static void sami_add_line(subtitle *current, char *buffer, char **pos) {
char *p = *pos;
*p = 0;
trail_space(buffer);
if (*buffer && current->lines < SUB_MAX_TEXT)
current->text[current->lines++] = strdup(buffer);
*pos = buffer;
}
static subtitle *sub_read_line_jacosub(stream_t* st, subtitle * current,
struct readline_args *args)
{
int utf16 = args->utf16;
char line1[LINE_LEN], line2[LINE_LEN], directive[LINE_LEN], *p, *q;
unsigned a1, a2, a3, a4, b1, b2, b3, b4, comment = 0;
static unsigned jacoTimeres = 30;
static int jacoShift = 0;
memset(current, 0, sizeof(subtitle));
memset(line1, 0, LINE_LEN);
memset(line2, 0, LINE_LEN);
memset(directive, 0, LINE_LEN);
while (!current->text[0]) {
if (!stream_read_line(st, line1, LINE_LEN, utf16)) {
return NULL;
}
if (sscanf
(line1, "%u:%u:%u.%u %u:%u:%u.%u %[^\n\r]", &a1, &a2, &a3, &a4,
&b1, &b2, &b3, &b4, line2) < 9) {
if (sscanf(line1, "@%u @%u %[^\n\r]", &a4, &b4, line2) < 3) {
if (line1[0] == '#') {
int hours = 0, minutes = 0, seconds, delta, inverter =
1;
unsigned units = jacoShift;
switch (toupper(line1[1])) {
case 'S':
if (isalpha(line1[2])) {
delta = 6;
} else {
delta = 2;
}
if (sscanf(&line1[delta], "%d", &hours)) {
if (hours < 0) {
hours *= -1;
inverter = -1;
}
if (sscanf(&line1[delta], "%*d:%d", &minutes)) {
if (sscanf
(&line1[delta], "%*d:%*d:%d",
&seconds)) {
sscanf(&line1[delta], "%*d:%*d:%*d.%d",
&units);
} else {
hours = 0;
sscanf(&line1[delta], "%d:%d.%d",
&minutes, &seconds, &units);
minutes *= inverter;
}
} else {
hours = minutes = 0;
sscanf(&line1[delta], "%d.%d", &seconds,
&units);
seconds *= inverter;
}
jacoShift =
((hours * 3600 + minutes * 60 +
seconds) * jacoTimeres +
units) * inverter;
}
break;
case 'T':
if (isalpha(line1[2])) {
delta = 8;
} else {
delta = 2;
}
sscanf(&line1[delta], "%u", &jacoTimeres);
break;
}
}
continue;
} else {
current->start =
(unsigned long) ((a4 + jacoShift) * 100.0 /
jacoTimeres);
current->end =
(unsigned long) ((b4 + jacoShift) * 100.0 /
jacoTimeres);
}
} else {
current->start =
(unsigned
long) (((a1 * 3600 + a2 * 60 + a3) * jacoTimeres + a4 +
jacoShift) * 100.0 / jacoTimeres);
current->end =
(unsigned
long) (((b1 * 3600 + b2 * 60 + b3) * jacoTimeres + b4 +
jacoShift) * 100.0 / jacoTimeres);
}
current->lines = 0;
p = line2;
while ((*p == ' ') || (*p == '\t')) {
++p;
}
if (isalpha(*p)||*p == '[') {
int cont, jLength;
if (sscanf(p, "%s %[^\n\r]", directive, line1) < 2)
return (subtitle *) ERR;
jLength = strlen(directive);
for (cont = 0; cont < jLength; ++cont) {
if (isalpha(*(directive + cont)))
*(directive + cont) = toupper(*(directive + cont));
}
if ((strstr(directive, "RDB") != NULL)
|| (strstr(directive, "RDC") != NULL)
|| (strstr(directive, "RLB") != NULL)
|| (strstr(directive, "RLG") != NULL)) {
continue;
}
if (strstr(directive, "JL") != NULL) {
current->alignment = SUB_ALIGNMENT_BOTTOMLEFT;
} else if (strstr(directive, "JR") != NULL) {
current->alignment = SUB_ALIGNMENT_BOTTOMRIGHT;
} else {
current->alignment = SUB_ALIGNMENT_BOTTOMCENTER;
}
snprintf(line2, sizeof(line2), "%s", line1);
p = line2;
}
for (q = line1; (!eol(*p)) && (current->lines < SUB_MAX_TEXT); ++p) {
switch (*p) {
case '{':
comment++;
break;
case '}':
if (comment) {
--comment;
//the next line to get rid of a blank after the comment
if ((*(p + 1)) == ' ')
p++;
}
break;
case '~':
if (!comment) {
*q = ' ';
++q;
}
break;
case ' ':
case '\t':
if ((*(p + 1) == ' ') || (*(p + 1) == '\t'))
break;
if (!comment) {
*q = ' ';
++q;
}
break;
case '\\':
if (*(p + 1) == 'n') {
*q = '\0';
q = line1;
current->text[current->lines++] = strdup(line1);
++p;
break;
}
if ((toupper(*(p + 1)) == 'C')
|| (toupper(*(p + 1)) == 'F')) {
++p,++p;
break;
}
if ((*(p + 1) == 'B') || (*(p + 1) == 'b') || (*(p + 1) == 'D') || //actually this means "insert current date here"
(*(p + 1) == 'I') || (*(p + 1) == 'i') || (*(p + 1) == 'N') || (*(p + 1) == 'T') || //actually this means "insert current time here"
(*(p + 1) == 'U') || (*(p + 1) == 'u')) {
++p;
break;
}
if ((*(p + 1) == '\\') ||
(*(p + 1) == '~') || (*(p + 1) == '{')) {
++p;
} else if (eol(*(p + 1))) {
if (!stream_read_line(st, directive, LINE_LEN, utf16))
return NULL;
trail_space(directive);
av_strlcat(line2, directive, LINE_LEN);
break;
}
default:
if (!comment) {
*q = *p;
++q;
}
} //-- switch
} //-- for
*q = '\0';
if (current->lines < SUB_MAX_TEXT)
current->text[current->lines] = strdup(line1);
} //-- while
if (current->lines < SUB_MAX_TEXT)
current->lines++;
return current;
}
static subtitle *sub_read_line_sami(stream_t* st, subtitle *current,
struct readline_args *args)
{
int utf16 = args->utf16;
static char line[LINE_LEN+1];
static char *s = NULL, *slacktime_s;
char text[LINE_LEN+1], *p=NULL, *q;
int state;
current->lines = current->start = current->end = 0;
current->alignment = SUB_ALIGNMENT_BOTTOMCENTER;
state = 0;
/* read the first line */
if (!s)
if (!(s = stream_read_line(st, line, LINE_LEN, utf16))) return 0;
do {
switch (state) {
case 0: /* find "START=" or "Slacktime:" */
slacktime_s = stristr (s, "Slacktime:");
if (slacktime_s)
args->sub_slacktime = strtol (slacktime_s+10, NULL, 0) / 10;
s = stristr (s, "Start=");
if (s) {
current->start = strtol (s + 6, &s, 0) / 10;
/* eat '>' */
for (; *s != '>' && *s != '\0'; s++);
s++;
state = 1; continue;
}
break;
case 1: /* find (optional) "<P", skip other TAGs */
for (; *s == ' ' || *s == '\t'; s++); /* strip blanks, if any */
if (*s == '\0') break;
if (*s != '<') { state = 3; p = text; continue; } /* not a TAG */
s++;
if (*s == 'P' || *s == 'p') { s++; state = 2; continue; } /* found '<P' */
for (; *s != '>' && *s != '\0'; s++); /* skip remains of non-<P> TAG */
if (*s == '\0')
break;
s++;
continue;
case 2: /* find ">" */
if ((s = strchr (s, '>'))) { s++; state = 3; p = text; continue; }
break;
case 3: /* get all text until '<' appears */
if (p - text >= LINE_LEN)
sami_add_line(current, text, &p);
if (*s == '\0') break;
else if (!strncasecmp (s, "<br>", 4)) {
sami_add_line(current, text, &p);
s += 4;
}
else if (*s == '{') { state = 5; ++s; continue; }
else if (*s == '<') { state = 4; }
else if (!strncasecmp (s, " ", 6)) { *p++ = ' '; s += 6; }
else if (*s == '\t') { *p++ = ' '; s++; }
else if (*s == '\r' || *s == '\n') { s++; }
else *p++ = *s++;
/* skip duplicated space */
if (p > text + 2) if (*(p-1) == ' ' && *(p-2) == ' ') p--;
continue;
case 4: /* get current->end or skip <TAG> */
q = stristr (s, "Start=");
if (q) {
current->end = strtol (q + 6, &q, 0) / 10 - 1;
*p = '\0'; trail_space (text);
if (text[0] != '\0')
current->text[current->lines++] = strdup (text);
if (current->lines > 0) { state = 99; break; }
state = 0; continue;
}
s = strchr (s, '>');
if (s) { s++; state = 3; continue; }
break;
case 5: /* get rid of {...} text, but read the alignment code */
if ((*s == '\\') && (*(s + 1) == 'a')) {
if (stristr(s, "\\a1") != NULL) {
current->alignment = SUB_ALIGNMENT_BOTTOMLEFT;
s = s + 3;
}
if (stristr(s, "\\a2") != NULL) {
current->alignment = SUB_ALIGNMENT_BOTTOMCENTER;
s = s + 3;
} else if (stristr(s, "\\a3") != NULL) {
current->alignment = SUB_ALIGNMENT_BOTTOMRIGHT;
s = s + 3;
} else if ((stristr(s, "\\a4") != NULL) || (stristr(s, "\\a5") != NULL) || (stristr(s, "\\a8") != NULL)) {
current->alignment = SUB_ALIGNMENT_TOPLEFT;
s = s + 3;
} else if (stristr(s, "\\a6") != NULL) {
current->alignment = SUB_ALIGNMENT_TOPCENTER;
s = s + 3;
} else if (stristr(s, "\\a7") != NULL) {
current->alignment = SUB_ALIGNMENT_TOPRIGHT;
s = s + 3;
} else if (stristr(s, "\\a9") != NULL) {
current->alignment = SUB_ALIGNMENT_MIDDLELEFT;
s = s + 3;
} else if (stristr(s, "\\a10") != NULL) {
current->alignment = SUB_ALIGNMENT_MIDDLECENTER;
s = s + 4;
} else if (stristr(s, "\\a11") != NULL) {
current->alignment = SUB_ALIGNMENT_MIDDLERIGHT;
s = s + 4;
}
}
if (*s == '}') state = 3;
++s;
continue;
}
/* read next line */
if (state != 99 && !(s = stream_read_line (st, line, LINE_LEN, utf16))) {
if (current->start > 0) {
break; // if it is the last subtitle
} else {
return 0;
}
}
} while (state != 99);
// For the last subtitle
if (current->end <= 0) {
current->end = current->start + args->sub_slacktime;
sami_add_line(current, text, &p);
}
return current;
}
Tsubtitle* TsubtitleParserSami::parse(Tstream &fd, int flags, REFERENCE_TIME start, REFERENCE_TIME stop)
{
wchar_t text[this->LINE_LEN + 1], *p = NULL, *q;
int state;
/* read the first line */
if (!s)
if ((s = fd.fgets(line, this->LINE_LEN)) == NULL) {
return NULL;
}
TsubtitleText current(this->format);
current.start = current.stop = 0;
state = 0;
do {
switch (state) {
case 0: /* find "START=" or "Slacktime:" */
slacktime_s = stristr(s, L"Slacktime:");
if (slacktime_s) {
sub_slacktime = strtol(slacktime_s + 10, NULL, 0) / 10;
}
s = (wchar_t*)stristr(s, L"Start=");
if (s) {
int sec1000 = strtol(s + 6, &s, 0);
current.start = this->hmsToTime(0, 0, sec1000 / 1000, (sec1000 % 1000) / 10);
/* eat '>' */
for (; *s != '>' && *s != '\0'; s++) {
;
}
s++;
state = 1;
continue;
}
break;
case 1: /* find (optionnal) "<P", skip other TAGs */
for (; *s == ' ' || *s == '\t'; s++) {
; /* strip blanks, if any */
}
if (*s == '\0') {
break;
}
if (*s != '<') {
state = 3; /* not a TAG */
p = text;
continue;
}
s++;
if (*s == 'P' || *s == 'p') {
s++; /* found '<P' */
state = 2;
continue;
}
for (; *s != '>' && *s != '\0'; s++) {
; /* skip remains of non-<P> TAG */
}
if (s == '\0') {
break;
}
s++;
continue;
case 2: /* find ">" */
if ((s = strchr(s, '>')) != NULL) {
s++;
state = 3;
p = text;
continue;
}
break;
case 3: /* get all text until '<' appears */
if (*s == '\0') {
break;
} else if (!_strnicmp(s, L"<br>", 4)) {
*p = '\0';
p = text;
trail_space(text);
if (text[0] != '\0') {
current.add(text);
}
s += 4;
} else if ((*s == '{') && !sub_no_text_pp) {
state = 5;
++s;
continue;
} else if (*s == '<') {
state = 4;
} else if (!_strnicmp(s, L" ", 6)) {
*p++ = ' ';
s += 6;
} else if (*s == '\t') {
*p++ = ' ';
s++;
} else if (*s == '\r' || *s == '\n') {
s++;
} else {
*p++ = *s++;
}
/* skip duplicated space */
if (p > text + 2) if (*(p - 1) == ' ' && *(p - 2) == ' ') {
p--;
}
continue;
case 4: /* get current->end or skip <TAG> */
q = (wchar_t*)stristr(s, L"Start=");
if (q) {
int sec1000 = strtol(q + 6, &q, 0);
current.stop = this->hmsToTime(0, 0, sec1000 / 1000, (sec1000 % 1000) / 10 - 1);
*p = '\0';
trail_space(text);
if (text[0] != '\0') {
current.add(text);
}
if (current.size() > 0) {
state = 99;
break;
}
state = 0;
continue;
}
s = strchr(s, '>');
if (s) {
s++;
state = 3;
continue;
}
break;
case 5: /* get rid of {...} text, but read the alignment code */
if ((*s == '\\') && (*(s + 1) == 'a') && !sub_no_text_pp) {
if (stristr(s, L"\\a1") != NULL) {
//current->alignment = SUB_ALIGNMENT_BOTTOMLEFT;
s = s + 3;
}
if (stristr(s, L"\\a2") != NULL) {
//current->alignment = SUB_ALIGNMENT_BOTTOMCENTER;
s = s + 3;
} else if (stristr(s, L"\\a3") != NULL) {
//current->alignment = SUB_ALIGNMENT_BOTTOMRIGHT;
s = s + 3;
} else if ((stristr(s, L"\\a4") != NULL) || (stristr(s, L"\\a5") != NULL) || (stristr(s, L"\\a8") != NULL)) {
//current->alignment = SUB_ALIGNMENT_TOPLEFT;
s = s + 3;
} else if (stristr(s, L"\\a6") != NULL) {
//current->alignment = SUB_ALIGNMENT_TOPCENTER;
s = s + 3;
} else if (stristr(s, L"\\a7") != NULL) {
//current->alignment = SUB_ALIGNMENT_TOPRIGHT;
s = s + 3;
} else if (stristr(s, L"\\a9") != NULL) {
//current->alignment = SUB_ALIGNMENT_MIDDLELEFT;
s = s + 3;
} else if (stristr(s, L"\\a10") != NULL) {
//current->alignment = SUB_ALIGNMENT_MIDDLECENTER;
s = s + 4;
} else if (stristr(s, L"\\a11") != NULL) {
//current->alignment = SUB_ALIGNMENT_MIDDLERIGHT;
s = s + 4;
}
}
if (*s == '}') {
state = 3;
}
++s;
continue;
}
/* read next line */
if (state != 99 && (s = fd.fgets(line, this->LINE_LEN)) == NULL) {
if (current.start > 0) {
break; // if it is the last subtitle
} else {
return NULL;
}
}
} while (state != 99);
// For the last subtitle
if (current.stop <= 0) {
current.stop = current.start + this->hmsToTime(0, 0, sub_slacktime / 1000, (sub_slacktime % 1000) / 10);
*p = '\0';
trail_space(text);
if (text[0] != '\0') {
current.add(text);
}
}
return store(current);
}
