charset_conv: make it possible to return an allocated string as guess

uchardet is written in C++, and thus doesn't appreciate the value of
using static strings, and internally stores the guessed charset as
allocated std::string. Add a minimal hack to deal with this. (I don't
appreciate that the code is potentially harder to understand by
returning either a static or allocated string, but I do appreciate for
not having to litter the existing code with strdups.)
This commit is contained in:
wm4 2015-08-01 23:25:50 +02:00
parent 17fe9d7c0d
commit 11f2be2bcc
3 changed files with 14 additions and 10 deletions

View File

@ -150,8 +150,9 @@ static const char *libguess_guess(struct mp_log *log, bstr buf,
// If user_cp doesn't refer to any known auto-detection (for example because // If user_cp doesn't refer to any known auto-detection (for example because
// it's a real iconv codepage), user_cp is returned without even looking at // it's a real iconv codepage), user_cp is returned without even looking at
// the buf data. // the buf data.
const char *mp_charset_guess(struct mp_log *log, bstr buf, const char *user_cp, // The return value may (but doesn't have to) be allocated under talloc_ctx.
int flags) const char *mp_charset_guess(void *talloc_ctx, struct mp_log *log, bstr buf,
const char *user_cp, int flags)
{ {
if (!mp_charset_requires_guess(user_cp)) if (!mp_charset_requires_guess(user_cp))
return user_cp; return user_cp;
@ -225,8 +226,11 @@ const char *mp_charset_guess(struct mp_log *log, bstr buf, const char *user_cp,
bstr mp_charset_guess_and_conv_to_utf8(struct mp_log *log, bstr buf, bstr mp_charset_guess_and_conv_to_utf8(struct mp_log *log, bstr buf,
const char *user_cp, int flags) const char *user_cp, int flags)
{ {
return mp_iconv_to_utf8(log, buf, mp_charset_guess(log, buf, user_cp, flags), void *tmp = talloc_new(NULL);
flags); const char *cp = mp_charset_guess(log, tmp, buf, user_cp, flags);
bstr res = mp_iconv_to_utf8(log, buf, cp, flags);
talloc_free(tmp);
return res;
} }
// Use iconv to convert buf to UTF-8. // Use iconv to convert buf to UTF-8.

View File

@ -14,8 +14,8 @@ enum {
bool mp_charset_is_utf8(const char *user_cp); bool mp_charset_is_utf8(const char *user_cp);
bool mp_charset_requires_guess(const char *user_cp); bool mp_charset_requires_guess(const char *user_cp);
const char *mp_charset_guess(struct mp_log *log, bstr buf, const char *user_cp, const char *mp_charset_guess(void *talloc_ctx, struct mp_log *log, bstr buf,
int flags); const char *user_cp, int flags);
bstr mp_charset_guess_and_conv_to_utf8(struct mp_log *log, bstr buf, bstr mp_charset_guess_and_conv_to_utf8(struct mp_log *log, bstr buf,
const char *user_cp, int flags); const char *user_cp, int flags);
bstr mp_iconv_to_utf8(struct mp_log *log, bstr buf, const char *cp, int flags); bstr mp_iconv_to_utf8(struct mp_log *log, bstr buf, const char *cp, int flags);

View File

@ -303,8 +303,8 @@ void sub_decode(struct dec_sub *sub, struct demux_packet *packet)
pthread_mutex_unlock(&sub->lock); pthread_mutex_unlock(&sub->lock);
} }
static const char *guess_sub_cp(struct mp_log *log, struct packet_list *subs, static const char *guess_sub_cp(struct mp_log *log, void *talloc_ctx,
const char *usercp) struct packet_list *subs, const char *usercp)
{ {
if (!mp_charset_requires_guess(usercp)) if (!mp_charset_requires_guess(usercp))
return usercp; return usercp;
@ -330,7 +330,7 @@ static const char *guess_sub_cp(struct mp_log *log, struct packet_list *subs,
memcpy(text.start + text.len + pkt->len, sep, sep_len); memcpy(text.start + text.len + pkt->len, sep, sep_len);
text.len += pkt->len + sep_len; text.len += pkt->len + sep_len;
} }
const char *guess = mp_charset_guess(log, text, usercp, 0); const char *guess = mp_charset_guess(talloc_ctx, log, text, usercp, 0);
talloc_free(text.start); talloc_free(text.start);
return guess; return guess;
} }
@ -455,7 +455,7 @@ bool sub_read_all_packets(struct dec_sub *sub, struct sh_stream *sh)
} }
if (opts->sub_cp && !sh->sub->is_utf8) if (opts->sub_cp && !sh->sub->is_utf8)
sub->charset = guess_sub_cp(sub->log, subs, opts->sub_cp); sub->charset = guess_sub_cp(sub->log, sub, subs, opts->sub_cp);
if (sub->charset && sub->charset[0] && !mp_charset_is_utf8(sub->charset)) if (sub->charset && sub->charset[0] && !mp_charset_is_utf8(sub->charset))
MP_INFO(sub, "Using subtitle charset: %s\n", sub->charset); MP_INFO(sub, "Using subtitle charset: %s\n", sub->charset);