mirror of
https://github.com/docling-project/docling-core.git
synced 2026-11-18 20:27:00 +00:00
fix(webvtt): escape cue text and voice names when exporting to WebVTT (#972)
* fix(webvtt): escape cue text and voice names when exporting to WebVTT Signed-off-by: Fahmid Arman <fahmid.brac@gmail.com> * test(webvtt): cover plain ampersand text and voice names in the export round trip Revert the serializer to the state on main, where the escaping was already added by #926, and keep the cases from the original change as parameters of test_webvtt_export_keeps_text_literal. Signed-off-by: Cesar Berrospi Ramis <ceb@zurich.ibm.com> --------- Signed-off-by: Fahmid Arman <fahmid.brac@gmail.com> Signed-off-by: Cesar Berrospi Ramis <ceb@zurich.ibm.com> Co-authored-by: Cesar Berrospi Ramis <ceb@zurich.ibm.com>
This commit is contained in:
1 parent
eb54920b13
commit
b11a1400a2
1 file changed
+24
-17
+24
-17
@@ -435,27 +435,34 @@ def test_webvtt_cue_html_character_references() -> None:
|
||||
assert str(vtt.cue_blocks[3]).rstrip("\n").endswith("R&foo;d &")
|
||||
|
||||
|
||||
def test_webvtt_export_keeps_text_literal() -> None:
|
||||
"""Test that document text resembling a character reference is exported literally."""
|
||||
@pytest.mark.parametrize(
|
||||
("text", "voice"),
|
||||
[
|
||||
("Ask © and & me", "Tom & Jerry > Co"),
|
||||
("x < y > z", "Tom & Jerry > Co"),
|
||||
("Tom & Jerry", None),
|
||||
("Hello", "Tom & Jerry > Co"),
|
||||
],
|
||||
)
|
||||
def test_webvtt_export_keeps_text_literal(text: str, voice: str | None) -> None:
|
||||
"""Test that document text and voice names resembling markup or references are exported literally."""
|
||||
doc = DoclingDocument(name="refs")
|
||||
texts = ["Ask © and & me", "x < y > z"]
|
||||
for idx, text in enumerate(texts):
|
||||
doc.add_text(
|
||||
label=DocItemLabel.TEXT,
|
||||
text=text,
|
||||
source=TrackSource(start_time=float(idx), end_time=idx + 1.0, voice="Tom & Jerry > Co"),
|
||||
)
|
||||
doc.add_text(
|
||||
label=DocItemLabel.TEXT,
|
||||
text=text,
|
||||
source=TrackSource(start_time=0.0, end_time=1.0, voice=voice),
|
||||
)
|
||||
|
||||
vtt = WebVTTFile.parse(doc.export_to_vtt())
|
||||
|
||||
assert len(vtt) == len(texts)
|
||||
for block, text in zip(vtt.cue_blocks, texts):
|
||||
voice = block.payload[0].component
|
||||
assert isinstance(voice, WebVTTCueVoiceSpan)
|
||||
assert voice.start_tag.annotation == "Tom & Jerry > Co"
|
||||
span = voice.internal_text.components[0].component
|
||||
assert isinstance(span, WebVTTCueTextSpan)
|
||||
assert span.text == text
|
||||
assert len(vtt) == 1
|
||||
component = vtt.cue_blocks[0].payload[0].component
|
||||
if voice is not None:
|
||||
assert isinstance(component, WebVTTCueVoiceSpan)
|
||||
assert component.start_tag.annotation == voice
|
||||
component = component.internal_text.components[0].component
|
||||
assert isinstance(component, WebVTTCueTextSpan)
|
||||
assert component.text == text
|
||||
|
||||
|
||||
def test_webvtt_cue_language_span_start_tag():
|
||||
|
||||
Reference in new issue
Block a user