X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?p=mymail.git;a=blobdiff_plain;f=mymail.c;h=a04fde8eab17dba46399cac7ee8a1fd266628866;hp=df64b5a6047d9ce3b224eb29dd40779d58d9ac4d;hb=HEAD;hpb=8702faa9b8718133b6daf349f61fee7b77348182 diff --git a/mymail.c b/mymail.c index df64b5a..a04fde8 100644 --- a/mymail.c +++ b/mymail.c @@ -21,11 +21,13 @@ /* - 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, to create a resulting mbox file. + mymail is a simple mail indexer. It can: + + (1) scan mbox files, and create a db file containing for each mail a + list of fields computed from its header. + + (2) read such a db file, gets all the mails matching regexp-defined + conditions on the fields, and generates a resulting mbox file. It is low-tech, simple, light and fast. @@ -35,6 +37,7 @@ #include #include +#include #include #include #include @@ -46,22 +49,33 @@ #include #define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file" -#define VERSION "0.9.5" +#define MYMAIL_VERSION "0.9.10" + +#define MYMAIL_DB_FORMAT_VERSION 1 #define MAX_NB_SEARCH_CONDITIONS 32 #define BUFFER_SIZE 65536 #define TOKEN_BUFFER_SIZE 1024 -#define LEADING_FROM_LINE_REGEXP_STRING "^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$" +#define LEADING_FROM_LINE_REGEXP_STRING "^From .*\\(Mon\\|Tue\\|Wed\\|Thu\\|Fri\\|Sat\\|Sun\\) \\(Jan\\|Feb\\|Mar\\|Apr\\|May\\|Jun\\|Jul\\|Aug\\|Sep\\|Oct\\|Nov\\|Dec\\) [ 0123][0-9] [0-9][0-9]:[0-9][0-9]:[0-9][0-9] [0-9][0-9][0-9][0-9]\n$" -/* Global variables! */ +/********************************************************************/ -int paranoid; -int quiet; -int ignore_dot_files; +struct alias_node { + char *alias, *value; + struct alias_node *next; +}; -regex_t leading_from_line_regexp; +/* Global variables! */ + +int global_quiet; +int global_use_leading_time; +int global_nb_mails_max; +int global_discard_mail_from_the_future; +regex_t global_leading_from_line_regexp; +struct alias_node *global_alias_list; +time_t global_current_time; /********************************************************************/ @@ -74,11 +88,14 @@ enum { ID_DATE, ID_PARTICIPANT, ID_BODY, - ID_INTERVAL, + ID_TIME_INTERVAL, + ID_MAIL_ID, + ID_REFERENCE_ID, + ID_THREAD_ID, MAX_ID }; -static char *field_names[] = { +static char *field_keys[] = { "mail", "lead", "from", @@ -87,16 +104,19 @@ static char *field_names[] = { "date", "part", "body", - "interval" + "interval", + "mailid", + "reference", + "thread" }; /********************************************************************/ 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; }; /********************************************************************/ @@ -144,6 +164,20 @@ static struct parsable_field fields_to_parse[] = { { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } }, + { + ID_MAIL_ID, + REG_ICASE, + "^message-id: ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + + { + ID_REFERENCE_ID, + REG_ICASE, + "^\\(in-reply-to\\|references\\): ", + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } + }, + }; /********************************************************************/ @@ -155,12 +189,15 @@ int xor(int a, int b) { const char *parse_token(char *token_buffer, size_t token_buffer_size, char separator, const char *string) { char *u = token_buffer; + while(*string == separator) { string++; } - while(u < token_buffer + token_buffer_size - 1 && *string && - *string != separator) { + + while(u < token_buffer + token_buffer_size - 1 && *string && *string != separator) { *(u++) = *(string++); } + while(*string == separator) { string++; } + *u = '\0'; return string; } @@ -182,14 +219,27 @@ char *default_value(char *current_value, } } +/********************************************************************/ + +void *safe_malloc(size_t n) { + void *p = malloc(n); + if(!p && n != 0) { + fprintf(stderr, + "mymail: cannot allocate memory: %s\n", strerror(errno)); + exit(EXIT_FAILURE); + } + return p; +} + FILE *safe_fopen(const char *path, const char *mode, const char *comment) { FILE *result = fopen(path, mode); if(result) { return result; } else { fprintf(stderr, - "mymail: Cannot open file '%s' (%s) with mode \"%s\".\n", - path, comment, mode); + "mymail: Cannot open file '%s' (%s) with mode \"%s\": %s\n", + path, comment, mode, + strerror(errno)); exit(EXIT_FAILURE); } } @@ -197,7 +247,7 @@ FILE *safe_fopen(const char *path, const char *mode, const char *comment) { /*********************************************************************/ void print_version(FILE *out) { - fprintf(out, "mymail version %s (%s)\n", VERSION, UNAME); + fprintf(out, "mymail version %s (%s)\n", MYMAIL_VERSION, UNAME); } void print_usage(FILE *out) { @@ -212,6 +262,11 @@ void print_usage(FILE *out) { 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, " -f, --do-not-discard-mails-from-the-future\n"); + fprintf(out, " do not ignore mails with a date more than 24h in the future\n"); fprintf(out, " -p , --db-pattern \n"); fprintf(out, " set the db filename pattern for recursive search\n"); fprintf(out, " -r , --db-root \n"); @@ -222,14 +277,17 @@ void print_usage(FILE *out) { fprintf(out, " set the mbox filename pattern for recursive search\n"); fprintf(out, " -s , --search \n"); fprintf(out, " search for matching mails in the db file\n"); - fprintf(out, " -d , --db-file-generate \n"); + fprintf(out, " -d , --db-file-output \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"); + fprintf(out, " -n , --nb-mails-max \n"); + fprintf(out, " set the maximum number of mails to extract\n"); fprintf(out, " -a , --default-search \n"); fprintf(out, " set the default search field\n"); + } /*********************************************************************/ @@ -238,56 +296,40 @@ int ignore_entry(const char *name) { 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; +int db_line_match_search(struct search_condition *condition, + int db_key, const char *db_value) { - 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 - ( + return + ( + (condition->db_key == db_key) - (condition->field_id == mbox_id) + || - || + (condition->db_key == ID_PARTICIPANT && (db_key == ID_LEADING_LINE || + db_key == ID_FROM || + db_key == ID_TO)) + || - (condition->field_id == ID_PARTICIPANT && (mbox_id == ID_LEADING_LINE || - mbox_id == ID_FROM || - mbox_id == ID_TO)) - || + (condition->db_key == ID_FROM && db_key == ID_LEADING_LINE) - (condition->field_id == ID_FROM && mbox_id == ID_LEADING_LINE) + || - ) + (condition->db_key == ID_THREAD_ID && (db_key == ID_MAIL_ID || + db_key == ID_REFERENCE_ID)) + ) - && + && - 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, @@ -308,16 +350,13 @@ void update_body_hits(char *mail_filename, int position_in_mail, 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(raw_mbox_line[0] == '\n') { header = 0; } 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++; } @@ -335,19 +374,21 @@ 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; + /* printf("Extract\n"); */ + mail_file = safe_fopen(mail_filename, "r", "mbox for mail extraction"); + /* fchmod(fileno(mail_file), 0x660); */ fseek(mail_file, position_in_mail, SEEK_SET); if(fgets(raw_mbox_line, BUFFER_SIZE, mail_file)) { fprintf(output_file, "%s", raw_mbox_line); while(1) { if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) || - (is_a_leading_from_line(raw_mbox_line)) - ) + is_a_leading_from_line(raw_mbox_line)) break; fprintf(output_file, "%s", raw_mbox_line); } @@ -356,26 +397,104 @@ 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; + + memset(&tm, 0, sizeof(struct tm)); + + if(db_key == ID_LEADING_LINE) { + c = db_value; + while(*c && *c != ' ') c++; + while(*c && *c == ' ') c++; + /* printf("From %s", db_value); */ + 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)) { + /* printf("Date: %s", db_value); */ + *t = mktime(&tm); + } + } + } + } +} + int search_in_db(const char *db_filename, + int nb_extracted_mails, 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_name[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; - int nb_extracted_mails; + const char *db_value; + int db_key; + int hits[MAX_NB_SEARCH_CONDITIONS]; + int nb_body_conditions, need_time; + time_t mail_time; - nb_extracted_mails = 0; + int m, n; - if(!quiet) { + if(!global_quiet) { printf("Searching in '%s' ... ", db_filename); fflush(stdout); } @@ -401,85 +520,87 @@ 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 = global_discard_mail_from_the_future; + 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_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) { - - /* 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); - } - } + while(nb_extracted_mails < global_nb_mails_max && + fgets(raw_db_line, BUFFER_SIZE, db_file)) { + + /* Removes the CR */ + char *s = raw_db_line; + while(*s && *s != '\n') { s++; } + *s = '\0'; + + 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] && + (!global_discard_mail_from_the_future || mail_time < global_current_time + 3600 * 24) && + 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); + strncpy(current_mail_filename, db_value, PATH_MAX + 1); 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) { - 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); } } } + /* printf("global_discard_mail_from_the_future = %d\n", + global_discard_mail_from_the_future); */ + + if(nb_extracted_mails < global_nb_mails_max && + current_mail_filename[0] && + (!global_discard_mail_from_the_future || mail_time < global_current_time + 3600 * 24) && + 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); } @@ -488,6 +609,7 @@ int search_in_db(const char *db_filename, } int recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp, + int nb_extracted_mails, int nb_search_conditions, struct search_condition *search_conditions, FILE *output_file) { @@ -495,7 +617,6 @@ int recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp, struct dirent *dir_e; struct stat sb; char subname[PATH_MAX + 1]; - int nb_extracted_mails = 0; if(lstat(entry_name, &sb) != 0) { fprintf(stderr, @@ -510,12 +631,14 @@ int recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp, dir = opendir(entry_name); if(dir) { - while((dir_e = readdir(dir))) { + while((dir_e = readdir(dir)) && + nb_extracted_mails < global_nb_mails_max) { if(!ignore_entry(dir_e->d_name)) { snprintf(subname, PATH_MAX, "%s/%s", entry_name, dir_e->d_name); - nb_extracted_mails += recursive_search_in_db(subname, db_filename_regexp, - nb_search_conditions, search_conditions, - output_file); + nb_extracted_mails = recursive_search_in_db(subname, db_filename_regexp, + nb_extracted_mails, + nb_search_conditions, search_conditions, + output_file); } } closedir(dir); @@ -526,8 +649,10 @@ int recursive_search_in_db(const char *entry_name, regex_t *db_filename_regexp, while(*s) { if(*s == '/') { filename = s+1; } s++; } if(regexec(db_filename_regexp, filename, 0, 0, 0) == 0) { - nb_extracted_mails += - search_in_db(entry_name, nb_search_conditions, search_conditions, output_file); + nb_extracted_mails = + search_in_db(entry_name, + nb_extracted_mails, + nb_search_conditions, search_conditions, output_file); } } @@ -544,7 +669,7 @@ void index_one_mbox_line(unsigned int nb_fields_to_parse, 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); } } @@ -570,17 +695,29 @@ void index_mbox(const char *mbox_filename, 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) { @@ -590,6 +727,7 @@ void index_mbox(const char *mbox_filename, } 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++) = ' '; @@ -601,17 +739,10 @@ void index_mbox(const char *mbox_filename, } 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); } @@ -679,7 +810,9 @@ static struct option long_options[] = { { "help", no_argument, 0, 'h' }, { "version", no_argument, 0, 'v' }, { "quiet", no_argument, 0, 'q' }, - { "db-file-generate", 1, 0, 'd' }, + { "use-leading-time", no_argument, 0, 't' }, + { "do-not-discard-mails-from-the-future", no_argument, 0, 'f' }, + { "db-file-output", 1, 0, 'd' }, { "db-pattern", 1, 0, 'p' }, { "db-root", 1, 0, 'r' }, { "db-list", 1, 0, 'l' }, @@ -688,11 +821,13 @@ static struct option long_options[] = { { "index", 0, 0, 'i' }, { "output", 1, 0, 'o' }, { "default-search", 1, 0, 'a' }, + { "nb-mails-max", 1, 0, 'n' }, { 0, 0, 0, 0 } }; struct time_criterion { char *label; + int day_criterion; int start_hour, end_hour; int past_week_day; }; @@ -701,37 +836,44 @@ 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 }, - - { "yesterday", 48, 24, -1 }, - - { "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 }, + { "1h", 0, 1, -1, -1 }, + { "2h", 0, 2, -1, -1 }, + { "4h", 0, 4, -1, -1 }, + { "8h", 0, 8, -1, -1 }, + { "24h", 0, 24, -1, -1 }, + { "48h", 0, 48, -1, -1 }, + { "week", 0, 24 * 7, -1, -1 }, + { "2weeks", 0, 24 * 14, -1, -1 }, + { "month", 0, 24 * 31, -1, -1 }, + { "semester", 0, 24 * 185, -1, -1 }, + { "trimester", 0, 24 * 92, -1, -1 }, + { "year", 0, 24 * 365, -1, -1 }, + + { "yesterday", 1, -1, -1, -1 }, + { "today", 1, -1, -1, 0 }, + + { "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 }, }; /*********************************************************************/ 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; - if(delta_day == 0) { delta_day = 7; } - return t - (delta_day * 3600 * 24 + tm->tm_sec + 60 * tm->tm_min + 3600 * tm->tm_hour); + tm = localtime(&global_current_time); + if(day > 0) { + delta_day = (7 + tm->tm_wday - day) % 7; + } else { + delta_day = - day; + } + return global_current_time - (delta_day * 3600 * 24 + tm->tm_sec + 60 * tm->tm_min + 3600 * tm->tm_hour); } void init_condition(struct search_condition *condition, const char *full_string, @@ -739,6 +881,14 @@ void init_condition(struct search_condition *condition, const char *full_string, char full_search_field[TOKEN_BUFFER_SIZE], *search_field; unsigned int k, m; const char *string; + struct alias_node *a; + + for(a = global_alias_list; a; a = a->next) { + if(strcmp(full_string, a->alias) == 0) { + full_string = a->value; + break; + } + } string = parse_token(full_search_field, TOKEN_BUFFER_SIZE, ' ', full_string); search_field = full_search_field; @@ -750,74 +900,142 @@ 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 = global_current_time - 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 = global_current_time - 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++) { - if(strncmp(field_names[m], search_field, strlen(search_field)) == 0) { - condition->field_id = m; + for(m = 0; (m < MAX_ID) && condition->db_key == -1; m++) { + if(strncmp(field_keys[m], search_field, strlen(search_field)) == 0) { + 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++) { - if(strncmp(field_names[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; + if(string[0] == '!') { string++; } } } - if(condition->field_id == -1) { + if(condition->db_key == -1) { fprintf(stderr, - "mymail: Syntax error in field name \"%s\".\n", + "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_names[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); + } +} + +const char *eat_space(const char *s) { + while(*s == ' ' || *s == '\t') { s++; } + return s; +} + +void read_rc_file(const char *rc_filename) { + char raw_line[BUFFER_SIZE]; + char command[TOKEN_BUFFER_SIZE], tmp_token[TOKEN_BUFFER_SIZE]; + + FILE *rc_file; + int line_number; + const char *s; + char *t; + + rc_file = fopen(rc_filename, "r"); + + if(rc_file) { + line_number = 1; + while(fgets(raw_line, BUFFER_SIZE, rc_file)) { + t = raw_line; + while(*t) { if(*t == '\n') { *t = '\0'; }; t++; } + + s = raw_line; + s = eat_space(s); + + if(*s && *s != '#') { + s = parse_token(command, TOKEN_BUFFER_SIZE, ' ', s); + + if(strcmp(command, "alias") == 0) { + struct alias_node *a = safe_malloc(sizeof(struct alias_node)); + a->next = global_alias_list; + global_alias_list = a; + if(s) { + s = eat_space(s); + s = parse_token(tmp_token, TOKEN_BUFFER_SIZE, '=', s); + a->alias = strdup(tmp_token); + if(s) { + s = eat_space(s); + a->value = strdup(s); + } else { + fprintf(stderr, "%s:%d syntax error, missing alias value.\n", + rc_filename, + line_number); + exit(EXIT_FAILURE); + } + } else { + fprintf(stderr, "%s:%d syntax error, missing alias key.\n", + rc_filename, + line_number); + exit(EXIT_FAILURE); + } + } else { + fprintf(stderr, "%s:%d syntax error, unknown command '%s'.\n", + rc_filename, + line_number, + command); + exit(EXIT_FAILURE); + } + } + + line_number++; + } + fclose(rc_file); } } @@ -833,6 +1051,7 @@ int main(int argc, char **argv) { char *mbox_filename_regexp_string = 0; char *default_search_field; char output_filename[PATH_MAX + 1]; + char rc_filename[PATH_MAX + 1]; int action_index = 0; int error = 0, show_help = 0; const unsigned int nb_fields_to_parse = @@ -841,24 +1060,54 @@ int main(int argc, char **argv) { unsigned int f, n; unsigned int nb_search_conditions; struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS]; + struct alias_node *a, *b; + + /* Group and others have no access */ + umask(S_IRWXG | S_IRWXO); - 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; + if(getenv("MYMAILRC")) { + sprintf(rc_filename, "%s", getenv("MYMAILRC")); + } else if(getenv("HOME")) { + sprintf(rc_filename, "%s/.mymailrc", getenv("HOME")); + } else { + rc_filename[0] = '\0'; + } + + global_alias_list = 0; + global_quiet = 0; + global_use_leading_time = 0; + global_nb_mails_max = 250; + global_discard_mail_from_the_future = 1; + global_current_time = time(0); + default_search_field = 0; - ignore_dot_files = 1; strncpy(output_filename, "", PATH_MAX); + if(rc_filename[0]) { + read_rc_file(rc_filename); + } + + /* + { + #warning Test code added on 2013 May 02 11:17:01 + struct alias_node *a; + for(a = global_alias_list; a; a = a->next) { + printf ("ALIAS [%s] [%s]\n", a->alias, a->value); + } + } + */ + setlocale(LC_ALL, ""); nb_search_conditions = 0; - while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:a:m:", + while ((c = getopt_long(argc, argv, "hvqtfip:s:d:r:l:o:a:m:", long_options, NULL)) != -1) { switch(c) { @@ -872,7 +1121,15 @@ int main(int argc, char **argv) { break; case 'q': - quiet = 1; + global_quiet = 1; + break; + + case 't': + global_use_leading_time = 1; + break; + + case 'f': + global_discard_mail_from_the_future = 0; break; case 'i': @@ -936,12 +1193,26 @@ int main(int argc, char **argv) { default_search_field = optarg; break; + case 'n': + global_nb_mails_max = atoi(optarg); + break; + default: error = 1; break; } } + if(error) { + print_usage(stderr); + exit(EXIT_FAILURE); + } + + if(show_help) { + print_usage(stdout); + exit(EXIT_SUCCESS); + } + /* Set all the values that may defined in the arguments, through environment variables, or hard-coded */ @@ -965,18 +1236,6 @@ int main(int argc, char **argv) { "MYMAIL_MBOX_PATTERN", 0); - /* Start the processing */ - - if(error) { - print_usage(stderr); - exit(EXIT_FAILURE); - } - - if(show_help) { - print_usage(stdout); - exit(EXIT_SUCCESS); - } - /* mbox indexing */ if(action_index) { @@ -1007,12 +1266,16 @@ int main(int argc, char **argv) { 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); } } - fprintf(db_file, "%s version_%s raw\n", MYMAIL_DB_MAGIC_TOKEN, VERSION); + fprintf(db_file, + "%s version_%s format_%d raw\n", + MYMAIL_DB_MAGIC_TOKEN, + MYMAIL_VERSION, + MYMAIL_DB_FORMAT_VERSION); while(optind < argc) { recursive_index_mbox(db_file, @@ -1044,7 +1307,7 @@ int main(int argc, char **argv) { output_file = safe_fopen(output_filename, "w", "result mbox"); } else { output_file = stdout; - quiet = 1; + global_quiet = 1; } if(nb_search_conditions > 0) { @@ -1062,9 +1325,10 @@ int main(int argc, char **argv) { exit(EXIT_FAILURE); } - nb_extracted_mails += recursive_search_in_db(db_root_path, &db_filename_regexp, - nb_search_conditions, search_conditions, - output_file); + nb_extracted_mails = recursive_search_in_db(db_root_path, &db_filename_regexp, + nb_extracted_mails, + nb_search_conditions, search_conditions, + output_file); regfree(&db_filename_regexp); } @@ -1081,8 +1345,10 @@ int main(int argc, char **argv) { s = parse_token(db_filename, PATH_MAX + 1, ';', s); if(db_filename[0]) { - nb_extracted_mails += - search_in_db(db_filename, nb_search_conditions, search_conditions, output_file); + nb_extracted_mails = + search_in_db(db_filename, + nb_extracted_mails, + nb_search_conditions, search_conditions, output_file); } } } @@ -1090,13 +1356,15 @@ int main(int argc, char **argv) { /* Search in all db files listed in the command arguments */ while(optind < argc) { - nb_extracted_mails += - search_in_db(argv[optind], nb_search_conditions, search_conditions, output_file); + nb_extracted_mails = + search_in_db(argv[optind], + nb_extracted_mails, + nb_search_conditions, search_conditions, output_file); optind++; } } - if(!quiet) { + if(!global_quiet) { if(nb_extracted_mails > 0) { printf("Found %d matching mails.\n", nb_extracted_mails); } else { @@ -1115,13 +1383,22 @@ int main(int argc, char **argv) { free_condition(&search_conditions[n]); } + a = global_alias_list; + while(a) { + b = a->next; + free(a->alias); + free(a->value); + free(a); + a = b; + } + free(db_filename); free(db_filename_regexp_string); free(db_root_path); free(db_filename_list); free(mbox_filename_regexp_string); - regfree(&leading_from_line_regexp); + regfree(&global_leading_from_line_regexp); exit(EXIT_SUCCESS); }