/*
* Verifies that a cue text start tag with an invalid name will be ignored.
* From http://dev.w3.org/html5/webvtt/#webvtt-tag-state As given by the algorithm defined and any connecting parts. (11/27/2012)
* Relevant cue text specification rules:
* 1. Characters that come after the "<" and before a white space are considered to be apart of the tag name.
* 2. Tag names that are not valid are ignored.
* 3. Characters that come after the beginning "<" brace and after a white space character are considered to be apart of the annotation.
*
* In this example the parser reads "iare</i" inside of the beginning "<" and ending ">" as the tag name.
*/
TEST_F(PayloadTagFormat,DISABLED_StartTagNoEndBrace)
{
loadVtt( "payload/tag-format/start-tag-no-end-brace.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 2 );
ASSERT_EQ( Node::Text, head->child( 0 )->kind() );
ASSERT_EQ( Node::Text, head->child( 1 )->kind() );
}
TEST_F(PayloadLangTag, InternalWithinTag)
{
loadVtt( "payload/lang-tag/internal-within-lang.vtt" );
const Node head = getHeadOfCue( 0 );
EXPECT_EQ( 1, head[ 0 ][ 0 ].childCount() );
EXPECT_EQ( Node::Bold, head[ 0 ][ 0 ].kind() );
EXPECT_STREQ( "en", head[ 0 ][ 0 ].lang().utf8() );
}
TEST_F(PayloadLangTag, LangWithoutAnnotation)
{
loadVtt( "payload/lang-tag/lang-without-annotation.vtt" );
const Node head = getHeadOfCue( 0 );
EXPECT_EQ( 1, head[ 0 ].childCount() );
EXPECT_EQ( Node::Lang, head[ 0 ].kind() );
EXPECT_STREQ( "", head[ 0 ].lang().utf8() );
}
/*
* Verifies that a voice cue text tag that does not have an end tag is parsed correctly.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-voice-span (11/27/2012)
* Voice tags consist of:
* 1. A cue span start tag "v" that requires an annotation.
* 2. Cue internal text representing the voice text.
* 3. A cue span end tag that can be ommitted for brevity if the voice span is the only tag in this webvtt cue text sequence.
*/
TEST_F(PayloadVoiceTag,DISABLED_VoiceTag)
{
loadVtt( "payload/v-tag/v-tag.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 1 );
ASSERT_EQ( Node::Voice, head->child( 0 )->kind() );
}
TEST_F(PayloadLangTag, BasicLangTag)
{
loadVtt( "payload/lang-tag/basic-lang.vtt" );
const Node head = getHeadOfCue( 0 );
EXPECT_EQ( 1, head.childCount() );
EXPECT_EQ( Node::Lang, head[ 0 ].kind() );
EXPECT_STREQ( "en", head[ 0 ].lang().utf8() );
}
TEST_F(PayloadLangTag, TwoLangTags)
{
loadVtt( "payload/lang-tag/two-lang-tags.vtt" );
const Node head = getHeadOfCue( 0 );
EXPECT_EQ( 2, head.childCount() );
EXPECT_EQ( Node::Lang, head[ 0 ].kind() );
EXPECT_STREQ( "en", head[ 0 ].lang().utf8() );
EXPECT_EQ( Node::Lang, head[ 1 ].kind() );
EXPECT_STREQ( "fe", head[ 1 ].lang().utf8() );
}
/*
* Verifies that a voice cue text tag that has an annotation is parsed correctly and contains the annotation.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-voice-span (11/27/2012)
* Voice tags consist of:
* 1. A cue span start tag "v" that requires an annotation.
* 2. Cue internal text representing the voice text.
* 3. A cue span end tag that can be ommitted for brevity if the voice span is the only tag in this webvtt cue text sequence.
*/
TEST_F(PayloadVoiceTag,DISABLED_VoiceTagAnnotation)
{
loadVtt( "payload/v-tag/v-tag-annotation.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 3 );
ASSERT_EQ( Node::Voice, head->child( 1 )->kind() );
String expectedString = String( (const byte *)"Annotation", 10 );
ASSERT_EQ( expectedString.text(), head->child( 1 )->toInternalNode()->annotation().text() );
}
/*
* Verifies that cue text end tags that are malformed will be ignored.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text-parsing-rules step "If token is an end tag" (11/27/2012)
* Relevant cue text specification rule(s):
* 1. End tags that do not close the current open tag are ignored.
* 2. End tags that are not formatted properly will be ignored.
* Implications:
* 1. Tags that are not closed properly are valid until the end of the current cue text.
*/
TEST_F(PayloadTagFormat,DISABLED_EndTagNoEndBrace)
{
loadVtt( "payload/tag-format/end-tag-no-end-brace.vtt.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 2 );
const InternalNode *italicNode = head->child( 1 )->toInternalNode();
ASSERT_EQ( Node::Italic, italicNode->kind() );
ASSERT_TRUE( italicNode->childCount() == 1 );
ASSERT_EQ( Node::Text, italicNode->child( 0 )->kind() );
}
/*
* Verifies that cue text end tags that are out of order will be ignored.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text-parsing-rules step "If token is an end tag" (11/27/2012)
* Relevant cue text specification rule(s):
* 1. End tags that do not close the current open tag are ignored.
* 2. End tags that are not formatted properly will be ignored.
* Implications:
* 1. Tags that are not closed properly are valid until the end of the current cue text.
*/
TEST_F(PayloadTagFormat,DISABLED_BadTagNesting)
{
loadVtt( "payload/tag-format/bad-tag-nesting.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 2 );
const InternalNode *italicNode = head->child( 1 )->toInternalNode();
ASSERT_EQ( Node::Italic, italicNode->kind() );
ASSERT_EQ( Node::Bold, italicNode->kind() );
ASSERT_EQ( Node::Text, italicNode->child( 0 )->kind() );
}
/*
* Verifies that a single subclass can be attached to a cue text voice start tag.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-span-start-tag (11/27/2012)
* Cue span start tags consist of the following:
* 1. A "<" character representing the beginning of the start tag.
* 2. The tag name.
* 3. Zero or more the following sequence representing a subclasses of the start tag
* 3.1. A full stop "." character.
* 3.2. A sequence of non-whitespace characters.
* 4. If the start tag requires an annotation then a space or tab character followed by a sequence of
* non-whitespace characters representing the annotation.
* 5. A ">" character repsenting the end of the start tag.
*/
TEST_F(PayloadVoiceTag,DISABLED_VoiceTagSingleSubclass)
{
loadVtt( "payload/v-tag/v-tag-single-subclass.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 3 );
ASSERT_EQ( Node::Voice, head->child( 1 )->kind() );
StringList cssClasses = head->child( 1 )->toInternalNode()->cssClasses();
String expectedString = String( (const byte *)"class", 5 );
ASSERT_TRUE( cssClasses.length() == 1 );
ASSERT_EQ( expectedString.text(), cssClasses.stringAtIndex( 0 ).text() );
}
/*
* Verifies that cue text end tags that are malformed will be ignored.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text-parsing-rules step "If token is an end tag" (11/27/2012)
* Relevant cue text specification rule(s):
* 1. End tags that do not close the current open tag are ignored.
* 2. End tags that are not formatted properly will be ignored.
* Implications:
* 1. Tags that are not closed properly are valid until the end of the current cue text.
*/
TEST_F(PayloadTagFormat,DISABLED_MultiTagNoEndTag)
{
loadVtt( "payload/tag-format/multi-tag-no-end-tag.vtt" );
const InternalNode *head = getHeadOfCue( 0 );
ASSERT_TRUE( head->childCount() == 2 );
ASSERT_EQ( Node::Text, head->child( 0 )->kind() );
const InternalNode *italicNode0 = head->child( 1 )->toInternalNode();
ASSERT_TRUE( italicNode0->childCount() == 2 );
ASSERT_EQ( Node::Italic, italicNode0->kind() );
ASSERT_EQ( Node::Text, italicNode0->child( 0 )->kind() );
const InternalNode *italicNode1 = italicNode0->child( 1 )->toInternalNode();
ASSERT_EQ( Node::Italic, italicNode1->kind() );
ASSERT_EQ( Node::Text, italicNode1->child( 0 )->kind() );
}
/**
* Verifies that multiple cue component are parsed correctly.
*
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text (11/27/2012)
* Cue text text consists of one or more cue text components optionally
* separated by a single line terminator which can be:
* 1. CR (U+000D)
* 2. LF (U+000A)
* 3. CRLF pair
*/
TEST_F(PayloadFormat, MultilineMultipleCueTextTag)
{
loadVtt( "payload/format/multiline-multiple-cue-text-tag.vtt", 1 );
const Node head = getHeadOfCue( 0 );
ASSERT_LE( 1, head.childCount() );
const Node underlineTag = head[0];
EXPECT_EQ( Node::Underline, underlineTag.kind() );
ASSERT_LE( 2, underlineTag.childCount() );
EXPECT_EQ( Node::Text, underlineTag[0].kind() );
const Node italicTag = underlineTag[1];
EXPECT_EQ( Node::Italic, italicTag.kind() );
ASSERT_LE( 2, italicTag.childCount() );
EXPECT_EQ( Node::Text, italicTag[0].kind() );
const Node boldTag = italicTag[1];
EXPECT_EQ( Node::Bold, boldTag.kind() );
}
/*
* Verifies that a cue text span with no tags will parse correctly.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text-span (11/27/2012)
*/
TEST_F(PayloadFormat, BasicCueText)
{
loadVtt( "payload/format/basic-cue-text.vtt", 1 );
ASSERT_EQ( Node::Text, getHeadOfCue( 0 )[ 0 ].kind() );
}
/*
* Verifies that multiple cue components can be put in one line, one after
* the other.
* http://dev.w3.org/html5/webvtt/#webvtt-cue-components
*/
TEST_F(PayloadFormat, MultipleCueTextTagTogether)
{
loadVtt( "payload/format/multiple-cue-tag-together.vtt", 1 );
ASSERT_EQ( 4, getHeadOfCue( 0 ).childCount() );
}
/* Verifies that cue text with a malformed line terminator is still parsed correctly.
* From http://dev.w3.org/html5/webvtt/#webvtt-parser-algorithm (12/02/2012)
* The WebVTT parser algorithm is as follows:
* [...] 50. Bad cue: Discard cue.
*/
TEST_F(PayloadFormat, MultilineBasicCueTextExtraLine)
{
loadVtt( "payload/format/multiline-extra-line-terminator.vtt", 1);
ASSERT_EQ( Node::Text, getHeadOfCue( 0 )[ 0 ].kind() );
}
/*
* Verifies that cue text separated by a CRLF line terminator is parsed correctly.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text (12/02/2012)
* Cue text consists of one or more cue text components optionally separated by a single line terminator which can be:
* 1. CR (U+000D)
* 2. LF (U+000A)
* 3. CRLF pair
*/
TEST_F(PayloadFormat, MultilineBasicCueTextCRLF)
{
loadVtt( "payload/format/multiline-basic-cue-text-crlf.vtt", 1 );
ASSERT_EQ( Node::Text, getHeadOfCue( 0 )[ 0 ].kind() );
}
/*
* Verifies that cue text with single CRLF pair characters will be parsed.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text (11/27/2012)
* Cue text text consists of one or more cue text components optionally separated by a single line terminator which can be:
* 1. CR (U+000D)
* 2. LF (U+000A)
* 3. CRLF pair
*/
TEST_F(PayloadFormat, MultilineCueTextCRLF)
{
loadVtt( "payload/format/multiline-cue-text-crlf.vtt", 1 );
ASSERT_EQ( 5, getHeadOfCue( 0 ).childCount() );
}
/*
Verifies that an incorrect class name cannot be used as a cue component.
* From http://dev.w3.org/html5/webvtt/#webvtt-cue-text-parsing-rules step "If token is an end tag" (11/27/2012)
* Relevant cue text specification rule(s):
* 1. End tags that do not close the current open tag are ignored.
* 2. End tags that are not formatted properly will be ignored.
* 3. Tags that do not have a valid tag name are ignored.
* Implications:
* 1. Tags that are not closed properly are valid until the end of the current cue text.
*/
TEST_F(PayloadTagFormat,DISABLED_BadTagName)
{
loadVtt( "payload/tag-format/incorrect-tag-name.vtt" );
ASSERT_TRUE( getHeadOfCue( 0 )->childCount() == 3 );
}
/*
* Verifies that a file with payload text containing a line feed in between the text
* will finish the parsing attempt gracefully.
* From http://dev.w3.org/html5/webvtt/#parsing (12/10/2012):
*
* ...
* 50. Bad cue: Discard cue.
* 51. Bad cue loop: If position is past the end of input, then jump to the step labeled end.
*/
TEST_F(PayloadFormat, MultilineCueTextExtraNewline)
{
loadVtt( "payload/format/multiline-cue-text-extra-newline.vtt", 1 );
ASSERT_EQ( 2, getHeadOfCue( 0 ).childCount() );
}
/*
* This test is to verify that multiple cue text tags can be nested within eachother.
* Based on the WebVTT cue components specification as of October 3, 2012.
* http://dev.w3.org/html5/webvtt/#webvtt-cue-components
*/
TEST_F(PayloadTagFormat,DISABLED_MultipleCueTextTag)
{
loadVtt( "payload/tag-format/multiple-cue-text-tag.vtt" );
ASSERT_EQ( Node::Italic, getHeadOfCue( 0 )->child( 0 )->kind() );
ASSERT_EQ( Node::Bold, getHeadOfCue( 0 )->child( 0 )->toInternalNode()->child( 0 )->kind() );
}