filter_sdh: combine skip_bracketed and skip_parenthesized

These two functions are almost exactly the same. The parenthesis variant
is essentially just a special case with more conditions to not remove
text. These can easily be combined together into one generic
skip_enclosed function to handle both cases. We also use char * instead
of char for the character comparison here since not everything is
neccesarily 1 byte and can fit into a char. This will be useful for the
following commits where we extend this logic further.
This commit is contained in:
Dudemanguy 2023-11-05 12:16:48 -06:00
parent 2193893be7
commit b7d85f0d4a
1 changed files with 29 additions and 68 deletions

View File

@ -86,7 +86,8 @@ static void copy_ass(struct sd_filter *sd, char **rpp, struct buffer *buf)
return;
}
static bool skip_bracketed(struct sd_filter *sd, char **rpp, struct buffer *buf);
static bool skip_enclosed(struct sd_filter *sd, char **rpp, struct buffer *buf,
const char *left, const char *right);
// check for speaker label, like MAN:
// normal subtitles may include mixed case text with : after so
@ -128,7 +129,7 @@ static void skip_speaker_label(struct sd_filter *sd, char **rpp, struct buffer *
copy_ass(sd, &rp, buf);
} else if (rp[0] == '[') {
// not uncommon with [xxxx]: which should also be skipped
if (!skip_bracketed(sd, &rp, buf)) {
if (!skip_enclosed(sd, &rp, buf, "[", "]")) {
buf->pos = old_pos;
return;
}
@ -174,94 +175,54 @@ static void skip_speaker_label(struct sd_filter *sd, char **rpp, struct buffer *
return;
}
// check for bracketed text, like [SOUND]
// and skip it while preserving ass tags
// any characters are allowed, brackets are seldom used in normal text
// Check for text enclosed in symbols, like (SOUND)
// and skip it while preserving ass tags.
// Parentheses are a special case since normal subtitles may have
// them so only upper case is accepted and lower case l which for
// some looks like upper case I. If sub_filter_SDH_harder is used,
// both upper and lower case is accepted.
//
// For other symbols, all text in between is removed.
//
// Parameters:
// rpp read pointer pointer to source string, updated on return
// buf write buffer
//
// scan in source string
// the first character in source string must by the starting '['
// the first character in source string must be the starting left symbol
// and copy ass tags to destination string but
// skipping bracketed text if it looks like SDH
// skipping enclosed text if it looks like SDH
//
// return true if bracketed text was removed.
// return true if enclosed text was removed.
// if not valid SDH read pointer and write buffer position will be unchanged
// otherwise they point to next position after text and next write position
static bool skip_bracketed(struct sd_filter *sd, char **rpp, struct buffer *buf)
static bool skip_enclosed(struct sd_filter *sd, char **rpp, struct buffer *buf,
const char *left, const char *right)
{
bool filter_harder = sd->opts->sub_filter_SDH_harder;
char *rp = *rpp;
int old_pos = buf->pos;
rp++; // skip past '['
// skip past valid data searching for ]
while (*rp && rp[0] != ']') {
rp++; // skip past the left character
// skip past valid data searching for the right character
bool only_digits = strcmp(left, "(") == 0;
while (*rp && rp[0] != right[0]) {
if (rp[0] == '{') {
copy_ass(sd, &rp, buf);
} else {
rp++;
}
}
if (!*rp) {
// ] was not found
buf->pos = old_pos;
return false;
}
rp++; // skip ]
// skip trailing spaces
while (rp[0] == ' ') {
rp++;
}
*rpp = rp;
return true;
}
// check for parenthesized text, like (SOUND)
// and skip it while preserving ass tags
// normal subtitles may include mixed case text in parentheses so
// only upper case is accepted and lower case l which for some
// looks like upper case I but if requested harder filtering
// both upper and lower case is accepted
//
// Parameters:
// rpp read pointer pointer to source string, updated on return
// buf write buffer
//
// scan in source string
// the first character in source string must be the starting '('
// and copy ass tags to destination string but
// skipping parenthesized text if it looks like SDH
//
// return true if parenthesized text was removed.
// if not valid SDH read pointer and write buffer position will be unchanged
// otherwise they point to next position after text and next write position
static bool skip_parenthesized(struct sd_filter *sd, char **rpp, struct buffer *buf)
{
int filter_harder = sd->opts->sub_filter_SDH_harder;
char *rp = *rpp;
int old_pos = buf->pos;
rp++; // skip past '('
// skip past valid data searching for )
bool only_digits = true;
while (*rp && rp[0] != ')') {
if (rp[0] == '{') {
copy_ass(sd, &rp, buf);
} else if ((mp_isalpha(rp[0]) &&
} else if (strcmp(left, "(") == 0 && ((mp_isalpha(rp[0]) &&
(filter_harder || mp_isupper(rp[0]) || rp[0] == 'l')) ||
mp_isdigit(rp[0]) ||
rp[0] == ' ' || rp[0] == '\'' || rp[0] == '#' ||
rp[0] == '.' || rp[0] == ',' ||
rp[0] == '-' || rp[0] == '"' || rp[0] == '\\') {
rp[0] == '-' || rp[0] == '"' || rp[0] == '\\')) {
if (!mp_isdigit(rp[0]))
only_digits = false;
rp++;
} else {
} else if (strcmp(left, "(") == 0) {
buf->pos = old_pos;
return false;
} else {
rp++;
}
}
if (!*rp) {
@ -274,7 +235,7 @@ static bool skip_parenthesized(struct sd_filter *sd, char **rpp, struct buffer *
buf->pos = old_pos;
return false;
}
rp++; // skip )
rp++; // skip right character
// skip trailing spaces
while (rp[0] == ' ') {
rp++;
@ -372,13 +333,13 @@ static char *filter_SDH(struct sd_filter *sd, char *data, int length, ptrdiff_t
while (*rp && !(rp[0] == '\\' && rp[1] == 'N')) {
copy_ass(sd, &rp, buf);
if (rp[0] == '[') {
if (!skip_bracketed(sd, &rp, buf)) {
if (!skip_enclosed(sd, &rp, buf, "[", "]")) {
append(sd, buf, rp[0]);
rp++;
line_with_text = true;
}
} else if (rp[0] == '(') {
if (!skip_parenthesized(sd, &rp, buf)) {
if (!skip_enclosed(sd, &rp, buf, "(", ")")) {
append(sd, buf, rp[0]);
rp++;
line_with_text = true;