From 9333eb48998e4cfecce193d1c4046934cf6c6b50 Mon Sep 17 00:00:00 2001
From: =?utf8?q?Sebastian=20Dr=C3=B6ge?= <slomo@circular-chaos.org>
Date: Mon, 5 May 2008 12:33:05 +0000
Subject: [PATCH] gst/subparse/samiparse.c: Remove trailing, leading and double
 whitespaces.

Original commit message from CVS:
* gst/subparse/samiparse.c: (handle_start_sync),
(end_sami_element), (characters_sami):
Remove trailing, leading and double whitespaces.
Correctly timestamp buffers and output the last buffer too.
* tests/check/elements/subparse.c: (GST_START_TEST),
(subparse_suite):
Add a simple unit test for SAMI parsing.
---
 ChangeLog                       | 11 ++++++++++
 gst/subparse/samiparse.c        | 45 ++++++++++++++++++++++++++++++++---------
 tests/check/elements/subparse.c | 37 +++++++++++++++++++++++++++++++++
 3 files changed, 83 insertions(+), 10 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 26d8796..7a722d8 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,16 @@
 2008-05-05  Sebastian Dröge  <slomo@circular-chaos.org>
 
+	* gst/subparse/samiparse.c: (handle_start_sync),
+	(end_sami_element), (characters_sami):
+	Remove trailing, leading and double whitespaces.
+	Correctly timestamp buffers and output the last buffer too.
+
+	* tests/check/elements/subparse.c: (GST_START_TEST),
+	(subparse_suite):
+	Add a simple unit test for SAMI parsing.
+
+2008-05-05  Sebastian Dröge  <slomo@circular-chaos.org>
+
 	Patch by: Young-Ho Cha <ganadist at chollian dot net>
 
 	* gst/subparse/samiparse.c: (handle_start_sync),
diff --git a/gst/subparse/samiparse.c b/gst/subparse/samiparse.c
index 4563279..1cceeaa 100644
--- a/gst/subparse/samiparse.c
+++ b/gst/subparse/samiparse.c
@@ -131,18 +131,13 @@ handle_start_sync (GstSamiContext * sctx, const xmlChar ** atts)
       if (!value)
         continue;
       if (!xmlStrncmp ((const xmlChar *) "start", key, 5)) {
-        /* FIXME: this is not correct. According to[0] the start parameter
-         * specifies the time when the text should be shown and not the time
-         * when the text should be hidden again.
-         * This essentially means that we have to look at the next sync element
-         * before pushing one buffer and push the last buffer with -1
-         * as duration after the body element is closed.
-         * [0] http://msdn.microsoft.com/en-us/library/ms971327.aspx
-         */
-        sctx->time1 = sctx->time2;
+        /* Only set a new start time if we don't have text pending */
+        if (sctx->resultbuf->len == 0)
+          sctx->time1 = sctx->time2;
+
         sctx->time2 = atoi ((const char *) value) * GST_MSECOND;
-        sctx->has_result = TRUE;
         g_string_append (sctx->resultbuf, sctx->buf->str);
+        sctx->has_result = (sctx->resultbuf->len != 0) ? TRUE : FALSE;
         g_string_truncate (sctx->buf, 0);
       }
     }
@@ -248,6 +243,19 @@ end_sami_element (void *ctx, const xmlChar * name)
 
   if (!xmlStrncmp ((const xmlChar *) "sync", name, 4)) {
     sctx->in_sync = FALSE;
+  } else if (!xmlStrncmp ((const xmlChar *) "body", name, 4)) {
+    /* We will usually have one buffer left when the body is closed
+     * as we need the next sync to actually send it */
+    if (sctx->buf->len != 0) {
+      /* Only set a new start time if we don't have text pending */
+      if (sctx->resultbuf->len == 0)
+        sctx->time1 = sctx->time2;
+
+      sctx->time2 = GST_CLOCK_TIME_NONE;
+      g_string_append (sctx->resultbuf, sctx->buf->str);
+      sctx->has_result = (sctx->resultbuf->len != 0) ? TRUE : FALSE;
+      g_string_truncate (sctx->buf, 0);
+    }
   } else if (!xmlStrncmp ((const xmlChar *) "font", name, 4)) {
     sami_context_pop_state (sctx, SPAN_TAG);
   } else if (!xmlStrncmp ((const xmlChar *) "ruby", name, 4)) {
@@ -262,12 +270,29 @@ characters_sami (void *ctx, const xmlChar * ch, int len)
 {
   GstSamiContext *sctx = (GstSamiContext *) ctx;
   gchar *escaped;
+  gchar *tmp;
+  gint i;
 
   /* Skip everything except content of the sync elements */
   if (!sctx->in_sync)
     return;
 
   escaped = g_markup_escape_text ((const gchar *) ch, len);
+  g_strstrip (escaped);
+
+  /* Remove double spaces forom the string as those are
+   * usually added by newlines and indention */
+  tmp = escaped;
+  for (i = 0; i <= strlen (escaped); i++) {
+    escaped[i] = *tmp;
+    if (*tmp != ' ') {
+      tmp++;
+      continue;
+    }
+    while (*tmp == ' ')
+      tmp++;
+  }
+
   if (has_tag (sctx->state, RT_TAG)) {
     g_string_append_c (sctx->rubybuf, ' ');
     g_string_append (sctx->rubybuf, escaped);
diff --git a/tests/check/elements/subparse.c b/tests/check/elements/subparse.c
index 4101df9..c0b5c45 100644
--- a/tests/check/elements/subparse.c
+++ b/tests/check/elements/subparse.c
@@ -579,6 +579,42 @@ GST_START_TEST (test_subviewer2)
 
 GST_END_TEST;
 
+GST_START_TEST (test_sami)
+{
+  SubParseInputChunk sami_input[] = {
+    {"<SAMI>\n"
+          "<HEAD>\n"
+          "    <TITLE>Subtitle</TITLE>\n"
+          "    <STYLE TYPE=\"text/css\">\n"
+          "    <!--\n"
+          "        P {margin-left:8pt; margin-right:8pt; margin-bottom:2pt; margin-top:2pt; text-align:center; font-size:12pt; font-weight:normal; color:black;}\n"
+          "        .CC {Name:English; lang:en-AU; SAMIType:CC;}\n"
+          "        #STDPrn {Name:Standard Print;}\n"
+          "        #LargePrn {Name:Large Print; font-size:24pt;}\n"
+          "        #SmallPrn {Name:Small Print; font-size:16pt;}\n"
+          "    -->\n"
+          "    </Style>\n"
+          "</HEAD>\n"
+          "<BODY>\n"
+          "    <SYNC Start=1000>\n"
+          "        <P Class=CC>\n"
+          "            This is a comment.<br>\n"
+          "            This is a second comment.\n",
+          1000 * GST_MSECOND, 2000 * GST_MSECOND,
+        "This is a comment.\nThis is a second comment."},
+    {"    <SYNC Start=2000>\n"
+          "        <P Class=CC>\n"
+          "            This is a third comment.<br>\n"
+          "            This is a fourth comment.\n" "</BODY>\n" "</SAMI>\n",
+          2000 * GST_MSECOND, GST_CLOCK_TIME_NONE,
+        "This is a third comment.\nThis is a fourth comment."}
+  };
+
+  do_test (sami_input, G_N_ELEMENTS (sami_input), "text/x-pango-markup");
+}
+
+GST_END_TEST;
+
 /* TODO:
  *  - add/modify tests so that lines aren't dogfed to the parsers in complete
  *    lines or sets of complete lines, but rather in random chunks
@@ -606,6 +642,7 @@ subparse_suite (void)
   tcase_add_test (tc_chain, test_mpl2);
   tcase_add_test (tc_chain, test_subviewer);
   tcase_add_test (tc_chain, test_subviewer2);
+  tcase_add_test (tc_chain, test_sami);
   return s;
 }
 
-- 
2.7.4