// 一致性语料：解析、序列化、变换与校验的单一事实来源。
///|
fn check(failures : Array[String], name : String, condition : Bool) -> Unit {
if !condition {
failures.push(name)
}
}
///|
fn parse_srt_ok(text : String) -> Subtitle? {
match parse_srt(text) {
Ok(subtitle) => Some(subtitle)
Err(_) => None
}
}
///|
fn parse_vtt_ok(text : String) -> Subtitle? {
match parse_vtt(text) {
Ok(subtitle) => Some(subtitle)
Err(_) => None
}
}
///|
const SAMPLE_SRT : String = "1\n00:00:01,000 --> 00:00:04,000\nHello\n\n2\n00:00:05,000 --> 00:00:08,000\nWorld\n"
///|
const SAMPLE_VTT : String = "WEBVTT\n\n1\n00:00:01.000 --> 00:00:04.000\nHello\n\n2\n00:00:05.000 --> 00:00:08.000\nWorld\n"
///|
/// 一致性语料条数。
pub fn conformance_case_count() -> Int {
17
}
///|
/// 运行一致性语料，返回失败描述（空数组表示全部通过）。
pub fn run_conformance() -> Array[String] {
let failures : Array[String] = []
let simple = "1\n00:00:01,000 --> 00:00:04,000\nHello\n"
match parse_srt_ok(simple) {
None => failures.push("parse-simple: 解析失败")
Some(subtitle) => {
check(failures, "parse-simple: 条数", subtitle.cues.length() == 1)
check(failures, "parse-simple: 开始", subtitle.cues[0].start_ms == 1000)
check(failures, "parse-simple: 结束", subtitle.cues[0].end_ms == 4000)
check(failures, "parse-simple: 文本", subtitle.cues[0].text == "Hello")
}
}
match parse_srt_ok(SAMPLE_SRT) {
None => failures.push("parse-two: 解析失败")
Some(subtitle) => {
check(failures, "parse-two: 条数", subtitle.cues.length() == 2)
check(
failures,
"parse-two: 第二条开始",
subtitle.cues[1].start_ms == 5000,
)
}
}
let multiline = "1\n00:00:01,000 --> 00:00:04,000\nline one\nline two\n"
match parse_srt_ok(multiline) {
None => failures.push("parse-multiline: 解析失败")
Some(subtitle) =>
check(
failures,
"parse-multiline: 文本",
subtitle.cues[0].text == "line one\nline two",
)
}
match
parse_srt_ok("\u{feff}1\r\n00:00:01,000 --> 00:00:04,000\r\nHello\r\n") {
None => failures.push("parse-bom-crlf: 解析失败")
Some(subtitle) =>
check(
failures,
"parse-bom-crlf: 文本",
subtitle.cues[0].text == "Hello",
)
}
match parse_srt_ok(SAMPLE_SRT) {
None => failures.push("roundtrip: 解析失败")
Some(subtitle) => {
let roundtrip = parse_srt_ok(serialize_srt(subtitle))
match roundtrip {
None => failures.push("roundtrip: 二次解析失败")
Some(again) =>
check(failures, "roundtrip: 条数", again.cues.length() == 2)
}
}
}
match parse_vtt_ok(SAMPLE_VTT) {
None => failures.push("vtt-parse: 解析失败")
Some(subtitle) =>
check(failures, "vtt-parse: 条数", subtitle.cues.length() == 2)
}
match parse_vtt_ok(SAMPLE_VTT) {
None => failures.push("vtt-serialize: 解析失败")
Some(subtitle) => {
let output = serialize_vtt(subtitle)
check(failures, "vtt-serialize: 头", output.has_prefix("WEBVTT\n\n"))
check(failures, "vtt-serialize: 点号", output.contains("00:00:01.000"))
}
}
match parse_srt_ok(SAMPLE_SRT) {
None => failures.push("shift: 解析失败")
Some(subtitle) => {
let moved = shift(subtitle, 1500)
check(failures, "shift: 开始", moved.cues[0].start_ms == 2500)
check(failures, "shift: 结束", moved.cues[0].end_ms == 5500)
}
}
match parse_srt_ok(simple) {
None => failures.push("shift-clamp: 解析失败")
Some(subtitle) => {
let moved = shift(subtitle, -5000)
check(failures, "shift-clamp: 不为负", moved.cues[0].start_ms == 0)
}
}
match parse_srt_ok(simple) {
None => failures.push("scale: 解析失败")
Some(subtitle) => {
let scaled = scale(subtitle, 200)
check(failures, "scale: 开始", scaled.cues[0].start_ms == 2000)
check(failures, "scale: 结束", scaled.cues[0].end_ms == 8000)
}
}
let overlap = "1\n00:00:01,000 --> 00:00:05,000\nA\n\n2\n00:00:04,000 --> 00:00:08,000\nB\n"
match parse_srt_ok(overlap) {
None => failures.push("validate-overlap: 解析失败")
Some(subtitle) => {
let issues = validate(subtitle, ValidateOptions::new())
let mut found = false
for issue in issues {
if issue_kind_name(issue.kind) == "overlap" {
found = true
}
}
check(failures, "validate-overlap: 检出", found)
let fixed = resolve_overlaps(subtitle, 100)
let fixed_issues = validate(fixed, ValidateOptions::new())
let mut still_overlap = false
for issue in fixed_issues {
if issue_kind_name(issue.kind) == "overlap" {
still_overlap = true
}
}
check(failures, "resolve-overlap: 已消除", !still_overlap)
}
}
let empty = "1\n00:00:01,000 --> 00:00:04,000\n \n"
match parse_srt_ok(empty) {
None => failures.push("validate-empty: 解析失败")
Some(subtitle) => {
let issues = validate(subtitle, ValidateOptions::new())
let mut found = false
for issue in issues {
if issue_kind_name(issue.kind) == "empty-text" {
found = true
}
}
check(failures, "validate-empty: 检出", found)
}
}
let short = "1\n00:00:01,000 --> 00:00:01,200\nA\n"
match parse_srt_ok(short) {
None => failures.push("validate-short: 解析失败")
Some(subtitle) => {
let issues = validate(subtitle, ValidateOptions::new())
let mut found = false
for issue in issues {
if issue_kind_name(issue.kind) == "too-short" {
found = true
}
}
check(failures, "validate-short: 检出", found)
}
}
let long_line = "1\n00:00:01,000 --> 00:00:05,000\n\{long_text()}\n"
match parse_srt_ok(long_line) {
None => failures.push("validate-long: 解析失败")
Some(subtitle) => {
let issues = validate(subtitle, ValidateOptions::new())
let mut found = false
for issue in issues {
if issue_kind_name(issue.kind) == "line-too-long" {
found = true
}
}
check(failures, "validate-long: 检出", found)
}
}
match parse_srt("1\nnot-a-timestamp\nHello\n") {
Ok(_) => failures.push("invalid-timestamp: 不应解析成功")
Err(_) => ()
}
match parse_srt_ok("1\n01:02:03,004 --> 01:02:04,005\nA\n") {
None => failures.push("format: 解析失败")
Some(subtitle) =>
check(
failures,
"format: SRT 时间",
serialize_srt(subtitle).contains("01:02:03,004"),
)
}
match
parse_vtt(
"WEBVTT\n\nNOTE a comment\n\n1\n00:00:01.000 --> 00:00:04.000\nHello\n",
) {
Err(_) => failures.push("vtt-note: 解析失败")
Ok(subtitle) =>
check(failures, "vtt-note: 跳过 NOTE", subtitle.cues.length() == 1)
}
failures
}
///|
fn long_text() -> String {
let builder = StringBuilder()
let mut i = 0
while i < 60 {
builder.write_string("x")
i = i + 1
}
builder.to_string()
}