From 11f2be2bcc264aa4f24cdad036231e34d09f5630 Mon Sep 17 00:00:00 2001
From: wm4 <wm4@nowhere>
Date: Sat, 1 Aug 2015 23:25:50 +0200
Subject: 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.)
---
 misc/charset_conv.c | 12 ++++++++----
 misc/charset_conv.h |  4 ++--
 sub/dec_sub.c       |  8 ++++----
 3 files changed, 14 insertions(+), 10 deletions(-)

diff --git a/misc/charset_conv.c b/misc/charset_conv.c
index 31f53ccecb..343fb7fd90 100644
--- a/misc/charset_conv.c
+++ b/misc/charset_conv.c
@@ -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
 // it's a real iconv codepage), user_cp is returned without even looking at
 // the buf data.
-const char *mp_charset_guess(struct mp_log *log, bstr buf, const char *user_cp,
-                             int flags)
+// The return value may (but doesn't have to) be allocated under talloc_ctx.
+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))
         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,
                                        const char *user_cp, int flags)
 {
-    return mp_iconv_to_utf8(log, buf, mp_charset_guess(log, buf, user_cp, flags),
-                            flags);
+    void *tmp = talloc_new(NULL);
+    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.
diff --git a/misc/charset_conv.h b/misc/charset_conv.h
index 93bd91cffe..bd76ae007a 100644
--- a/misc/charset_conv.h
+++ b/misc/charset_conv.h
@@ -14,8 +14,8 @@ enum {
 
 bool mp_charset_is_utf8(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,
-                             int flags);
+const char *mp_charset_guess(void *talloc_ctx, struct mp_log *log, bstr buf,
+                             const char *user_cp, int flags);
 bstr mp_charset_guess_and_conv_to_utf8(struct mp_log *log, bstr buf,
                                        const char *user_cp, int flags);
 bstr mp_iconv_to_utf8(struct mp_log *log, bstr buf, const char *cp, int flags);
diff --git a/sub/dec_sub.c b/sub/dec_sub.c
index 1800cf34d9..20fa9eff83 100644
--- a/sub/dec_sub.c
+++ b/sub/dec_sub.c
@@ -303,8 +303,8 @@ void sub_decode(struct dec_sub *sub, struct demux_packet *packet)
     pthread_mutex_unlock(&sub->lock);
 }
 
-static const char *guess_sub_cp(struct mp_log *log, struct packet_list *subs,
-                                const char *usercp)
+static const char *guess_sub_cp(struct mp_log *log, void *talloc_ctx,
+                                struct packet_list *subs, const char *usercp)
 {
     if (!mp_charset_requires_guess(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);
         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);
     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)
-        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))
         MP_INFO(sub, "Using subtitle charset: %s\n", sub->charset);
-- 
cgit v1.2.3