* printf("LINE.H %s", raw_line); */ - /* } else { */ - /* printf("LINE.B %s", raw_line); */ - /* } */ + last_mbox_line_was_empty = (raw_mbox_line[0] == '\n'); if(in_header) { - int f; - regmatch_t matches; - for(f = 0; f < nb_fields_to_parse; f++) { - if(regexec(&fields_to_parse[f].regexp, raw_line, 1, &matches, 0) == 0) { - fprintf(fields_to_parse[f].db_file, "%s:%d %s", - input_filename, position_in_file, - raw_line + matches.rm_eo); + if(new_header) { + fprintf(db_file, "mail %lu %s\n", position_in_file, mbox_filename); + new_header = 0; + } + + if(raw_mbox_line[0] == ' ' || raw_mbox_line[0] == '\t') { + char *start = raw_mbox_line; + while(*start == ' ' || *start == '\t') start++; + *(end_of_full_line++) = ' '; + strcpy(end_of_full_line, start); + while(*end_of_full_line && *end_of_full_line != '\n') { + end_of_full_line++; } + *end_of_full_line = '\0'; } + + 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); + } + */ + + if(full_line[0]) { + index_one_mbox_line(nb_fields_to_parse, fields_to_parse, full_line, db_file); + } + + end_of_full_line = full_line; + strcpy(end_of_full_line, raw_mbox_line); + while(*end_of_full_line && *end_of_full_line != '\n') { + end_of_full_line++; + } + *end_of_full_line = '\0'; + } + } - position_in_file += strlen(raw_line); + position_in_file += strlen(raw_mbox_line); } fclose(file); } -int ignore_entry(const char *name) { - return - strcmp(name, ".") == 0 || - strcmp(name, "..") == 0 || - (name[0] == '.' && name[1] != '/'); -} - -void process_entry(const char *dir_name, - int nb_fields_to_parse, struct parsable_field *fields_to_parse) { +void recursive_index_mbox(FILE *db_file, + const char *entry_name, + int nb_fields_to_parse, struct parsable_field *fields_to_parse) { DIR *dir; struct dirent *dir_e; struct stat sb; char subname[PATH_MAX + 1]; - if(lstat(dir_name, &sb) != 0) { + if(lstat(entry_name, &sb) != 0) { fprintf(stderr, - "mymail: Can not stat \"%s\": %s\n", - dir_name, + "mymail: Cannot stat \"%s\": %s\n", + entry_name, strerror(errno)); exit(EXIT_FAILURE); - } else { } - if(S_ISLNK(sb.st_mode)) { - return; - } - - dir = opendir(dir_name); + dir = opendir(entry_name); if(dir) { - printf("Processing directory '%s'.\n", dir_name); while((dir_e = readdir(dir))) { if(!ignore_entry(dir_e->d_name)) { - snprintf(subname, PATH_MAX, "%s/%s", dir_name, dir_e->d_name); - process_entry(subname, nb_fields_to_parse, fields_to_parse); + snprintf(subname, PATH_MAX, "%s/%s", entry_name, dir_e->d_name); + recursive_index_mbox(db_file, subname, nb_fields_to_parse, fields_to_parse); } } closedir(dir); } else { - if(S_ISREG(sb.st_mode)) { - /* printf("Processing regular file '%s'.\n", dir_name); */ - read_file(dir_name, nb_fields_to_parse, fields_to_parse); - } + index_mbox(entry_name, nb_fields_to_parse, fields_to_parse, db_file); } } @@ -175,40 +668,181 @@ void process_entry(const char *dir_name, /* For long options that have no equivalent short option, use a non-character as a pseudo short option, starting with CHAR_MAX + 1. */ -enum -{ +enum { OPT_BASH_MODE = CHAR_MAX + 1 }; static struct option long_options[] = { { "help", no_argument, 0, 'h' }, - { "db-prefix", 1, 0, 'p' }, + { "version", no_argument, 0, 'v' }, + { "quiet", no_argument, 0, 'q' }, + { "db-file", 1, 0, 'd' }, + { "db-pattern", 1, 0, 'p' }, + { "db-root", 1, 0, 'r' }, + { "db-list", 1, 0, 'l' }, + { "search", 1, 0, 's' }, + { "index", 0, 0, 'i' }, + { "output", 1, 0, 'o' }, { 0, 0, 0, 0 } }; -static struct parsable_field fields_to_parse[] = { - { - "from", - "^[Ff][Rr][Oo][Mm]: *", - { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, 0 - }, +/*********************************************************************/ - { - "dest", - "^\\([Tt][Oo]\\|[Cc][Cc]\\|[Bb][Cc][Cc]\\): *", - { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, 0 - }, -}; +void init_condition(struct search_condition *condition, char *string) { + char full_search_field[TOKEN_BUFFER_SIZE], *search_field; + int m; + + string = parse_token(full_search_field, TOKEN_BUFFER_SIZE, ' ', string); + search_field = full_search_field; + + if(search_field[0] == '!') { + search_field++; + condition->negation = 1; + } else { + condition->negation = 0; + } + + /* Last 8 hours */ + + if(strcmp(search_field, "8h") == 0) { + condition->field_id = ID_INTERVAL; + condition->interval_start = time(0) - 3600 * 8; + condition->interval_stop = 0; + } + + /* Today and yesterday */ + + 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; + } + + 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 { + 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) { + 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; + 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\\|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$", + 0)) { + fprintf(stderr, + "mymail: Cannot compile leading \"from\" line regexp. That is strange.\n"); + exit(EXIT_FAILURE); + } + + paranoid = 0; + action_index = 0; + db_filename = 0; + db_root_path = 0; + db_filename_list = 0; + quiet = 0; setlocale(LC_ALL, ""); - while ((c = getopt_long(argc, argv, "hp:", + nb_search_conditions = 0; + + while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:", long_options, NULL)) != -1) { switch(c) { @@ -217,8 +851,45 @@ int main(int argc, char **argv) { show_help = 1; break; + case 'v': + print_version(stdout); + break; + + case 'q': + quiet = 1; + break; + + case 'i': + action_index = 1; + break; + + case 'd': + db_filename = strdup(optarg); + break; + case 'p': - db_filename_prefix = strdup(optarg); + db_filename_regexp_string = strdup(optarg); + break; + + case 'o': + strncpy(output_filename, optarg, PATH_MAX); + break; + + case 'r': + db_root_path = strdup(optarg); + break; + + case 'l': + db_filename_list = strdup(optarg); + break; + + case 's': + if(nb_search_conditions == MAX_NB_SEARCH_CONDITIONS) { + fprintf(stderr, "mymail: Too many search patterns.\n"); + exit(EXIT_FAILURE); + } + init_condition(&search_conditions[nb_search_conditions], optarg); + nb_search_conditions++; break; default: @@ -227,55 +898,174 @@ int main(int argc, char **argv) { } } - if(!db_filename_prefix) { - db_filename_prefix = strdup("/tmp/mymail_"); + if(!db_filename) { + char *default_db_filename = getenv("MYMAIL_DB_FILE"); + + if(!default_db_filename) { + default_db_filename = "mymail.db"; + } + + db_filename = strdup(default_db_filename); + } + + if(!db_filename_regexp_string) { + char *default_db_filename_regexp_string = getenv("MYMAIL_DB_PATTERN"); + + if(!default_db_filename_regexp_string) { + default_db_filename_regexp_string = "^mymail.db$"; + } + + db_filename_regexp_string = strdup(default_db_filename_regexp_string); + } + + if(!db_root_path) { + char *default_db_root_path = getenv("MYMAIL_DB_ROOT"); + + if(default_db_root_path) { + db_root_path = strdup(default_db_root_path); + } + } + + if(!db_filename_list) { + char *default_db_filename_list = getenv("MYMAIL_DB_LIST"); + + if(default_db_filename_list) { + db_filename_list = strdup(default_db_filename_list); + } + } + + if(output_filename[0]) { + output_file = fopen(output_filename, "w"); + + if(!output_file) { + fprintf(stderr, + "mymail: Cannot open result file \"%s\" for writing: %s\n", + output_filename, + strerror(errno)); + exit(EXIT_FAILURE); + } + } else { + output_file = stdout; + quiet = 1; } if(error) { - usage(stderr); + print_usage(stderr); exit(EXIT_FAILURE); } if(show_help) { - usage(stdout); + print_usage(stdout); exit(EXIT_SUCCESS); } - for(f = 0; f < nb_fields_to_parse; f++) { - char db_filename[BUFFER_SIZE]; - sprintf(db_filename, "%s%s", db_filename_prefix, fields_to_parse[f].name); - fields_to_parse[f].db_file = fopen(db_filename, "w"); - if(!fields_to_parse[f].db_file) { + if(action_index) { + FILE *db_file; + + db_file = fopen(db_filename, "w"); + + if(!db_file) { fprintf(stderr, - "mymail: Can not open \"%s\" for writing: %s\n", + "mymail: Cannot open \"%s\" for writing: %s\n", db_filename, strerror(errno)); exit(EXIT_FAILURE); } - printf("Initialized %s.\n", db_filename); + for(f = 0; f < nb_fields_to_parse; f++) { + if(regcomp(&fields_to_parse[f].regexp, + fields_to_parse[f].regexp_string, + REG_ICASE)) { + 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]); + exit(EXIT_FAILURE); + } + } + + fprintf(db_file, "%s version_%s raw\n", MYMAIL_DB_MAGIC_TOKEN, VERSION); - if(regcomp(&fields_to_parse[f].regexp, - fields_to_parse[f].regexp_string, - REG_ICASE)) { - fprintf(stderr, - "mymail: Syntax error in regexp \"%s\" for field \"%s\".\n", - fields_to_parse[f].regexp_string, - fields_to_parse[f].name); - exit(EXIT_FAILURE); + while(optind < argc) { + recursive_index_mbox(db_file, + argv[optind], + nb_fields_to_parse, fields_to_parse); + optind++; + } + + fflush(db_file); + fclose(db_file); + + for(f = 0; f < nb_fields_to_parse; f++) { + regfree(&fields_to_parse[f].regexp); } } - while(optind < argc) { - process_entry(argv[optind], - nb_fields_to_parse, fields_to_parse); - optind++; + else { + + if(nb_search_conditions > 0) { + + /* Recursive search if db_root_path is set */ + + if(db_root_path) { + regex_t db_filename_regexp; + if(regcomp(&db_filename_regexp, + db_filename_regexp_string, + 0)) { + fprintf(stderr, + "mymail: Syntax error in regexp \"%s\".\n", + db_filename_regexp_string); + exit(EXIT_FAILURE); + } + + recursive_search_in_db(db_root_path, &db_filename_regexp, + nb_search_conditions, search_conditions, + output_file); + + regfree(&db_filename_regexp); + } + + /* Search in all db files listed in db_filename_list */ + + if(db_filename_list) { + char db_filename[PATH_MAX + 1]; + char *s; + + s = db_filename_list; + + while(*s) { + s = parse_token(db_filename, PATH_MAX + 1, ';', s); + + if(db_filename[0]) { + 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) { + search_in_db(argv[optind], nb_search_conditions, search_conditions, output_file); + optind++; + } + } } - for(f = 0; f < nb_fields_to_parse; f++) { - fclose(fields_to_parse[f].db_file); - regfree(&fields_to_parse[f].regexp); + for(n = 0; n < nb_search_conditions; n++) { + free_condition(&search_conditions[n]); + } + + if(output_file != stdout) { + fflush(output_file); + fclose(output_file); } + free(db_filename); + free(db_filename_regexp_string); + free(db_root_path); + free(db_filename_list); + + regfree(&leading_from_line_regexp); + exit(EXIT_SUCCESS); }