#include <time.h>
#define MYMAIL_DB_MAGIC_TOKEN "mymail_index_file"
-#define VERSION "0.9.2"
+#define VERSION "0.9.5"
#define MAX_NB_SEARCH_CONDITIONS 32
#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$"
-regex_t leading_from_line_regexp;
-
/* Global variables! */
-int paranoid;
-int quiet;
-char *default_search_field;
-int ignore_dot_files;
+int global_quiet;
+int global_use_leading_time;
+
+regex_t global_leading_from_line_regexp;
/********************************************************************/
MAX_ID
};
-static char *field_names[] = {
+static char *field_keys[] = {
"mail",
"lead",
"from",
return (a && !b) || (!a && b);
}
-char *parse_token(char *token_buffer, size_t token_buffer_size,
- char separator, char *string) {
+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++;
+ while(*string == separator) { string++; }
*u = '\0';
return string;
}
}
}
-/********************************************************************/
-
-/* 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;
}
/*********************************************************************/
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");
/*********************************************************************/
-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 ||
- (ignore_dot_files && name[0] == '.' && name[1] != '/');
+ (name[0] == '.' && name[1] != '/');
}
int is_a_leading_from_line(char *mbox_line) {
return
-
- /*
-
- The mbox man page in qmail documentation states:
-
- > The reader should not attempt to take advantage of the fact
- > that every From_ line (past the beginning of the file) is
- > preceded by a blank line.
-
- */
-
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, char *mbox_value) {
+ int mbox_id, const 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));
+ const char *c;
+ time_t t;
+ struct tm tm;
+ if(global_use_leading_time) {
+ if(mbox_id == ID_LEADING_LINE) {
+ c = mbox_value;
+ while(*c && *c != ' ') c++; while(*c && *c == ' ') c++;
+ strptime(c, "%a %b %e %k:%M:%S %Y", &tm);
+ t = mktime(&tm);
+ return (t >= condition->interval_start &&
+ (condition->interval_stop == 0 ||
+ t <= condition->interval_stop));
+ } else {
+ return 0;
+ }
} else {
- return 0;
+ if(mbox_id == ID_DATE) {
+ strptime(mbox_value, "%a, %d %b %Y %k:%M:%S", &tm);
+ t = mktime(&tm);
+ return (t >= condition->interval_start &&
+ (condition->interval_stop == 0 ||
+ t <= condition->interval_stop));
+ } else {
+ return 0;
+ }
}
} else {
return
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);
- }
+ mail_file = safe_fopen(mail_filename, "r", "mbox for body scan");
fseek(mail_file, position_in_mail, SEEK_SET);
char raw_mbox_line[BUFFER_SIZE];
FILE *mail_file;
- mail_file = fopen(mail_filename, "r");
-
- if(!mail_file) {
- fprintf(stderr,
- "mymail: Cannot open mbox '%s' for mail extraction.\n",
- mail_filename);
- exit(EXIT_FAILURE);
- }
-
+ 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)) {
char raw_db_line[BUFFER_SIZE];
char current_mail_filename[PATH_MAX + 1];
unsigned long int current_position_in_mail;
- char mbox_name[TOKEN_BUFFER_SIZE], *mbox_value;
+ char mbox_key[TOKEN_BUFFER_SIZE];
+ const char *mbox_value;
int mbox_id;
int already_written, m, n;
int nb_body_conditions, nb_fulfilled_body_conditions;
nb_extracted_mails = 0;
- if(!quiet) {
+ if(!global_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);
- }
+ db_file = safe_fopen(db_filename, "r", "index file for search");
/* First, check the db file leading line integrity */
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);
+ mbox_value = parse_token(mbox_key, TOKEN_BUFFER_SIZE, ' ', raw_db_line);
- if(strcmp("mail", mbox_name) == 0) {
+ if(strcmp("mail", mbox_key) == 0) {
char position_in_file_string[TOKEN_BUFFER_SIZE];
if(current_mail_filename[0]) {
/* 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);
- }
- 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++;
+ 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++;
+ }
}
}
else {
mbox_id = -1;
for(m = 0; (m < MAX_ID) && mbox_id == -1; m++) {
- if(strncmp(field_names[m], mbox_name, strlen(mbox_name)) == 0) {
+ if(strncmp(field_keys[m], mbox_key, strlen(mbox_key)) == 0) {
mbox_id = m;
}
}
fclose(db_file);
- if(!quiet) {
+ if(!global_quiet) {
printf("done.\n");
fflush(stdout);
}
exit(EXIT_FAILURE);
}
+ /* printf("recursive_search_in_db %s\n", entry_name); */
+
dir = opendir(entry_name);
if(dir) {
/*********************************************************************/
-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);
}
}
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;
while(fgets(raw_mbox_line, BUFFER_SIZE, file)) {
if(is_a_leading_from_line(raw_mbox_line)) {
+ /* This starts a new mail */
if(in_header) {
fprintf(stderr,
"Got a ^\"From \" in the header in %s:%lu.\n",
mbox_filename, position_in_file);
fprintf(stderr, "%s", raw_mbox_line);
- if(paranoid) { exit(EXIT_FAILURE); }
}
+
+ /* printf("LEADING_LINE %s", raw_mbox_line); */
+
in_header = 1;
new_header = 1;
} else if(raw_mbox_line[0] == '\n') {
- if(in_header) { in_header = 0; }
+ if(in_header) {
+ in_header = 0;
+ /* We leave the header, index the current line */
+ if(full_line[0]) {
+ /* printf("INDEX %s\n", full_line); */
+ index_one_mbox_line(nb_fields_to_parse, fields_to_parse, full_line, db_file);
+ }
+ end_of_full_line = full_line;
+ *end_of_full_line = '\0';
+ }
}
if(in_header) {
}
if(raw_mbox_line[0] == ' ' || raw_mbox_line[0] == '\t') {
+ /* Continuation of a line */
char *start = raw_mbox_line;
while(*start == ' ' || *start == '\t') start++;
*(end_of_full_line++) = ' ';
}
else {
- /*
- if(!((raw_mbox_line[0] >= 'a' && raw_mbox_line[0] <= 'z') ||
- (raw_mbox_line[0] >= 'A' && raw_mbox_line[0] <= 'Z'))) {
- fprintf(stderr,
- "Header line syntax error %s:%lu.\n",
- mbox_filename, position_in_file);
- fprintf(stderr, "%s", raw_mbox_line);
- }
- */
+ /* Start a new header line, not a continuation */
if(full_line[0]) {
+ /* printf("INDEX %s\n", full_line); */
index_one_mbox_line(nb_fields_to_parse, fields_to_parse, full_line, db_file);
}
{ "help", no_argument, 0, 'h' },
{ "version", no_argument, 0, 'v' },
{ "quiet", no_argument, 0, 'q' },
+ { "use-leading-time", no_argument, 0, 't' },
{ "db-file-generate", 1, 0, 'd' },
{ "db-pattern", 1, 0, 'p' },
{ "db-root", 1, 0, 'r' },
{ 0, 0, 0, 0 }
};
+struct time_criterion {
+ char *label;
+ int start_hour, end_hour;
+ int past_week_day;
+};
+
/*********************************************************************/
-void init_condition(struct search_condition *condition, char *full_string) {
+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 },
+
+};
+
+/*********************************************************************/
+
+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);
+}
+
+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;
- int m;
- char *string;
+ unsigned int k, m;
+ const char *string;
string = parse_token(full_search_field, TOKEN_BUFFER_SIZE, ' ', full_string);
search_field = full_search_field;
condition->negation = 0;
}
- /* Recently */
-
- if(strcmp(search_field, "8h") == 0) {
- condition->field_id = ID_INTERVAL;
- condition->interval_start = time(0) - 3600 * 8;
- condition->interval_stop = 0;
- }
-
- 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, "week") == 0) {
- condition->field_id = ID_INTERVAL;
- condition->interval_start = time(0) - 3600 * 24 * 7;
- condition->interval_stop = 0;
- }
-
- else if(strcmp(search_field, "month") == 0) {
- condition->field_id = ID_INTERVAL;
- condition->interval_start = time(0) - 3600 * 24 * 31;
- condition->interval_stop = 0;
- }
-
- else if(strcmp(search_field, "year") == 0) {
- condition->field_id = ID_INTERVAL;
- condition->interval_start = time(0) - 3600 * 24 * 365;
- condition->interval_stop = 0;
- }
-
- /* Yesterday */
-
- 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 */
+ condition->field_id = -1;
- 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;
- }
+ /* Recently */
- 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;
+ 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;
+ if(time_criteria[k].end_hour >= 0) {
+ condition->interval_stop = time(0) - 3600 * time_criteria[k].end_hour;
+ } else {
+ condition->interval_stop = 0;
+ }
+ } else {
+ condition->interval_start = time_for_past_day(time_criteria[k].past_week_day);
+ condition->interval_stop = condition->interval_start + 3600 * 24;
+ }
+ break;
+ }
}
- else {
-
- /* header-related conditions */
+ if(condition->field_id == -1) {
- 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_names[m], search_field, strlen(search_field)) == 0) {
+ 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_names[m],
+ if(strncmp(field_keys[m],
default_search_field, strlen(default_search_field)) == 0) {
condition->field_id = m;
}
}
string = full_string;
+ if(string[0] == '!') { string++; }
}
}
if(condition->field_id == -1) {
fprintf(stderr,
- "mymail: Syntax error in field name \"%s\".\n",
+ "mymail: Syntax error in field key \"%s\".\n",
search_field);
exit(EXIT_FAILURE);
}
fprintf(stderr,
"mymail: Syntax error in regexp \"%s\" for field \"%s\".\n",
string,
- field_names[condition->field_id]);
+ field_keys[condition->field_id]);
exit(EXIT_FAILURE);
}
}
/*********************************************************************/
int main(int argc, char **argv) {
- char *db_filename;
- char *db_filename_regexp_string;
- char *db_root_path;
- char *db_filename_list;
- char *mbox_filename_regexp_string;
+ 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;
+ 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, n;
- int nb_search_conditions;
+ unsigned int f, n;
+ unsigned int nb_search_conditions;
struct search_condition search_conditions[MAX_NB_SEARCH_CONDITIONS];
- if(regcomp(&leading_from_line_regexp, LEADING_FROM_LINE_REGEXP_STRING, 0)) {
+ if(regcomp(&global_leading_from_line_regexp, LEADING_FROM_LINE_REGEXP_STRING, 0)) {
fprintf(stderr,
"mymail: Cannot compile leading \"from\" line regexp. That is strange.\n");
exit(EXIT_FAILURE);
}
- paranoid = 0;
- action_index = 0;
- db_filename = 0;
- db_filename_regexp_string = 0;
- db_root_path = 0;
- db_filename_list = 0;
- mbox_filename_regexp_string = 0;
- quiet = 0;
+ global_quiet = 0;
+ global_use_leading_time = 0;
default_search_field = 0;
- ignore_dot_files = 1;
+ strncpy(output_filename, "", PATH_MAX);
setlocale(LC_ALL, "");
nb_search_conditions = 0;
- while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:a:",
+ while ((c = getopt_long(argc, argv, "hvqip:s:d:r:l:o:a:m:",
long_options, NULL)) != -1) {
switch(c) {
break;
case 'q':
- quiet = 1;
+ global_quiet = 1;
+ break;
+
+ case 't':
+ global_use_leading_time = 1;
break;
case 'i':
fprintf(stderr, "mymail: Too many search patterns.\n");
exit(EXIT_FAILURE);
}
- init_condition(&search_conditions[nb_search_conditions], optarg);
+ init_condition(&search_conditions[nb_search_conditions], optarg, default_search_field);
nb_search_conditions++;
break;
db_filename_regexp_string = default_value(db_filename_regexp_string,
"MYMAIL_DB_FILE",
- "^mymail.db$");
+ "\\.db$");
db_root_path = default_value(db_root_path,
"MYMAIL_DB_ROOT",
mbox_filename_regexp_string = default_value(mbox_filename_regexp_string,
"MYMAIL_MBOX_PATTERN",
- "mbox$");
+ 0);
/* Start the processing */
mbox_filename_regexp = 0;
}
- 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);
- }
+ 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,
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);
}
}
int nb_extracted_mails = 0;
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);
- }
+ output_file = safe_fopen(output_filename, "w", "result mbox");
} else {
output_file = stdout;
- quiet = 1;
+ global_quiet = 1;
}
if(nb_search_conditions > 0) {
if(db_filename_list) {
char db_filename[PATH_MAX + 1];
- char *s;
+ const char *s;
s = db_filename_list;
}
}
- if(!quiet) {
+ if(!global_quiet) {
if(nb_extracted_mails > 0) {
printf("Found %d matching mails.\n", nb_extracted_mails);
} else {
free(db_filename_list);
free(mbox_filename_regexp_string);
- regfree(&leading_from_line_regexp);
+ regfree(&global_leading_from_line_regexp);
exit(EXIT_SUCCESS);
}