static int
parse_external_byte (RIP_MANAGER_INFO* rmi, External_Process* ep,
TRACK_INFO* ti, char c)
{
int got_metadata = 0;
if (c != '\r' && c != '\n') {
if (ep->line_buf_idx < MAX_EXT_LINE_LEN-1) {
ep->line_buf[ep->line_buf_idx++] = c;
ep->line_buf[ep->line_buf_idx] = 0;
}
} else {
if (!strcmp (".",ep->line_buf)) {
/* Found end of record! */
mchar tmp_raw_metadata[MAX_TRACK_LEN];
gstring_from_string (rmi, ti->artist, MAX_TRACK_LEN, ep->artist_buf,
CODESET_METADATA);
gstring_from_string (rmi, ti->album, MAX_TRACK_LEN, ep->album_buf,
CODESET_METADATA);
gstring_from_string (rmi, ti->title, MAX_TRACK_LEN, ep->title_buf,
CODESET_METADATA);
g_snprintf (tmp_raw_metadata, MAX_TRACK_LEN, "%s - %s",
ti->artist, ti->title);
string_from_gstring (rmi, ti->raw_metadata, MAX_TRACK_LEN,
tmp_raw_metadata, CODESET_METADATA);
ti->have_track_info = 1;
ti->save_track = TRUE;
ep->artist_buf[0] = 0;
ep->album_buf[0] = 0;
ep->title_buf[0] = 0;
got_metadata = 1;
} else if (!strncmp ("ARTIST=", ep->line_buf, strlen("ARTIST="))) {
strcpy (ep->artist_buf, &ep->line_buf[strlen("ARTIST=")]);
} else if (!strncmp ("ALBUM=", ep->line_buf, strlen("ALBUM="))) {
strcpy (ep->album_buf, &ep->line_buf[strlen("ALBUM=")]);
} else if (!strncmp ("TITLE=", ep->line_buf, strlen("TITLE="))) {
strcpy (ep->title_buf, &ep->line_buf[strlen("TITLE=")]);
}
ep->line_buf[0] = 0;
ep->line_buf_idx = 0;
}
return got_metadata;
}
/* This mega-function reads in the rules file, and loads
all the rules into the rmi->parse_rules data structure */
void init_metadata_parser(RIP_MANAGER_INFO * rmi, char *rules_file)
{
FILE *fp;
int ri; /* Rule index */
int rn; /* Number of rules allocated */
if (!rules_file || !*rules_file) {
use_default_rules(rmi);
return;
}
fp = fopen(rules_file, "r");
if (!fp) {
use_default_rules(rmi);
return;
}
rmi->parse_rules = 0;
ri = rn = 0;
while (1) {
char rule_buf[MAX_RULE_SIZE];
char match_buf[MAX_RULE_SIZE];
char subst_buf[MAX_RULE_SIZE];
mchar w_match_buf[MAX_RULE_SIZE];
mchar w_subst_buf[MAX_RULE_SIZE];
char *rbp;
char *rp;
int got_command;
int rc;
/* Allocate memory for rule, if necessary. */
/* If there are no more rules in the file, */
/* this rule will become the sentinel null rule */
if (ri + 1 != rn) {
rmi->parse_rules = realloc(rmi->parse_rules, (ri + 1) * sizeof(Parse_Rule));
memset(&rmi->parse_rules[ri], 0, sizeof(Parse_Rule));
rn = ri + 1;
}
/* Get next line from file */
rp = fgets(rule_buf, 2048, fp);
if (!rp)
break;
/* Skip leading whitespace */
rbp = rule_buf;
while (*rbp && isspace(*rbp))
rbp++;
if (!*rbp)
continue;
/* Get command */
got_command = 0;
switch (*rbp++) {
case 'm':
got_command = 1;
rmi->parse_rules[ri].cmd = PARSERULE_CMD_MATCH;
break;
case 's':
got_command = 1;
rmi->parse_rules[ri].cmd = PARSERULE_CMD_SUBST;
break;
case '#':
got_command = 0;
break;
default:
got_command = 0;
printf("Warning: malformed command in rules file:\n%s\n", rule_buf);
break;
}
if (!got_command)
continue;
/* Skip past fwd slash */
if (*rbp++ != '/') {
printf("Warning: malformed command in rules file:\n%s\n", rule_buf);
continue;
}
/* Parse match string */
rbp = parse_escaped_string(match_buf, rbp);
debug_printf("match_buf=%s\n", match_buf);
if (!rbp) {
printf("Warning: malformed command in rules file:\n%s\n", rule_buf);
continue;
}
/* Parse subst string */
if (rmi->parse_rules[ri].cmd == PARSERULE_CMD_SUBST) {
rbp = parse_escaped_string(subst_buf, rbp);
debug_printf("subst_buf=%s\n", subst_buf);
if (!rbp) {
printf("Warning: malformed command in rules file:\n%s\n", rule_buf);
continue;
}
}
/* Parse flags */
rc = parse_flags(&rmi->parse_rules[ri], rbp);
if (!rc) {
printf("Warning: malformed command in rules file:\n%s\n", rule_buf);
continue;
}
/* Compile the rule */
debug_printf("Compiling the rule\n");
gstring_from_string(rmi, w_match_buf, MAX_RULE_SIZE, match_buf, CODESET_UTF8);
if (!compile_rule(&rmi->parse_rules[ri], w_match_buf)) {
printf("Warning: malformed regular expression:\n%s\n", match_buf);
continue;
}
/* Copy rule strings */
debug_printf("Copying rule string (1)\n");
debug_mprintf(m_("String is ") m_S m_("\n"), w_match_buf);
rmi->parse_rules[ri].match = mstrdup(w_match_buf);
debug_printf("Copying rule string (2)\n");
if (rmi->parse_rules[ri].cmd == PARSERULE_CMD_SUBST) {
debug_printf("Copying rule string (3)\n");
gstring_from_string(rmi, w_subst_buf, MAX_RULE_SIZE, subst_buf, CODESET_UTF8);
debug_printf("Copying rule string (4)\n");
rmi->parse_rules[ri].subst = mstrdup(w_subst_buf);
debug_printf("Copying rule string (5)\n");
}
debug_printf("End of loop\n");
ri++;
}
fclose(fp);
}
void parse_metadata(RIP_MANAGER_INFO * rmi, TRACK_INFO * ti)
{
int i;
int eflags;
int rc;
int matched;
mchar query_string[MAX_TRACK_LEN];
Parse_Rule *rulep;
/* Has any m/.../s rule matched? */
BOOL save_track_matched = FALSE;
/* Has any m/.../x rule matched? */
BOOL exclude_track_matched = FALSE;
ti->artist[0] = 0;
ti->title[0] = 0;
ti->album[0] = 0;
ti->composed_metadata[0] = 0;
ti->save_track = TRUE;
/* Loop through rules, if we find a matching rule, then use it */
/* For now, only default rules supported with ascii
regular expressions. */
debug_printf("Converting query string to wide\n");
gstring_from_string(rmi, query_string, MAX_TRACK_LEN, ti->raw_metadata, CODESET_METADATA);
for (rulep = rmi->parse_rules; rulep->cmd; rulep++) {
#if !defined (USE_GLIB_REGEX)
regmatch_t pmatch[MAX_SUBMATCHES + 1];
#endif
eflags = 0;
if (rulep->cmd == PARSERULE_CMD_MATCH) {
debug_mprintf(m_("Testing match rule: ") m_S m_(" vs. ") m_S m_("\n"),
query_string, rulep->match);
if (rulep->flags & PARSERULE_SKIP) {
#if defined (USE_GLIB_REGEX)
rc = g_regex_match(rulep->reg, query_string, 0, NULL);
matched = rc;
#else
rc = mregexec(rulep->reg, query_string, 0, NULL, eflags);
matched = !rc;
#endif
if (!matched) {
continue;
}
/* GCS FIX: We need to return to the
caller that the metadata should be dropped. */
debug_printf("Skip rule matched\n");
ti->save_track = FALSE;
ti->have_track_info = 0;
return;
} else if (rulep->flags & PARSERULE_SAVE) {
#if defined (USE_GLIB_REGEX)
rc = g_regex_match(rulep->reg, query_string, 0, NULL);
matched = rc;
#else
rc = mregexec(rulep->reg, query_string, 0, NULL, eflags);
matched = !rc;
#endif
if (!matched) {
if (!save_track_matched)
ti->save_track = FALSE;
continue;
}
if (!exclude_track_matched) {
ti->save_track = TRUE;
save_track_matched = TRUE;
}
} else if (rulep->flags & PARSERULE_EXCLUDE) {
#if defined (USE_GLIB_REGEX)
rc = g_regex_match(rulep->reg, query_string, 0, NULL);
matched = rc;
#else
rc = mregexec(rulep->reg, query_string, 0, NULL, eflags);
matched = !rc;
#endif
if (matched && !save_track_matched) {
/* Rule matched => Exclude track */
ti->save_track = FALSE;
exclude_track_matched = TRUE;
}
} else {
#if defined (USE_GLIB_REGEX)
GMatchInfo *match_info;
gint nmatch;
rc = g_regex_match(rulep->reg, query_string, 0, &match_info);
if (rc == 0) {
/* Didn't match rule. */
continue;
}
nmatch = g_match_info_get_match_count(match_info);
debug_printf("Got %d matches\n", nmatch);
for (i = 0; i < nmatch; i++) {
gchar *match = g_match_info_fetch(match_info, i);
debug_printf("[%d] = %s\n", i, match);
g_free(match);
}
copy_rule_result(ti->artist, match_info, rulep->artist_idx);
copy_rule_result(ti->title, match_info, rulep->title_idx);
copy_rule_result(ti->album, match_info, rulep->album_idx);
copy_rule_result(ti->track_p, match_info, rulep->trackno_idx);
copy_rule_result(ti->year, match_info, rulep->year_idx);
g_match_info_free(match_info);
#else
eflags = 0;
rc = mregexec(rulep->reg, query_string, MAX_SUBMATCHES + 1, pmatch, eflags);
if (rc != 0) {
/* Didn't match rule. */
continue;
}
for (i = 0; i < MAX_SUBMATCHES + 1; i++) {
debug_printf("pmatch[%d]: (so,eo) = (%d,%d)\n", i,
pmatch[i].rm_so, pmatch[i].rm_eo);
}
copy_rule_result(ti->artist, query_string, pmatch, rulep->artist_idx);
copy_rule_result(ti->title, query_string, pmatch, rulep->title_idx);
copy_rule_result(ti->album, query_string, pmatch, rulep->album_idx);
copy_rule_result(ti->track_p, query_string, pmatch, rulep->trackno_idx);
copy_rule_result(ti->year, query_string, pmatch, rulep->year_idx);
#endif
ti->have_track_info = 1;
compose_metadata(rmi, ti);
debug_mprintf(m_("Parsed track info.\n")
m_("ARTIST: ") m_S m_("\n")
m_("TITLE: ") m_S m_("\n")
m_("ALBUM: ") m_S m_("\n")
m_("TRACK: ") m_S m_("\n")
m_("YEAR: ") m_S m_("\n"),
ti->artist, ti->title, ti->album, ti->track_p, ti->year);
return;
}
} else if (rulep->cmd == PARSERULE_CMD_SUBST) {
#if defined (USE_GLIB_REGEX)
GMatchInfo *match_info;
gint start_pos, end_pos;
gchar *tmp, *subst_string;
debug_mprintf(m_("Testing subst rule: ") m_S m_(" vs. ") m_S m_("\n"),
query_string, rulep->match);
rc = g_regex_match(rulep->reg, query_string, 0, &match_info);
if (rc == 0) {
/* Didn't match rule. */
continue;
}
rc = g_match_info_fetch_pos(match_info, 0, &start_pos, &end_pos);
if (!rc) {
debug_printf("g_match_info_fetch_pos returned 0\n");
g_match_info_free(match_info);
continue;
}
debug_printf("Matched at (%d,%d)\n", start_pos, end_pos);
if (start_pos == -1) {
g_match_info_free(match_info);
continue;
}
tmp = g_strndup(query_string, start_pos);
tmp[start_pos] = 0;
subst_string = g_strconcat(tmp, rulep->subst, &tmp[end_pos], NULL);
g_free(tmp);
g_match_info_free(match_info);
mstrncpy(query_string, subst_string, MAX_TRACK_LEN);
#else
mchar subst_string[MAX_TRACK_LEN];
int used, left;
debug_mprintf(m_("Testing subst rule: ") m_S m_(" vs. ") m_S m_("\n"),
query_string, rulep->match);
rc = mregexec(rulep->reg, query_string, 1, pmatch, eflags);
if (rc != 0) {
/* Didn't match rule. */
continue;
}
/* Update the query string and continue. */
debug_printf("Matched at (%d,%d)\n", pmatch[0].rm_so, pmatch[0].rm_eo);
mstrncpy(subst_string, query_string, pmatch[0].rm_so + 1);
debug_mprintf(m_("(1) subst_string = ") m_S m_("\n"), subst_string);
used = pmatch[0].rm_so;
left = MAX_TRACK_LEN - used;
mstrncpy(subst_string + used, rulep->subst, left);
debug_mprintf(m_("(2) subst_string = ") m_S m_("\n"), subst_string);
used += mstrlen(rulep->subst);
left = MAX_TRACK_LEN - used;
mstrncpy(subst_string + used, query_string + pmatch[0].rm_eo, left);
debug_mprintf(m_("(3) subst_string = ") m_S m_("\n"), subst_string);
mstrncpy(query_string, subst_string, MAX_TRACK_LEN);
debug_mprintf(m_("(4) query_string = ") m_S m_("\n"), query_string);
#endif
}
}
debug_printf("Fell through while parsing data...\n");
mstrncpy(ti->title, query_string, MAX_TRACK_LEN);
ti->have_track_info = 1;
compose_metadata(rmi, ti);
}
/* Return 1 if the page is a header page */
static int vorbis_process(RIP_MANAGER_INFO * rmi, stream_processor * stream, ogg_page * page, TRACK_INFO * ti)
{
ogg_packet packet;
misc_vorbis_info *inf = stream->data;
int i, header = 0;
int k;
ogg_stream_pagein(&stream->os, page);
if (inf->doneheaders < 3)
header = 1;
while (ogg_stream_packetout(&stream->os, &packet) > 0) {
if (inf->doneheaders < 3) {
if (vorbis_synthesis_headerin(&inf->vi, &inf->vc, &packet) < 0) {
warn(_("Warning: Could not decode vorbis header "
"packet - invalid vorbis stream (%d)\n"), stream->num);
continue;
}
inf->doneheaders++;
if (inf->doneheaders == 3) {
if (ogg_page_granulepos(page) != 0 || ogg_stream_packetpeek(&stream->os, NULL) == 1)
warn(_("Warning: Vorbis stream %d does not have headers "
"correctly framed. Terminal header page contains "
"additional packets or has non-zero granulepos\n"), stream->num);
debug_printf("Vorbis headers parsed for stream %d, "
"information follows...\n", stream->num);
debug_printf("Version: %d\n", inf->vi.version);
k = 0;
while (releases[k].vendor_string) {
if (!strcmp(inf->vc.vendor, releases[k].vendor_string)) {
debug_printf("Vendor: %s (%s)\n", inf->vc.vendor, releases[k].desc);
break;
}
k++;
}
if (!releases[k].vendor_string)
debug_printf("Vendor: %s\n", inf->vc.vendor);
debug_printf("Channels: %d\n", inf->vi.channels);
debug_printf("Rate: %ld\n\n", inf->vi.rate);
if (inf->vi.bitrate_nominal > 0)
debug_printf("Nominal bitrate: %f kb/s\n",
(double)inf->vi.bitrate_nominal / 1000.0);
else
debug_printf("Nominal bitrate not set\n");
if (inf->vi.bitrate_upper > 0)
debug_printf("Upper bitrate: %f kb/s\n",
(double)inf->vi.bitrate_upper / 1000.0);
else
debug_printf("Upper bitrate not set\n");
if (inf->vi.bitrate_lower > 0)
debug_printf("Lower bitrate: %f kb/s\n",
(double)inf->vi.bitrate_lower / 1000.0);
else
debug_printf("Lower bitrate not set\n");
if (inf->vc.comments > 0)
debug_printf("User comments section follows...\n");
for (i = 0; i < inf->vc.comments; i++) {
char *sep = strchr(inf->vc.user_comments[i], '=');
char *decoded;
int j;
int broken = 0;
unsigned char *val;
int bytes;
int remaining;
if (sep == NULL) {
warn(_("Warning: Comment %d in stream %d is invalidly "
"formatted, does not contain '=': \"%s\"\n"),
i, stream->num, inf->vc.user_comments[i]);
continue;
}
for (j = 0; j < sep - inf->vc.user_comments[i]; j++) {
if (inf->vc.user_comments[i][j] < 0x20 ||
inf->vc.user_comments[i][j] > 0x7D) {
warn(_("Warning: Invalid comment fieldname in "
"comment %d (stream %d): \"%s\"\n"),
i, stream->num, inf->vc.user_comments[i]);
broken = 1;
break;
}
}
if (broken)
continue;
val = (unsigned char *)inf->vc.user_comments[i];
j = sep - inf->vc.user_comments[i] + 1;
while (j < inf->vc.comment_lengths[i]) {
remaining = inf->vc.comment_lengths[i] - j;
if ((val[j] & 0x80) == 0)
bytes = 1;
else if ((val[j] & 0x40) == 0x40) {
if ((val[j] & 0x20) == 0)
bytes = 2;
else if ((val[j] & 0x10) == 0)
bytes = 3;
else if ((val[j] & 0x08) == 0)
bytes = 4;
else if ((val[j] & 0x04) == 0)
bytes = 5;
else if ((val[j] & 0x02) == 0)
bytes = 6;
else {
warn(_("Warning: Illegal UTF-8 sequence in "
"comment %d (stream %d): length "
"marker wrong\n"), i, stream->num);
broken = 1;
break;
}
} else {
warn(_("Warning: Illegal UTF-8 sequence in comment "
"%d (stream %d): length marker wrong\n"),
i, stream->num);
broken = 1;
break;
}
if (bytes > remaining) {
warn(_("Warning: Illegal UTF-8 sequence in comment "
"%d (stream %d): too few bytes\n"), i, stream->num);
broken = 1;
break;
}
switch (bytes) {
case 1:
/* No more checks needed */
break;
case 2:
if ((val[j + 1] & 0xC0) != 0x80)
broken = 1;
if ((val[j] & 0xFE) == 0xC0)
broken = 1;
break;
case 3:
if (!((val[j] == 0xE0 && val[j + 1] >= 0xA0 &&
val[j + 1] <= 0xBF &&
(val[j + 2] & 0xC0) == 0x80) ||
(val[j] >= 0xE1 && val[j] <= 0xEC &&
(val[j + 1] & 0xC0) == 0x80 &&
(val[j + 2] & 0xC0) == 0x80) ||
(val[j] == 0xED && val[j + 1] >= 0x80 &&
val[j + 1] <= 0x9F &&
(val[j + 2] & 0xC0) == 0x80) ||
(val[j] >= 0xEE && val[j] <= 0xEF &&
(val[j + 1] & 0xC0) == 0x80 &&
(val[j + 2] & 0xC0) == 0x80)))
broken = 1;
if (val[j] == 0xE0 && (val[j + 1] & 0xE0) == 0x80)
broken = 1;
break;
case 4:
if (!((val[j] == 0xF0 && val[j + 1] >= 0x90 &&
val[j + 1] <= 0xBF &&
(val[j + 2] & 0xC0) == 0x80 &&
(val[j + 3] & 0xC0) == 0x80) ||
(val[j] >= 0xF1 && val[j] <= 0xF3 &&
(val[j + 1] & 0xC0) == 0x80 &&
(val[j + 2] & 0xC0) == 0x80 &&
(val[j + 3] & 0xC0) == 0x80) ||
(val[j] == 0xF4 && val[j + 1] >= 0x80 &&
val[j + 1] <= 0x8F &&
(val[j + 2] & 0xC0) == 0x80 &&
(val[j + 3] & 0xC0) == 0x80)))
broken = 1;
if (val[j] == 0xF0 && (val[j + 1] & 0xF0) == 0x80)
broken = 1;
break;
/* 5 and 6 aren't actually allowed at this point */
case 5:
broken = 1;
break;
case 6:
broken = 1;
break;
}
if (broken) {
warn(_("Warning: Illegal UTF-8 sequence in comment "
"%d (stream %d): invalid sequence\n"), i, stream->num);
broken = 1;
break;
}
j += bytes;
}
if (!broken) {
if (utf8_decode(sep + 1, &decoded) < 0) {
warn(_("Warning: Failure in utf8 decoder. This "
"should be impossible\n"));
continue;
}
*sep = 0;
debug_printf("\t%s=%s\n", inf->vc.user_comments[i], decoded);
/* GCS FIX: Need case insensitive compare */
if (!strcmp(inf->vc.user_comments[i], "artist")
|| !strcmp(inf->vc.user_comments[i], "ARTIST")
|| !strcmp(inf->vc.user_comments[i], "Artist")) {
/* GCS FIX: This is a bit funky, maybe I need
to get rid of the ogg built-in utf8 decoder */
gstring_from_string(rmi, ti->artist,
MAX_TRACK_LEN, decoded, CODESET_LOCALE);
} else if (!strcmp(inf->vc.user_comments[i], "title")
|| !strcmp(inf->vc.user_comments[i], "TITLE")
|| !strcmp(inf->vc.user_comments[i], "Title")) {
/* GCS FIX: This is a bit funky, maybe I need
to get rid of the ogg built-in utf8 decoder */
gstring_from_string(rmi, ti->title, MAX_TRACK_LEN,
decoded, CODESET_LOCALE);
ti->have_track_info = 1;
} else if (!strcmp(inf->vc.user_comments[i], "album")
|| !strcmp(inf->vc.user_comments[i], "ALBUM")
|| !strcmp(inf->vc.user_comments[i], "Album")) {
/* GCS FIX: This is a bit funky, maybe I need
to get rid of the ogg built-in utf8 decoder */
gstring_from_string(rmi, ti->album, MAX_TRACK_LEN,
decoded, CODESET_LOCALE);
} else if (!strcmp(inf->vc.user_comments[i], "tracknumber")
|| !strcmp(inf->vc.user_comments[i], "TRACKNUMBER")
|| !strcmp(inf->vc.user_comments[i], "Tracknumber")) {
/* GCS FIX: This is a bit funky, maybe I need
to get rid of the ogg built-in utf8 decoder */
gstring_from_string(rmi, ti->track_p,
MAX_TRACK_LEN, decoded, CODESET_LOCALE);
}
free(decoded);
}
}
/* Done looping through vorbis comments. If we didn't find
a title, give a default title. */
if (!ti->have_track_info) {
strncpy(ti->title, "Title Unknown", MAX_TRACK_LEN);
ti->have_track_info = 1;
}
}
}
}
if (!header) {
ogg_int64_t gp = ogg_page_granulepos(page);
if (gp > 0) {
if (gp < inf->lastgranulepos)
#ifdef _WIN32
warn(_("Warning: granulepos in stream %d decreases from %I64d to %I64d"),
stream->num, inf->lastgranulepos, gp);
#else
warn(_("Warning: granulepos in stream %d decreases from %lld to %lld"),
stream->num, inf->lastgranulepos, gp);
#endif
inf->lastgranulepos = gp;
} else {
warn(_
("Negative granulepos on vorbis stream outside of headers. This file was created by a buggy encoder\n"));
}
if (inf->firstgranulepos < 0) { /* Not set yet */
}
inf->bytes += page->header_len + page->body_len;
}
return header;
}
