diff options
| author | Edward Hervey <edward.hervey@collabora.co.uk> | 2012-06-10 11:36:32 +0200 |
|---|---|---|
| committer | Sebastian Dröge <sebastian.droege@collabora.co.uk> | 2012-06-26 12:46:46 +0200 |
| commit | ee9cc1cb4329fc0db8ed0211dbc881dae9ced4ed (patch) | |
| tree | 2e22a08b83537ca2af8e731875c7728071b3729d /gst | |
| parent | 97a241ca26b603d23862c32004b60ff99da5da4d (diff) | |
mpegtspacketizer: Don't use GByteArray
Using a simple array comes to the same result for a fraction of the cost
Diffstat (limited to 'gst')
| -rw-r--r-- | gst/mpegtsdemux/mpegtspacketizer.c | 44 |
1 files changed, 27 insertions, 17 deletions
diff --git a/gst/mpegtsdemux/mpegtspacketizer.c b/gst/mpegtsdemux/mpegtspacketizer.c index 51c91e303..5c04cbb89 100644 --- a/gst/mpegtsdemux/mpegtspacketizer.c +++ b/gst/mpegtsdemux/mpegtspacketizer.c @@ -2829,7 +2829,7 @@ convert_to_utf8 (const gchar * text, gint length, guint start, const gchar * encoding, gboolean is_multibyte, GError ** error) { gchar *new_text; - GByteArray *sb; + gchar *tmp, *pos; gint i; g_return_val_if_fail (text != NULL, NULL); @@ -2837,8 +2837,7 @@ convert_to_utf8 (const gchar * text, gint length, guint start, text += start; - /* FIXME : GByteArray has a big overhead */ - sb = g_byte_array_sized_new (length * 1.1); + pos = tmp = g_malloc (length * 2); if (is_multibyte) { if (length == -1) { @@ -2851,12 +2850,15 @@ convert_to_utf8 (const gchar * text, gint length, guint start, /* skip it */ break; case 0xE08A:{ - guint8 nl[] = { 0x00, 0x0A }; /* new line */ - g_byte_array_append (sb, nl, 2); + pos[0] = 0x00; /* 0x00 0x0A is a new line */ + pos[1] = 0x0A; + pos += 2; break; } default: - g_byte_array_append (sb, (guint8 *) text, 2); + pos[0] = text[0]; + pos[1] = text[1]; + pos += 2; break; } @@ -2872,12 +2874,15 @@ convert_to_utf8 (const gchar * text, gint length, guint start, /* skip it */ break; case 0xE08A:{ - guint8 nl[] = { 0x00, 0x0A }; /* new line */ - g_byte_array_append (sb, nl, 2); + pos[0] = 0x00; /* 0x00 0x0A is a new line */ + pos[1] = 0x0A; + pos += 2; break; } default: - g_byte_array_append (sb, (guint8 *) text, 2); + pos[0] = text[0]; + pos[1] = text[1]; + pos += 2; break; } @@ -2895,10 +2900,12 @@ convert_to_utf8 (const gchar * text, gint length, guint start, /* skip it */ break; case 0x8A: - g_byte_array_append (sb, (guint8 *) "\n", 1); + *pos = '\n'; + pos += 1; break; default: - g_byte_array_append (sb, &code, 1); + *pos = *text; + pos += 1; break; } @@ -2914,10 +2921,12 @@ convert_to_utf8 (const gchar * text, gint length, guint start, /* skip it */ break; case 0x8A: - g_byte_array_append (sb, (guint8 *) "\n", 1); + *pos = '\n'; + pos += 1; break; default: - g_byte_array_append (sb, &code, 1); + *pos = *text; + pos += 1; break; } @@ -2926,15 +2935,16 @@ convert_to_utf8 (const gchar * text, gint length, guint start, } } - if (sb->len > 0) { + if (pos > tmp) { + gsize bread = 0; new_text = - g_convert ((gchar *) sb->data, sb->len, "utf-8", encoding, NULL, NULL, - error); + g_convert (tmp, pos - tmp, "utf-8", encoding, &bread, NULL, error); + GST_DEBUG ("Converted to : %s", new_text); } else { new_text = g_strdup (""); } - g_byte_array_free (sb, TRUE); + g_free (tmp); return new_text; } |