mirror of
https://github.com/storytold/craft-libs.git
synced 2026-11-13 06:04:40 +00:00
Dylan M. Taylor
58b02d4e6a
Carved from FilmCraft's filmcraft-captions (crates/captions at 4bec22bd): the file readers and writers, CEA-608/708 coding, format detection and the cue document. Rendering and the mapping onto FilmCraft's caption tracks stay there. - time: the crate has its own exact Tick (1/254016000000 s) and rational FrameRate, the subset of filmcraft-time the formats use. - TTML is read through craft-xml instead of roxmltree: no entity expansion, nesting bounded by the reader (128 levels, as before). - parse refuses input over 64 MiB and keeps at most 1,000,000 cues. - MCC and EBU STL name the caller's program (WriteOptions::program). - plain_text (inline tag stripping) moves with the SCC/MCC writers that need it.
234 lines
11 KiB
Rust
234 lines
11 KiB
Rust
use std::collections::{HashMap, HashSet};
|
||
|
||
use craft_captions::FrameRate;
|
||
use craft_captions::{Cue, Document, Format, WriteOptions, parse, write};
|
||
|
||
const TT: &str = "http://www.w3.org/ns/ttml";
|
||
const META: &str = "http://www.w3.org/ns/ttml#metadata";
|
||
|
||
#[test]
|
||
fn imports_speaker_names_from_significant_ttml_metadata() {
|
||
for format in [Format::Ttml, Format::Dfxp] {
|
||
for metadata in [
|
||
r#"<ttm:agent xml:id="speaker1" type="person"><ttm:name type="alias">Ann</ttm:name><ttm:name type="full">Ann Lee</ttm:name></ttm:agent>"#
|
||
.to_string(),
|
||
r#"<metadata><ttm:agent xml:id="speaker1" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent></metadata>"#.to_string(),
|
||
r#"<ttm:agent xml:id="speaker1" type="person"><ttm:name type="full">Ann<!-- editorial note --> Lee</ttm:name></ttm:agent>"#.to_string(),
|
||
] {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}"><head>{metadata}</head><body><div><p begin="1s" end="2s" ttm:agent="speaker1">Hello</p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), format).unwrap();
|
||
assert_eq!(doc.cues.len(), 1);
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann Lee"));
|
||
assert_eq!(doc.cues[0].text, "Hello");
|
||
}
|
||
}
|
||
}
|
||
|
||
#[test]
|
||
fn exported_speakers_have_unique_resolvable_ids_and_exact_names() {
|
||
let rate = FrameRate::FPS_25;
|
||
let speakers = [
|
||
"Ann Lee",
|
||
"Ann_Lee",
|
||
"O’Connor & <Guest>",
|
||
"_Ann",
|
||
"山田 花子",
|
||
"Ann Lee",
|
||
"",
|
||
" Ann Lee ",
|
||
"Ann\rLee",
|
||
"Ann\r\nLee",
|
||
"Ann\nLee",
|
||
"Ann\tLee",
|
||
];
|
||
let ids = ["agent1", "agent2", "speaker1", "speaker2", "c5", "c6", "c7", "c8", "c9", "c10", "c11", "c12"];
|
||
let doc = Document {
|
||
cues: speakers
|
||
.iter()
|
||
.enumerate()
|
||
.map(|(i, name)| Cue {
|
||
start: rate.tick_of(i as i64 * 50),
|
||
end: rate.tick_of(i as i64 * 50 + 25),
|
||
text: format!("Line {}", i + 1),
|
||
speaker: Some((*name).into()),
|
||
id: Some(ids[i].into()),
|
||
..Default::default()
|
||
})
|
||
.collect(),
|
||
..Default::default()
|
||
};
|
||
for format in [Format::Ttml, Format::Dfxp] {
|
||
let bytes = write(&doc, format, WriteOptions { rate: Some(rate), ..Default::default() });
|
||
assert_eq!(bytes, write(&doc, format, WriteOptions { rate: Some(rate), ..Default::default() }), "stable output");
|
||
let text = std::str::from_utf8(&bytes).unwrap();
|
||
// names by namespace: tt, ttm and xml prefixes whatever the writer used
|
||
let canonical = craft_xml::Canonical::new(&[("tt", TT), ("ttm", META)]);
|
||
let xml = craft_xml::Options::canonical(&canonical).strict().parse(text.as_bytes()).unwrap();
|
||
let mut all_ids = HashSet::new();
|
||
for id in xml.descendants().filter_map(|n| n.attr("xml:id")) {
|
||
assert!(all_ids.insert(id), "duplicate xml:id {id}");
|
||
}
|
||
let mut agents = HashMap::new();
|
||
for agent in xml.find_all("ttm:agent") {
|
||
assert_eq!(agent.attr("type"), Some("other"));
|
||
let name = agent.child("ttm:name").unwrap();
|
||
assert_eq!(name.attr("type"), Some("full"));
|
||
assert_eq!(name.attr("xml:space"), Some("preserve"));
|
||
agents.insert(agent.attr("xml:id").unwrap(), name.text());
|
||
}
|
||
assert_eq!(agents.len(), 11, "each exact name is defined once");
|
||
let paragraphs: Vec<_> = xml.find_all("tt:p").collect();
|
||
assert_eq!(paragraphs.len(), speakers.len());
|
||
for (paragraph, expected) in paragraphs.iter().zip(speakers) {
|
||
let id = paragraph.attr("ttm:agent").unwrap();
|
||
assert_eq!(agents.get(id).map(|n| n.as_ref()), Some(expected), "reference {id}");
|
||
}
|
||
let back = parse(&bytes, format).unwrap();
|
||
assert_eq!(back.cues, doc.cues, "{format:?} names and cue content survive");
|
||
}
|
||
}
|
||
|
||
#[test]
|
||
fn imports_multiple_agent_names_in_reference_order() {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}"><head>
|
||
<ttm:agent xml:id="ann" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent>
|
||
<ttm:agent xml:id="guest" type="person"><ttm:name type="alias">O’Connor</ttm:name></ttm:agent>
|
||
</head><body><div><p begin="1s" end="2s" ttm:agent="guest ann">Together</p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("O’Connor, Ann Lee"));
|
||
}
|
||
|
||
#[test]
|
||
fn keeps_legacy_underscore_speaker_references_readable() {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}"><body><div>
|
||
<p begin="1s" end="2s" ttm:agent="Ann_Lee">Hello</p>
|
||
</div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann Lee"));
|
||
}
|
||
|
||
#[test]
|
||
fn timed_span_can_name_its_own_speaker() {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}"><head>
|
||
<ttm:agent xml:id="ann" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent>
|
||
<ttm:agent xml:id="guest" type="person"><ttm:name type="full">Guest</ttm:name></ttm:agent>
|
||
</head><body><div><p ttm:agent="ann"><span begin="1s" end="2s" ttm:agent="guest">Hello</span>
|
||
<span begin="3s" end="4s">Welcome</span></p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues.len(), 2);
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Guest"));
|
||
assert_eq!(doc.cues[1].speaker.as_deref(), Some("Ann Lee"));
|
||
}
|
||
|
||
#[test]
|
||
fn foreign_metadata_does_not_replace_ttml_speakers() {
|
||
let real = r#"<ttm:agent xml:id="speaker1" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent>"#;
|
||
for (metadata, attributes) in [
|
||
(
|
||
format!(r#"{real}<metadata><app:agent id="speaker1"><app:name type="full">Library editor</app:name></app:agent></metadata>"#),
|
||
r#"ttm:agent="speaker1""#,
|
||
),
|
||
(real.into(), r#"app:agent="library" ttm:agent="speaker1""#),
|
||
] {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}" xmlns:app="urn:example:caption-library"><head>{metadata}</head><body><div><p begin="1s" end="2s" {attributes}>Hello</p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann Lee"), "{xml}");
|
||
}
|
||
// A TTML agent nested inside another application's metadata container is not significant.
|
||
let xml = format!(
|
||
r#"<tt xmlns="{TT}" xmlns:ttm="{META}" xmlns:app="urn:example:caption-library"><head><app:metadata>{real}</app:metadata></head><body><div><p begin="1s" end="2s" ttm:agent="speaker1">Hello</p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("speaker1"));
|
||
}
|
||
|
||
#[test]
|
||
fn imports_draft_metadata_and_legacy_unqualified_speaker_references() {
|
||
for (tt, metadata, attributes) in [
|
||
("http://www.w3.org/2006/10/ttaf1", "http://www.w3.org/2006/10/ttaf1#metadata", r#"ttm:agent="speaker1""#),
|
||
(TT, META, r#"agent="speaker1""#),
|
||
] {
|
||
let xml = format!(
|
||
r#"<tt xmlns="{tt}" xmlns:ttm="{metadata}"><head><ttm:agent xml:id="speaker1" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent></head><body><div><p begin="1s" end="2s" {attributes}>Hello</p></div></body></tt>"#
|
||
);
|
||
let doc = parse(xml.as_bytes(), Format::Ttml).unwrap();
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann Lee"));
|
||
}
|
||
}
|
||
|
||
#[test]
|
||
fn keeps_april_2006_draft_speaker_references_readable() {
|
||
let xml = r#"<?xml version="1.0" encoding="UTF-8"?>
|
||
<tt xml:lang="en" xmlns="http://www.w3.org/2006/04/ttaf1" xmlns:ttm="http://www.w3.org/2006/04/ttaf1#metadata">
|
||
<head><metadata>
|
||
<ttm:agent id="Ann_Lee" type="person"><ttm:name type="full">Ann Lee</ttm:name></ttm:agent>
|
||
</metadata></head>
|
||
<body><div><p begin="1s" end="2s" ttm:agent="Ann_Lee">Hello</p></div></body>
|
||
</tt>"#;
|
||
for format in [Format::Ttml, Format::Dfxp] {
|
||
let doc = parse(xml.as_bytes(), format).unwrap();
|
||
assert_eq!(doc.cues.len(), 1);
|
||
assert_eq!(doc.cues[0].text, "Hello");
|
||
assert_eq!(doc.cues[0].start, FrameRate::FPS_25.tick_of(25));
|
||
assert_eq!(doc.cues[0].end, FrameRate::FPS_25.tick_of(50));
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann Lee"), "the documented draft keeps its previously readable speaker");
|
||
}
|
||
}
|
||
|
||
#[test]
|
||
fn reads_april_2006_agent_definitions_with_unqualified_ids() {
|
||
let xml = r#"<tt xml:lang="en" xmlns="http://www.w3.org/2006/04/ttaf1" xmlns:ttm="http://www.w3.org/2006/04/ttaf1#metadata">
|
||
<head><ttm:agent id="speaker1" type="person"><ttm:name type="full">Ann-Marie Lee</ttm:name></ttm:agent></head>
|
||
<body><div><p begin="1s" end="2s" ttm:agent="speaker1">Welcome</p></div></body></tt>"#;
|
||
for format in [Format::Ttml, Format::Dfxp] {
|
||
let doc = parse(xml.as_bytes(), format).unwrap();
|
||
assert_eq!(doc.cues.len(), 1);
|
||
assert_eq!(doc.cues[0].text, "Welcome");
|
||
assert_eq!(doc.cues[0].speaker.as_deref(), Some("Ann-Marie Lee"), "the name resolves through the draft's unqualified id");
|
||
}
|
||
}
|
||
|
||
fn assert_full_speaker_name_with_type(attributes: &str) {
|
||
let xml = format!(
|
||
r#"<tt xml:lang="en" xmlns="{TT}" xmlns:ttm="{META}" xmlns:app="urn:example:caption-library">
|
||
<head><ttm:agent xml:id="speaker1" type="person">
|
||
<ttm:name type="alias">Ann</ttm:name><ttm:name {attributes}>Ann Lee</ttm:name>
|
||
</ttm:agent></head><body><div><p begin="1s" end="2s" ttm:agent="speaker1">Hello</p></div></body></tt>"#
|
||
);
|
||
let mut speakers = Vec::new();
|
||
for format in [Format::Ttml, Format::Dfxp] {
|
||
let doc = parse(xml.as_bytes(), format).unwrap();
|
||
assert_eq!(doc.cues.len(), 1);
|
||
assert_eq!(doc.cues[0].text, "Hello");
|
||
assert_eq!(doc.cues[0].start, FrameRate::FPS_25.tick_of(25));
|
||
assert_eq!(doc.cues[0].end, FrameRate::FPS_25.tick_of(50));
|
||
speakers.push(doc.cues[0].speaker.clone());
|
||
}
|
||
assert_eq!(speakers, [Some("Ann Lee".into()), Some("Ann Lee".into())], "TTML and DFXP prefer the full name: {attributes}");
|
||
}
|
||
|
||
#[test]
|
||
fn name_type_without_application_attribute_prefers_full_name() {
|
||
assert_full_speaker_name_with_type(r#"type="full""#);
|
||
}
|
||
|
||
#[test]
|
||
fn name_type_after_application_attribute_prefers_full_name() {
|
||
assert_full_speaker_name_with_type(r#"app:type="contact" type="full""#);
|
||
}
|
||
|
||
#[test]
|
||
fn name_type_before_application_attribute_prefers_full_name() {
|
||
assert_full_speaker_name_with_type(r#"type="full" app:type="contact""#);
|
||
}
|