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.
+ conditions on the fields, to create a resulting mbox file.
It is low-tech, simple, light and fast.
#include <limits.h>
#include <dirent.h>
#include <regex.h>
+#include <time.h>
#define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file"
-#define VERSION "0.9"
+#define VERSION "0.9.6"
-#define MAX_NB_SEARCH_PATTERNS 10
+#define MAX_NB_SEARCH_CONDITIONS 32
#define BUFFER_SIZE 65536
+#define TOKEN_BUFFER_SIZE 1024
-char *db_filename;
-char *db_root_path;
+#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$"
-int paranoid;
-int action_index;
+/* Global variables! */
+
+int global_quiet;
+int global_use_leading_time;
+
+regex_t global_leading_from_line_regexp;
/********************************************************************/
enum {
ID_MAIL = 0,
+ ID_LEADING_LINE,
ID_FROM,
- ID_DEST,
+ ID_TO,
ID_SUBJECT,
+ ID_DATE,
ID_PARTICIPANT,
+ ID_BODY,
+ ID_TIME_INTERVAL,
MAX_ID
};
-static char *field_names[] = {
+static char *field_keys[] = {
"mail",
+ "lead",
"from",
- "dest",
+ "to",
"subject",
- "part"
+ "date",
+ "part",
+ "body",
+ "interval"
};
/********************************************************************/
-struct search_request {
+struct search_condition {
int field_id;
int negation;
regex_t regexp;
+ time_t interval_start, interval_stop;
};
/********************************************************************/
struct parsable_field {
int id;
+ int cflags;
char *regexp_string;
regex_t regexp;
};
static struct parsable_field fields_to_parse[] = {
+ {
+ ID_LEADING_LINE,
+ 0,
+ "^From ",
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
+ },
+
{
ID_FROM,
- "^\\(From \\|[Ff][Rr][Oo][Mm]:\\|[R][r][E][e][P][p][L][l][Y][y]-[T][t][O][o]:\\)",
+ REG_ICASE,
+ "^\\(from\\|reply-to\\|sender\\|return-path\\): ",
{ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
},
{
- ID_DEST,
- "^\\([Tt][Oo]\\|[Cc][Cc]\\|[Bb][Cc][Cc]\\): ",
+ ID_TO,
+ REG_ICASE,
+ "^\\(to\\|cc\\|bcc\\): ",
{ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
},
{
ID_SUBJECT,
- "^[Ss][Uu][Bb][Jj][Ee][Cc][Tt]: ",
+ REG_ICASE,
+ "^subject: ",
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
+ },
+
+ {
+ ID_DATE,
+ REG_ICASE,
+ "^date: ",
{ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
},
return (a && !b) || (!a && b);
}
-char *segment_next_field(char *current) {
- while(*current && *current != ' ') current++;
- *current = '\0'; current++;
- while(*current && *current == ' ') current++;
- return current;
+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) {
+ *(u++) = *(string++);
+ }
+ while(*string == separator) { string++; }
+ *u = '\0';
+ return string;
}
-void remove_eof(char *c) {
- while(*c && *c != '\n' && *c != '\r') c++;
- *c = '\0';
+char *default_value(char *current_value,
+ const char *env_variable,
+ const char *hard_default_value) {
+ if(current_value) {
+ return current_value;
+ } else {
+ char *env_value = getenv(env_variable);
+ if(env_value) {
+ return strdup(env_value);
+ } else if(hard_default_value) {
+ return strdup(hard_default_value);
+ } else {
+ return 0;
+ }
+ }
}
-/********************************************************************/
-
-/* malloc with error checking. */
-
-void *safe_malloc(size_t n) {
- void *p = malloc(n);
- if(!p && n != 0) {
+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 allocate memory: %s\n", strerror(errno));
+ "mymail: Cannot open file '%s' (%s) with mode \"%s\".\n",
+ path, comment, mode);
exit(EXIT_FAILURE);
}
- return p;
}
/*********************************************************************/
print_version(out);
fprintf(out, "Written by Francois Fleuret <francois@fleuret.org>.\n");
fprintf(out, "\n");
- fprintf(out, "Usage: mymail [options] [<mbox dir1> [<mbox dir2> ...]]\n");
+ fprintf(out, "Usage: mymail [options] [<mbox dir1> [<mbox dir2> ...]|<db file1> [<db file2> ...]]\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, " -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");
+ fprintf(out, " set the db root path for recursive search\n");
+ fprintf(out, " -l <db filename list>, --db-list <db filename list>\n");
+ fprintf(out, " set the semicolon-separated list of db files for search\n");
+ fprintf(out, " -m <mbox filename pattern>, --mbox-pattern <mbox filename pattern>\n");
+ fprintf(out, " set the mbox filename pattern for recursive search\n");
+ fprintf(out, " -s <search pattern>, --search <search pattern>\n");
+ fprintf(out, " search for matching mails in the db file\n");
+ fprintf(out, " -d <db filename>, --db-file-generate <db filename>\n");
+ fprintf(out, " set the db filename for indexing\n");
fprintf(out, " -i, --index\n");
fprintf(out, " index mails\n");
- fprintf(out, " -s <search pattern>, --search <search pattern>\n");
- fprintf(out, " search for matching mails in the data-base file\n");
- fprintf(out, " -d <db filename>, --db-file <db filename>\n");
- fprintf(out, " set the data-base file\n");
- fprintf(out, " -r <db root path>, --db-root <db root path>\n");
- fprintf(out, " set the data-base root path for recursive search\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");
}
/*********************************************************************/
int ignore_entry(const char *name) {
return
- /* strcmp(name, ".") == 0 || */
- /* strcmp(name, "..") == 0 || */
+ strcmp(name, ".") == 0 ||
+ strcmp(name, "..") == 0 ||
(name[0] == '.' && name[1] != '/');
}
-int mbox_line_match_search(struct search_request *request,
- int mbox_id, char *mbox_value) {
+int is_a_leading_from_line(char *mbox_line) {
return
- (request->field_id == mbox_id ||
- (request->field_id == ID_PARTICIPANT && (mbox_id == ID_FROM || mbox_id == ID_DEST)))
+ strncmp(mbox_line, "From ", 5) == 0 &&
+ regexec(&global_leading_from_line_regexp, mbox_line, 0, 0, 0) == 0;
+}
+
+int db_line_match_search(struct search_condition *condition,
+ int mbox_id, const char *mbox_value) {
+
+ 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(&request->regexp, mbox_value, 0, 0, 0) == 0;
+
+ regexec(&condition->regexp, mbox_value, 0, 0, 0) == 0;
}
-void search_in_db(int nb_search_patterns,
- struct search_request *search_requests,
- FILE *db_file) {
- int hits[MAX_NB_SEARCH_PATTERNS];
- char raw_db_line[BUFFER_SIZE];
+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;
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;
- int mbox_id;
- int already_written, m, n;
- int last_mbox_line_was_empty;
+ int nb_body_hits;
- current_position_in_mail = 0;
- already_written = 0;
+ nb_body_hits = 0;
- for(n = 0; n < nb_search_patterns; n++) { hits[n] = 0; }
+ header = 1;
+ mail_file = safe_fopen(mail_filename, "r", "mbox for body scan");
- while(fgets(raw_db_line, BUFFER_SIZE, db_file)) {
- mbox_name = raw_db_line;
- mbox_value = segment_next_field(raw_db_line);
+ fseek(mail_file, position_in_mail, SEEK_SET);
- if(strcmp("mail", mbox_name) == 0) {
- char *position_in_file_string;
- char *mail_filename;
+ 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; } */
- for(n = 0; n < nb_search_patterns && xor(hits[n], search_requests[n].negation); n++);
+ if(raw_mbox_line[0] == '\n') { header = 0; }
- /* for(n = 0; n < nb_search_patterns && */
- /* ((hits[n] && !search_requests[n].negation) || */
- /* (!hits[n] && search_requests[n].negation)); n++); */
+ 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(n == nb_search_patterns) {
- FILE *mail_file;
+ if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) ||
+ (is_a_leading_from_line(raw_mbox_line)))
+ break;
+ }
+ }
- mail_file = fopen(current_mail_filename, "r");
+ fclose(mail_file);
+}
- if(!mail_file) {
- fprintf(stderr, "mymail: Cannot open mbox '%s'.\n", current_mail_filename);
- exit(EXIT_FAILURE);
- }
+void extract_mail(const char *mail_filename, unsigned long int position_in_mail,
+ FILE *output_file) {
+ char raw_mbox_line[BUFFER_SIZE];
+ FILE *mail_file;
+
+ mail_file = safe_fopen(mail_filename, "r", "mbox for mail extraction");
+ 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))
+ )
+ break;
+ fprintf(output_file, "%s", raw_mbox_line);
+ }
+ }
- fseek(mail_file, current_position_in_mail, SEEK_SET);
+ fclose(mail_file);
+}
- if(fgets(raw_mbox_line, BUFFER_SIZE, mail_file)) {
- last_mbox_line_was_empty = 1;
- printf("%s", raw_mbox_line);
- while(1) {
- if(!fgets(raw_mbox_line, BUFFER_SIZE, mail_file) ||
- (last_mbox_line_was_empty && strncmp(raw_mbox_line, "From ", 5) == 0)) break;
- last_mbox_line_was_empty = (raw_mbox_line[0] == '\n');
- printf("%s", raw_mbox_line);
- }
+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 db_id, const char *db_value, time_t *t) {
+ const char *c;
+ struct tm tm;
- fclose(mail_file);
+ if(db_id == 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_id == 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);
+ }
}
+ }
+ }
+}
- for(n = 0; n < nb_search_patterns; n++) { hits[n] = 0; }
+int search_in_db(const char *db_filename,
+ int nb_search_conditions,
+ struct search_condition *search_conditions,
+ FILE *output_file) {
- position_in_file_string = mbox_value;
- mail_filename = segment_next_field(mbox_value);
- current_position_in_mail = atol(position_in_file_string);
- strcpy(current_mail_filename, mail_filename);
+ FILE *db_file;
+ char raw_db_line[BUFFER_SIZE];
+ char current_mail_filename[PATH_MAX + 1];
+ char db_key[TOKEN_BUFFER_SIZE];
+ char position_in_file_string[TOKEN_BUFFER_SIZE];
+ unsigned long int current_position_in_mail;
+ const char *db_value;
+ int db_id;
+ 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;
+
+ if(!global_quiet) {
+ printf("Searching in '%s' ... ", db_filename);
+ fflush(stdout);
+ }
- remove_eof(current_mail_filename);
- already_written = 0;
+ db_file = safe_fopen(db_filename, "r", "index file for search");
+
+ /* First, check the db file leading line integrity */
+
+ 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);
+ }
+
+ /* Then parse the said db file */
+
+ current_position_in_mail = 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) {
+ 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)) {
+ db_value = parse_token(db_key, TOKEN_BUFFER_SIZE, ' ', raw_db_line);
+
+ if(strcmp("mail", db_key) == 0) {
+ 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++;
+ }
+ }
+
+ for(n = 0; n < nb_search_conditions; n++) { hits[n] = 0; }
+ 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);
}
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_id = -1;
+ for(m = 0; (m < MAX_ID) && db_id == -1; m++) {
+ if(strncmp(field_keys[m], db_key, strlen(db_key)) == 0) {
+ db_id = m;
}
}
- for(n = 0; n < nb_search_patterns; n++) {
- hits[n] |= mbox_line_match_search(&search_requests[n],
- mbox_id, mbox_value);
+
+ for(n = 0; n < nb_search_conditions; n++) {
+ hits[n] |= db_line_match_search(&search_conditions[n],
+ db_id, db_value);
}
+
+ if(need_time) {
+ update_time(db_id, 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) {
+ printf("done.\n");
+ fflush(stdout);
+ }
+
+ return nb_extracted_mails;
}
-void recursive_search_in_db(const char *entry_name,
- int nb_search_patterns,
- struct search_request *search_requests) {
+int 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 raw_db_line[BUFFER_SIZE];
char subname[PATH_MAX + 1];
+ int nb_extracted_mails = 0;
if(lstat(entry_name, &sb) != 0) {
fprintf(stderr,
exit(EXIT_FAILURE);
}
+ /* printf("recursive_search_in_db %s\n", entry_name); */
+
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,
- nb_search_patterns,
- search_requests);
+ nb_extracted_mails += recursive_search_in_db(subname, db_filename_regexp,
+ nb_search_conditions, search_conditions,
+ output_file);
}
}
closedir(dir);
- } else {
+ }
+
+ else {
const char *s = entry_name, *filename = entry_name;
while(*s) { if(*s == '/') { filename = s+1; } s++; }
- if(strcmp(filename, db_filename) == 0) {
- FILE *db_file = fopen(entry_name, "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",
- 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(nb_search_patterns, search_requests, db_file);
-
- fclose(db_file);
+ 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);
}
}
+
+ return nb_extracted_mails;
}
/*********************************************************************/
-void index_one_mbox_line(int nb_fields_to_parse, struct parsable_field *fields_to_parse,
+void index_one_mbox_line(unsigned int nb_fields_to_parse,
+ struct parsable_field *fields_to_parse,
char *raw_mbox_line, FILE *db_file) {
regmatch_t matches;
- int f;
+ unsigned 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],
+ field_keys[fields_to_parse[f].id],
raw_mbox_line + matches.rm_eo);
}
}
char raw_mbox_line[BUFFER_SIZE], full_line[BUFFER_SIZE];
char *end_of_full_line;
FILE *file;
- int in_header, new_header, last_mbox_line_was_empty;
+ int in_header, new_header;
unsigned long int position_in_file;
- file = fopen(mbox_filename, "r");
-
- if(!file) {
- fprintf(stderr, "mymail: Cannot open '%s'.\n", mbox_filename);
- if(paranoid) { exit(EXIT_FAILURE); }
- return;
- }
+ file = safe_fopen(mbox_filename, "r", "mbox for indexing");
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_mbox_line, BUFFER_SIZE, file)) {
- if(last_mbox_line_was_empty && strncmp(raw_mbox_line, "From ", 5) == 0) {
+ 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';
+ }
}
- last_mbox_line_was_empty = (raw_mbox_line[0] == '\n');
-
if(in_header) {
if(new_header) {
fprintf(db_file, "mail %lu %s\n", position_in_file, 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++) = ' ';
}
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);
}
}
void recursive_index_mbox(FILE *db_file,
- const char *entry_name,
+ const char *entry_name, regex_t *mbox_filename_regexp,
int nb_fields_to_parse, struct parsable_field *fields_to_parse) {
DIR *dir;
struct dirent *dir_e;
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_index_mbox(db_file, subname, nb_fields_to_parse, fields_to_parse);
+ recursive_index_mbox(db_file, subname, mbox_filename_regexp,
+ nb_fields_to_parse, fields_to_parse);
}
}
closedir(dir);
} else {
- index_mbox(entry_name, nb_fields_to_parse, fields_to_parse, db_file);
+ const char *s = entry_name, *filename = s;
+ while(*s) { if(*s == '/') { filename = s+1; }; s++; }
+ if(!mbox_filename_regexp || regexec(mbox_filename_regexp, filename, 0, 0, 0) == 0) {
+ index_mbox(entry_name, nb_fields_to_parse, fields_to_parse, db_file);
+ }
}
}
static struct option long_options[] = {
{ "help", no_argument, 0, 'h' },
{ "version", no_argument, 0, 'v' },
- { "db-file", 1, 0, 'd' },
+ { "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' },
+ { "db-list", 1, 0, 'l' },
+ { "mbox-pattern", 1, 0, 'm' },
{ "search", 1, 0, 's' },
{ "index", 0, 0, 'i' },
+ { "output", 1, 0, 'o' },
+ { "default-search", 1, 0, 'a' },
{ 0, 0, 0, 0 }
};
+struct time_criterion {
+ char *label;
+ int day_criterion;
+ int start_hour, end_hour;
+ int past_week_day;
+};
+
+/*********************************************************************/
+
+static struct time_criterion time_criteria[] = {
+
+ { "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", 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);
+ 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);
+}
+
+void init_condition(struct search_condition *condition, const char *full_string,
+ const char *default_search_field) {
+ char full_search_field[TOKEN_BUFFER_SIZE], *search_field;
+ unsigned int k, m;
+ const 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;
+ }
+
+ condition->field_id = -1;
+
+ /* 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_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;
+ }
+ }
+ break;
+ }
+ }
+
+ if(condition->field_id == -1) {
+
+ /* No time condition matched, look for the search fields */
+
+ for(m = 0; (m < MAX_ID) && condition->field_id == -1; m++) {
+ if(strncmp(field_keys[m], search_field, strlen(search_field)) == 0) {
+ condition->field_id = m;
+ }
+ }
+
+ /* None match, if there is a default search field, re-run the search with it */
+
+ if(condition->field_id == -1) {
+ if(default_search_field) {
+ for(m = 0; (m < MAX_ID) && condition->field_id == -1; 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 key \"%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_keys[condition->field_id]);
+ exit(EXIT_FAILURE);
+ }
+ }
+}
+
+void free_condition(struct search_condition *condition) {
+ if(condition->field_id != ID_TIME_INTERVAL) {
+ regfree(&condition->regexp);
+ }
+}
+
+/*********************************************************************/
+/*********************************************************************/
/*********************************************************************/
int main(int argc, char **argv) {
+ char *db_filename = 0;
+ char *db_filename_regexp_string = 0;
+ char *db_root_path = 0;
+ char *db_filename_list = 0;
+ char *mbox_filename_regexp_string = 0;
+ char *default_search_field;
+ char output_filename[PATH_MAX + 1];
+ int action_index = 0;
int error = 0, show_help = 0;
- const int nb_fields_to_parse = sizeof(fields_to_parse) / sizeof(struct parsable_field);
+ const unsigned int nb_fields_to_parse =
+ sizeof(fields_to_parse) / sizeof(struct parsable_field);
char c;
- int f;
- int nb_search_patterns;
- char *search_pattern[MAX_NB_SEARCH_PATTERNS];
+ unsigned int f, n;
+ unsigned int nb_search_conditions;
+ struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS];
- /* for(f = 0; f < argc; f++) { */
- /* printf("arg %d \"%s\"\n", f, argv[f]); */
- /* } */
+ 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;
- action_index = 0;
- db_filename = 0;
- db_root_path = 0;
+ global_quiet = 0;
+ global_use_leading_time = 0;
+ default_search_field = 0;
+ strncpy(output_filename, "", PATH_MAX);
setlocale(LC_ALL, "");
- nb_search_patterns = 0;
+ nb_search_conditions = 0;
- while ((c = getopt_long(argc, argv, "hvip:s:d:r:",
+ while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:a:m:",
long_options, NULL)) != -1) {
switch(c) {
print_version(stdout);
break;
+ case 'q':
+ global_quiet = 1;
+ break;
+
+ case 't':
+ global_use_leading_time = 1;
+ break;
+
case 'i':
action_index = 1;
break;
case 'd':
+ if(db_filename) {
+ fprintf(stderr, "mymail: Can not set the db filename twice.\n");
+ exit(EXIT_FAILURE);
+ }
db_filename = strdup(optarg);
break;
+ case 'p':
+ if(db_filename_regexp_string) {
+ fprintf(stderr, "mymail: Can not set the db filename pattern twice.\n");
+ exit(EXIT_FAILURE);
+ }
+ db_filename_regexp_string = strdup(optarg);
+ break;
+
+ case 'm':
+ if(mbox_filename_regexp_string) {
+ fprintf(stderr, "mymail: Can not set the mbox filename pattern twice.\n");
+ exit(EXIT_FAILURE);
+ }
+ mbox_filename_regexp_string = strdup(optarg);
+ break;
+
+ case 'o':
+ strncpy(output_filename, optarg, PATH_MAX);
+ break;
+
case 'r':
+ if(db_root_path) {
+ fprintf(stderr, "mymail: Can not set the db root path twice.\n");
+ exit(EXIT_FAILURE);
+ }
db_root_path = strdup(optarg);
break;
+ case 'l':
+ if(db_filename_list) {
+ fprintf(stderr, "mymail: Can not set the db filename list twice.\n");
+ exit(EXIT_FAILURE);
+ }
+ db_filename_list = strdup(optarg);
+ break;
+
case 's':
- if(nb_search_patterns == MAX_NB_SEARCH_PATTERNS) {
+ if(nb_search_conditions == MAX_NB_SEARCH_CONDITIONS) {
fprintf(stderr, "mymail: Too many search patterns.\n");
exit(EXIT_FAILURE);
}
- search_pattern[nb_search_patterns++] = strdup(optarg);
+ init_condition(&search_conditions[nb_search_conditions], optarg, default_search_field);
+ nb_search_conditions++;
+ break;
+
+ case 'a':
+ default_search_field = optarg;
break;
default:
}
}
- if(!db_filename) {
- char *default_db_filename = getenv("MYMAIL_DB_FILE");
+ /* Set all the values that may defined in the arguments, through
+ environment variables, or hard-coded */
- if(!default_db_filename) {
- default_db_filename = "mymail.db";
- }
+ db_filename = default_value(db_filename,
+ "MYMAIL_DB_FILE",
+ "mymail.db");
- db_filename = strdup(default_db_filename);
- }
+ db_filename_regexp_string = default_value(db_filename_regexp_string,
+ "MYMAIL_DB_FILE",
+ "\\.db$");
- if(!db_root_path) {
- char *default_db_root_path = getenv("MYMAIL_DB_ROOT");
+ db_root_path = default_value(db_root_path,
+ "MYMAIL_DB_ROOT",
+ 0);
- if(default_db_root_path) {
- db_root_path = strdup(default_db_root_path);
- }
- }
+ db_filename_list = default_value(db_filename_list,
+ "MYMAIL_DB_LIST",
+ 0);
+
+ mbox_filename_regexp_string = default_value(mbox_filename_regexp_string,
+ "MYMAIL_MBOX_PATTERN",
+ 0);
+
+ /* Start the processing */
if(error) {
print_usage(stderr);
exit(EXIT_SUCCESS);
}
+ /* mbox indexing */
+
if(action_index) {
FILE *db_file;
+ regex_t mbox_filename_regexp_static;
+ regex_t *mbox_filename_regexp;
- db_file = fopen(db_filename, "w");
-
- if(!db_file) {
- fprintf(stderr,
- "mymail: Cannot open \"%s\" for writing: %s\n",
- db_filename,
- strerror(errno));
- exit(EXIT_FAILURE);
+ if(mbox_filename_regexp_string) {
+ if(regcomp(&mbox_filename_regexp_static,
+ mbox_filename_regexp_string,
+ 0)) {
+ fprintf(stderr,
+ "mymail: Syntax error in regexp \"%s\".\n",
+ mbox_filename_regexp_string);
+ exit(EXIT_FAILURE);
+ }
+ mbox_filename_regexp = &mbox_filename_regexp_static;
+ } else {
+ mbox_filename_regexp = 0;
}
+ db_file = safe_fopen(db_filename, "w", "index file for indexing");
+
for(f = 0; f < nb_fields_to_parse; f++) {
if(regcomp(&fields_to_parse[f].regexp,
fields_to_parse[f].regexp_string,
- REG_ICASE)) {
+ fields_to_parse[f].cflags)) {
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);
}
}
while(optind < argc) {
recursive_index_mbox(db_file,
- argv[optind],
+ argv[optind], mbox_filename_regexp,
nb_fields_to_parse, fields_to_parse);
optind++;
}
+ fflush(db_file);
fclose(db_file);
+ if(mbox_filename_regexp) {
+ regfree(mbox_filename_regexp);
+ }
+
for(f = 0; f < nb_fields_to_parse; f++) {
regfree(&fields_to_parse[f].regexp);
}
}
+ /* Mail search */
+
else {
- if(!db_root_path) {
- fprintf(stderr,
- "mymail: db root path is not set\n");
- exit(EXIT_FAILURE);
+ FILE *output_file;
+ int nb_extracted_mails = 0;
+
+ if(output_filename[0]) {
+ output_file = safe_fopen(output_filename, "w", "result mbox");
+ } else {
+ output_file = stdout;
+ global_quiet = 1;
}
- if(nb_search_patterns > 0) {
- struct search_request search_requests[MAX_NB_SEARCH_PATTERNS];
- char *search_regexp_string;
- int m, n;
+ if(nb_search_conditions > 0) {
- for(n = 0; n < nb_search_patterns; n++) {
- search_regexp_string = segment_next_field(search_pattern[n]);
+ /* Recursive search if db_root_path is set */
- if(search_pattern[n][0] == '!') {
- search_pattern[n]++;
- search_requests[n].negation = 1;
- } else {
- search_requests[n].negation = 0;
+ 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);
}
- search_requests[n].field_id = -1;
- for(m = 0; (m < MAX_ID) && search_requests[n].field_id == -1; m++) {
- if(strncmp(field_names[m], search_pattern[n], strlen(search_pattern[n])) == 0) {
- search_requests[n].field_id = m;
- }
- }
+ nb_extracted_mails += recursive_search_in_db(db_root_path, &db_filename_regexp,
+ nb_search_conditions, search_conditions,
+ output_file);
- if(regcomp(&search_requests[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_requests[n].field_id]);
- exit(EXIT_FAILURE);
+ regfree(&db_filename_regexp);
+ }
+
+ /* Search in all db files listed in db_filename_list */
+
+ if(db_filename_list) {
+ char db_filename[PATH_MAX + 1];
+ const char *s;
+
+ s = db_filename_list;
+
+ while(*s) {
+ 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);
+ }
}
}
- recursive_search_in_db(db_root_path,
- nb_search_patterns, search_requests);
+ /* Search in all db files listed in the command arguments */
- for(n = 0; n < nb_search_patterns; n++) {
- free(search_pattern[n]);
+ while(optind < argc) {
+ nb_extracted_mails +=
+ search_in_db(argv[optind], nb_search_conditions, search_conditions, output_file);
+ optind++;
}
}
+
+ if(!global_quiet) {
+ if(nb_extracted_mails > 0) {
+ printf("Found %d matching mails.\n", nb_extracted_mails);
+ } else {
+ printf("No matching mail found.\n");
+ }
+ }
+
+ fflush(output_file);
+
+ if(output_file != stdout) {
+ fclose(output_file);
+ }
+ }
+
+ for(n = 0; n < nb_search_conditions; n++) {
+ free_condition(&search_conditions[n]);
}
free(db_filename);
+ free(db_filename_regexp_string);
free(db_root_path);
+ free(db_filename_list);
+ free(mbox_filename_regexp_string);
+
+ regfree(&global_leading_from_line_regexp);
exit(EXIT_SUCCESS);
}