1142 lines
39 KiB
Rust
1142 lines
39 KiB
Rust
use super::*;
|
|
|
|
const ONE_CHANNEL: [FeedRequest; 1] = [FeedRequest {
|
|
channel: 0,
|
|
start: 1,
|
|
count: PAGE_SIZE,
|
|
}];
|
|
|
|
fn full_progress(requests: &[FeedRequest]) -> Vec<FeedProgress> {
|
|
requests
|
|
.iter()
|
|
.map(|&request| FeedProgress {
|
|
request,
|
|
count: request.count,
|
|
})
|
|
.collect()
|
|
}
|
|
|
|
#[test]
|
|
fn feed_cursor_shares_60_rows_across_17_channels_with_continuous_uneven_offsets() {
|
|
let mut cursor = FeedCursor::default();
|
|
assert!(!cursor.has_more());
|
|
let mut offsets = [1; 17];
|
|
for _ in 0..3 {
|
|
let requests = cursor.plan(17).unwrap();
|
|
let planned = cursor.clone();
|
|
assert_eq!(
|
|
cursor.plan(17).unwrap(),
|
|
requests,
|
|
"retrying must preserve the plan"
|
|
);
|
|
assert_eq!(
|
|
cursor, planned,
|
|
"planning must not advance offsets or rotation"
|
|
);
|
|
assert_eq!(requests.len(), 17);
|
|
assert_eq!(
|
|
requests.iter().map(|request| request.count).sum::<usize>(),
|
|
PAGE_SIZE
|
|
);
|
|
assert_eq!(
|
|
requests.iter().filter(|request| request.count == 4).count(),
|
|
9
|
|
);
|
|
assert_eq!(
|
|
requests.iter().filter(|request| request.count == 3).count(),
|
|
8
|
|
);
|
|
assert!(requests.windows(2).all(|pair| {
|
|
(pair[0].start, pair[0].count, pair[0].channel)
|
|
< (pair[1].start, pair[1].count, pair[1].channel)
|
|
}));
|
|
for request in &requests {
|
|
assert_eq!(request.start, offsets[request.channel]);
|
|
offsets[request.channel] += request.count;
|
|
}
|
|
cursor.accept(&full_progress(&requests)).unwrap();
|
|
}
|
|
// Rotating the nine remainder rows avoids permanently favoring the first channels.
|
|
assert!(offsets.iter().max().unwrap() - offsets.iter().min().unwrap() <= 1);
|
|
let saved = cursor.clone();
|
|
assert!(cursor.plan(16).is_err());
|
|
assert_eq!(cursor, saved);
|
|
}
|
|
|
|
#[test]
|
|
fn exhausted_channel_quota_moves_to_survivors_without_restarting_offsets() {
|
|
let mut cursor = FeedCursor::default();
|
|
let first = cursor.plan(3).unwrap();
|
|
let mut progress = full_progress(&first);
|
|
progress
|
|
.iter_mut()
|
|
.find(|entry| entry.request.channel == 1)
|
|
.unwrap()
|
|
.count = 0;
|
|
cursor.accept(&progress).unwrap();
|
|
let second = cursor.plan(3).unwrap();
|
|
assert_eq!(
|
|
second,
|
|
[
|
|
FeedRequest {
|
|
channel: 0,
|
|
start: 21,
|
|
count: 30
|
|
},
|
|
FeedRequest {
|
|
channel: 2,
|
|
start: 21,
|
|
count: 30
|
|
},
|
|
]
|
|
);
|
|
let mut progress = full_progress(&second);
|
|
progress[0].count = 29;
|
|
cursor.accept(&progress).unwrap();
|
|
let last = cursor.plan(3).unwrap();
|
|
assert_eq!(
|
|
last,
|
|
[FeedRequest {
|
|
channel: 2,
|
|
start: 51,
|
|
count: PAGE_SIZE
|
|
}]
|
|
);
|
|
cursor
|
|
.accept(&[FeedProgress {
|
|
request: last[0],
|
|
count: 0,
|
|
}])
|
|
.unwrap();
|
|
assert!(!cursor.has_more());
|
|
assert!(cursor.plan(3).unwrap().is_empty());
|
|
let mut empty = FeedCursor::default();
|
|
assert!(empty.plan(0).unwrap().is_empty());
|
|
assert!(!empty.has_more());
|
|
}
|
|
|
|
#[test]
|
|
fn more_than_60_channels_rotate_fairly_and_unvisited_channels_do_not_exhaust() {
|
|
let mut cursor = FeedCursor::default();
|
|
let mut seen = [false; 137];
|
|
let mut offsets = vec![1; 137];
|
|
for _ in 0..3 {
|
|
let requests = cursor.plan(137).unwrap();
|
|
assert_eq!(requests.len(), PAGE_SIZE);
|
|
for request in &requests {
|
|
assert_eq!(request.count, 1);
|
|
assert_eq!(request.start, offsets[request.channel]);
|
|
seen[request.channel] = true;
|
|
offsets[request.channel] += 1;
|
|
}
|
|
cursor.accept(&full_progress(&requests)).unwrap();
|
|
}
|
|
assert!(seen.iter().all(|seen| *seen));
|
|
let mut cursor = FeedCursor::default();
|
|
for expected_remaining in [77, 17, 0] {
|
|
let requests = cursor.plan(137).unwrap();
|
|
assert_eq!(
|
|
requests.iter().map(|request| request.count).sum::<usize>(),
|
|
PAGE_SIZE
|
|
);
|
|
let progress: Vec<_> = requests
|
|
.iter()
|
|
.map(|&request| FeedProgress { request, count: 0 })
|
|
.collect();
|
|
cursor.accept(&progress).unwrap();
|
|
assert_eq!(
|
|
cursor
|
|
.positions
|
|
.as_ref()
|
|
.unwrap()
|
|
.iter()
|
|
.filter(|offset| offset.is_some())
|
|
.count(),
|
|
expected_remaining,
|
|
);
|
|
assert_eq!(cursor.has_more(), expected_remaining != 0);
|
|
}
|
|
assert!(cursor.plan(137).unwrap().is_empty());
|
|
}
|
|
|
|
#[test]
|
|
fn cursor_rejects_invalid_or_stale_progress_and_overflow_atomically() {
|
|
let mut cursor = FeedCursor::default();
|
|
assert!(cursor.accept(&[]).is_err());
|
|
let requests = cursor.plan(3).unwrap();
|
|
let valid = full_progress(&requests);
|
|
let saved = cursor.clone();
|
|
let mut bad = valid.clone();
|
|
bad[2].count += 1;
|
|
assert!(cursor.accept(&bad).is_err());
|
|
assert_eq!(cursor, saved);
|
|
bad = valid.clone();
|
|
bad[2].request.channel = 99;
|
|
assert!(cursor.accept(&bad).is_err());
|
|
assert_eq!(cursor, saved);
|
|
bad = valid.clone();
|
|
bad[2].request = bad[0].request;
|
|
assert!(cursor.accept(&bad).is_err());
|
|
assert_eq!(cursor, saved);
|
|
bad = valid.clone();
|
|
bad[2].request.start += 1;
|
|
assert!(cursor.accept(&bad).is_err());
|
|
assert_eq!(cursor, saved);
|
|
bad = valid.clone();
|
|
bad.reverse();
|
|
assert!(cursor.accept(&bad).is_err());
|
|
assert_eq!(cursor, saved);
|
|
assert!(cursor.accept(&valid[..2]).is_err());
|
|
assert_eq!(cursor, saved);
|
|
cursor.accept(&valid).unwrap();
|
|
let accepted = cursor.clone();
|
|
assert!(cursor.accept(&valid).is_err());
|
|
assert_eq!(cursor, accepted);
|
|
|
|
let mut overflow = FeedCursor {
|
|
positions: Some(vec![Some(1), Some(usize::MAX - 19), Some(1)]),
|
|
rotation: 0,
|
|
};
|
|
let requests = overflow.plan(3).unwrap();
|
|
let saved = overflow.clone();
|
|
assert!(overflow.accept(&full_progress(&requests)).is_err());
|
|
assert_eq!(overflow, saved);
|
|
}
|
|
|
|
#[cfg(unix)]
|
|
mod subprocesses {
|
|
use super::*;
|
|
|
|
fn controlled(script: &str) -> Command {
|
|
let mut command = isolated_command(OsStr::new("/bin/sh"));
|
|
command.args(["-c", &format!("kill -STOP $$; {script}")]);
|
|
command
|
|
}
|
|
|
|
fn wait_stopped(extraction: &Extraction) {
|
|
wait_stopped_pid(extraction.child.as_ref().unwrap().id());
|
|
}
|
|
|
|
fn wait_stopped_pid(pid: u32) {
|
|
let pid = pid as i32;
|
|
let mut status = 0;
|
|
loop {
|
|
// SAFETY: wait only for this test's owned child, without reaping its exit.
|
|
let result = unsafe { libc::waitpid(pid, &mut status, libc::WUNTRACED) };
|
|
if result == -1 && io::Error::last_os_error().kind() == io::ErrorKind::Interrupted {
|
|
continue;
|
|
}
|
|
assert_eq!(result, pid);
|
|
assert!(libc::WIFSTOPPED(status));
|
|
break;
|
|
}
|
|
}
|
|
|
|
fn release(extraction: &mut Extraction) {
|
|
let child = extraction.child.as_mut().unwrap();
|
|
// SAFETY: this child is stopped at the deterministic test barrier.
|
|
assert_eq!(unsafe { libc::kill(child.id() as i32, libc::SIGCONT) }, 0);
|
|
child.wait().unwrap();
|
|
// Synchronize pipe draining too, then leave the real result for Job::poll.
|
|
// Blocking is confined to tests; production polling never waits for output.
|
|
let result = extraction.output.recv().unwrap();
|
|
let (sender, receiver) = mpsc::channel();
|
|
sender.send(result).unwrap();
|
|
extraction.output = receiver;
|
|
}
|
|
|
|
fn take_output(extraction: &mut Extraction) -> Receiver<io::Result<(String, String)>> {
|
|
let (_, receiver) = mpsc::channel();
|
|
std::mem::replace(&mut extraction.output, receiver)
|
|
}
|
|
|
|
fn assert_reaped(pid: u32) {
|
|
let mut status = 0;
|
|
// SAFETY: query only this test's child; ECHILD proves cancellation reaped it.
|
|
assert_eq!(
|
|
unsafe { libc::waitpid(pid as i32, &mut status, libc::WNOHANG) },
|
|
-1
|
|
);
|
|
assert_eq!(
|
|
io::Error::last_os_error().raw_os_error(),
|
|
Some(libc::ECHILD)
|
|
);
|
|
}
|
|
|
|
#[test]
|
|
fn a_later_spawn_failure_cancels_and_reaps_already_started_batches() {
|
|
static NEXT: std::sync::atomic::AtomicU64 = std::sync::atomic::AtomicU64::new(0);
|
|
let directory = std::env::temp_dir().join(format!(
|
|
"tubular-extractor-spawn-{}-{}",
|
|
std::process::id(),
|
|
NEXT.fetch_add(1, std::sync::atomic::Ordering::Relaxed)
|
|
));
|
|
std::fs::create_dir(&directory).unwrap();
|
|
let ready = directory.join("ready");
|
|
let mut first = isolated_command(OsStr::new("/bin/sh"));
|
|
first
|
|
.args(["-c", "printf '%s' $$ > \"$1\"; kill -STOP $$", "test"])
|
|
.arg(&ready);
|
|
let mut started_pid = None;
|
|
let commands = std::iter::once(first).chain(std::iter::once_with(|| {
|
|
// The file only communicates identity; SIGSTOP is the completion barrier.
|
|
let pid = loop {
|
|
if let Ok(text) = std::fs::read_to_string(&ready)
|
|
&& let Ok(pid) = text.parse::<u32>()
|
|
{
|
|
break pid;
|
|
}
|
|
thread::yield_now();
|
|
};
|
|
wait_stopped_pid(pid);
|
|
started_pid = Some(pid);
|
|
isolated_command(directory.join("missing-extractor").as_os_str())
|
|
}));
|
|
let error = Job::spawn_all(commands).err().unwrap();
|
|
assert_eq!(error.kind(), io::ErrorKind::NotFound);
|
|
assert_reaped(started_pid.unwrap());
|
|
std::fs::remove_dir_all(directory).unwrap();
|
|
}
|
|
|
|
#[test]
|
|
fn queued_range_groups_keep_four_workers_and_preserve_parser_request_order() {
|
|
let requests: Vec<_> = (0..6)
|
|
.map(|channel| FeedRequest {
|
|
channel,
|
|
start: channel + 1,
|
|
count: channel + 1,
|
|
})
|
|
.collect();
|
|
let commands = requests.iter().map(|request| {
|
|
let output =
|
|
serde_json::json!({"entries": vec![serde_json::Value::Null; request.count]});
|
|
controlled(&format!("printf '%s' '{output}'"))
|
|
});
|
|
let mut job = Job::spawn_all(commands).unwrap();
|
|
assert_eq!(job.children.len(), MAX_EXTRACTORS);
|
|
assert_eq!(job.queued.len(), 2);
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
release(&mut job.children[3]);
|
|
assert!(job.poll().is_none());
|
|
assert_eq!(job.queued.len(), 1);
|
|
assert_eq!(
|
|
job.children
|
|
.iter()
|
|
.filter(|child| child.completed.is_none())
|
|
.count(),
|
|
MAX_EXTRACTORS
|
|
);
|
|
wait_stopped(&job.children[4]);
|
|
release(&mut job.children[4]);
|
|
release(&mut job.children[1]);
|
|
assert!(job.poll().is_none());
|
|
assert!(job.queued.is_empty());
|
|
assert!(
|
|
job.children
|
|
.iter()
|
|
.filter(|child| child.completed.is_none())
|
|
.count()
|
|
<= MAX_EXTRACTORS
|
|
);
|
|
wait_stopped(&job.children[5]);
|
|
release(&mut job.children[5]);
|
|
assert!(job.poll().is_none());
|
|
release(&mut job.children[2]);
|
|
assert!(job.poll().is_none());
|
|
release(&mut job.children[0]);
|
|
let output = job.poll().unwrap().unwrap();
|
|
let page = feed_results(&output, &requests).unwrap();
|
|
assert_eq!(page.feed_progress, full_progress(&requests));
|
|
assert!(page.videos.is_empty());
|
|
assert!(job.poll().is_none());
|
|
}
|
|
|
|
#[test]
|
|
fn queued_spawn_failure_discards_outputs_and_cancels_running_and_queued_groups() {
|
|
let mut job = Job::spawn_all([
|
|
controlled("printf '%s' 'completed output'"),
|
|
controlled("printf '%s' 'pending one'"),
|
|
controlled("printf '%s' 'pending two'"),
|
|
controlled("printf '%s' 'pending three'"),
|
|
isolated_command(OsStr::new("/dev/null/tubular-missing-extractor")),
|
|
controlled("printf '%s' 'must never start'"),
|
|
])
|
|
.unwrap();
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
let pids: Vec<_> = job.children[1..]
|
|
.iter()
|
|
.map(|child| child.child.as_ref().unwrap().id())
|
|
.collect();
|
|
let outputs: Vec<_> = job.children[1..].iter_mut().map(take_output).collect();
|
|
release(&mut job.children[0]);
|
|
assert!(
|
|
job.poll()
|
|
.unwrap()
|
|
.unwrap_err()
|
|
.starts_with("Cannot start yt-dlp:")
|
|
);
|
|
assert!(job.children.is_empty());
|
|
assert!(job.queued.is_empty());
|
|
for output in outputs {
|
|
assert!(output.recv().unwrap().is_ok());
|
|
}
|
|
for pid in pids {
|
|
assert_reaped(pid);
|
|
}
|
|
assert!(job.poll().is_none());
|
|
}
|
|
|
|
#[test]
|
|
fn aggregate_waits_for_all_batches_and_merges_in_request_not_completion_order() {
|
|
let mut job = Job::spawn_all([
|
|
controlled("printf '%s' '{\"batch\":1}'"),
|
|
controlled("printf '%s' '{\"batch\":2}'"),
|
|
controlled("printf '%s' '{\"batch\":3}'"),
|
|
])
|
|
.unwrap();
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
assert!(job.poll().is_none());
|
|
release(&mut job.children[2]);
|
|
assert!(
|
|
job.poll().is_none(),
|
|
"the last batch cannot complete a page"
|
|
);
|
|
release(&mut job.children[1]);
|
|
assert!(job.poll().is_none(), "the first batch is still outstanding");
|
|
release(&mut job.children[0]);
|
|
let output = job.poll().unwrap().unwrap();
|
|
let batches: Vec<_> = serde_json::Deserializer::from_str(&output)
|
|
.into_iter::<serde_json::Value>()
|
|
.map(Result::unwrap)
|
|
.collect();
|
|
assert_eq!(
|
|
batches,
|
|
[
|
|
serde_json::json!({"batch":1}),
|
|
serde_json::json!({"batch":2}),
|
|
serde_json::json!({"batch":3}),
|
|
]
|
|
);
|
|
assert!(job.poll().is_none());
|
|
}
|
|
|
|
#[test]
|
|
fn a_failed_group_discards_completed_output_and_cancels_running_and_queued_siblings() {
|
|
let mut job = Job::spawn_all([
|
|
controlled("printf '%s' '{\"complete\":true}'"),
|
|
controlled("printf '%s' '{\"partial\":true}'; printf '%s' 'batch failed' >&2; exit 7"),
|
|
controlled("/bin/sh -c 'kill -STOP $$' & wait"),
|
|
controlled("printf '%s' 'pending sibling'"),
|
|
controlled("printf '%s' 'queued sibling'"),
|
|
controlled("printf '%s' 'must never start'"),
|
|
])
|
|
.unwrap();
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
release(&mut job.children[0]);
|
|
assert!(job.poll().is_none());
|
|
wait_stopped(&job.children[4]);
|
|
assert_eq!(job.queued.len(), 1);
|
|
let pids: Vec<_> = job.children[2..]
|
|
.iter()
|
|
.map(|child| child.child.as_ref().unwrap().id())
|
|
.collect();
|
|
let outputs: Vec<_> = job.children[2..].iter_mut().map(take_output).collect();
|
|
release(&mut job.children[1]);
|
|
assert_eq!(job.poll().unwrap().unwrap_err(), "batch failed");
|
|
assert!(job.children.is_empty());
|
|
assert!(job.queued.is_empty());
|
|
for output in outputs {
|
|
assert!(output.recv().unwrap().is_ok());
|
|
}
|
|
for pid in pids {
|
|
assert_reaped(pid);
|
|
}
|
|
assert!(job.poll().is_none());
|
|
}
|
|
|
|
#[test]
|
|
fn unreadable_batch_output_fails_the_whole_page_and_cancels_siblings() {
|
|
let mut job = Job::spawn_all([
|
|
controlled("printf '\\377'"),
|
|
controlled("printf '%s' 'must not finish'"),
|
|
])
|
|
.unwrap();
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
let pending_pid = job.children[1].child.as_ref().unwrap().id();
|
|
let pending_output = take_output(&mut job.children[1]);
|
|
release(&mut job.children[0]);
|
|
assert!(
|
|
job.poll()
|
|
.unwrap()
|
|
.unwrap_err()
|
|
.starts_with("Cannot read yt-dlp output:")
|
|
);
|
|
assert!(pending_output.recv().unwrap().is_ok());
|
|
assert_reaped(pending_pid);
|
|
}
|
|
|
|
#[test]
|
|
fn dropping_a_page_cancels_and_reaps_running_groups_and_drops_queued_commands() {
|
|
let mut job = Job::spawn_all([
|
|
controlled("/bin/sh -c 'kill -STOP $$' & wait"),
|
|
controlled("printf '%s' 'must not finish'"),
|
|
controlled("printf '%s' 'pending two'"),
|
|
controlled("printf '%s' 'pending three'"),
|
|
controlled("printf '%s' 'queued one must never start'"),
|
|
controlled("printf '%s' 'queued two must never start'"),
|
|
])
|
|
.unwrap();
|
|
assert_eq!(job.children.len(), MAX_EXTRACTORS);
|
|
assert_eq!(job.queued.len(), 2);
|
|
for child in &job.children {
|
|
wait_stopped(child);
|
|
}
|
|
let pids: Vec<_> = job
|
|
.children
|
|
.iter()
|
|
.map(|child| child.child.as_ref().unwrap().id())
|
|
.collect();
|
|
let outputs: Vec<_> = job.children.iter_mut().map(take_output).collect();
|
|
drop(job);
|
|
for output in outputs {
|
|
assert!(output.recv().unwrap().is_ok());
|
|
}
|
|
for pid in pids {
|
|
assert_reaped(pid);
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn dropping_after_the_leader_exits_still_cancels_inherited_pipe_owners() {
|
|
let mut job = Job::spawn_all([controlled("/bin/sh -c 'kill -STOP $$' & exit 0")]).unwrap();
|
|
wait_stopped(&job.children[0]);
|
|
let child = job.children[0].child.as_mut().unwrap();
|
|
let pid = child.id();
|
|
// SAFETY: resume this test's stopped leader, which exits without its descendant.
|
|
assert_eq!(unsafe { libc::kill(pid as i32, libc::SIGCONT) }, 0);
|
|
assert!(child.wait().unwrap().success());
|
|
let output = take_output(&mut job.children[0]);
|
|
drop(job);
|
|
// EOF proves the descendant that inherited stdout/stderr was also canceled.
|
|
assert!(output.recv().unwrap().is_ok());
|
|
assert_reaped(pid);
|
|
}
|
|
}
|
|
#[test]
|
|
fn video_urls_preserve_provider_identity_and_reject_invalid_keys() {
|
|
assert_eq!(
|
|
video_url("jNQXAC9IVRw").unwrap(),
|
|
"https://www.youtube.com/watch?v=jNQXAC9IVRw"
|
|
);
|
|
assert_eq!(
|
|
video_url("miruro:opaque_ID-7:3").unwrap(),
|
|
"https://barelystarted.miruro.tv/watch/opaque_ID-7/episode?ep=3"
|
|
);
|
|
assert_eq!(
|
|
video_url("anilist:1:3").unwrap(),
|
|
"https://zokoanime.video/stream/ani/1/3/sub"
|
|
);
|
|
for id in [
|
|
"",
|
|
"--help",
|
|
"jNQXAC9IVRw&list=other",
|
|
"miruro:id:0",
|
|
"miruro:id:03",
|
|
"miruro:../id:3",
|
|
"anilist:01:3",
|
|
"anilist:1:03",
|
|
"anilist:1:0",
|
|
"anilist:1:2147483648",
|
|
] {
|
|
assert_eq!(
|
|
video_url(id).unwrap_err().kind(),
|
|
io::ErrorKind::InvalidInput
|
|
);
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn browser_open_rejects_unsafe_video_ids_before_spawning() {
|
|
for id in ["", "--help", "jNQXAC9IVRw&list=other", "日本語"] {
|
|
assert!(matches!(
|
|
open_in_browser(id),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn publication_dates_prefer_exact_metadata_and_handle_utc_boundaries_and_invalid_dates() {
|
|
for feed in [false, true] {
|
|
for (upload_date, timestamp, expected) in [
|
|
(Some("20240229"), Some(0), Some("2024-02-29")),
|
|
(Some("20000229"), None, Some("2000-02-29")),
|
|
(Some("19000229"), None, None),
|
|
(Some("20230229"), Some(0), Some("~1970-01-01")),
|
|
(Some("20241301"), None, None),
|
|
(Some("日本語"), None, None),
|
|
(None, Some(-1), Some("~1969-12-31")),
|
|
(None, Some(1_709_251_199), Some("~2024-02-29")),
|
|
(None, Some(1_709_251_200), Some("~2024-03-01")),
|
|
(None, Some(i64::MAX), None),
|
|
(None, None, None),
|
|
] {
|
|
let output = serde_json::json!({"entries":[{
|
|
"id": "jNQXAC9IVRw", "title": "Video",
|
|
"upload_date": upload_date, "timestamp": timestamp,
|
|
}]})
|
|
.to_string();
|
|
let page = if feed {
|
|
feed_results(&output, &ONE_CHANNEL)
|
|
} else {
|
|
search_results(&output)
|
|
}
|
|
.unwrap();
|
|
assert_eq!(page.videos[0].publication_date.as_deref(), expected);
|
|
}
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn thumbnails_bound_binary_input_and_reject_unsafe_ids() {
|
|
let bytes = [0xff, 0xd8, 0x00, 0xfe];
|
|
assert_eq!(read_thumbnail(bytes.as_slice()).unwrap(), bytes);
|
|
assert_eq!(
|
|
read_thumbnail(io::repeat(0).take(MAX_THUMBNAIL_BYTES as u64))
|
|
.unwrap()
|
|
.len(),
|
|
MAX_THUMBNAIL_BYTES
|
|
);
|
|
assert!(read_thumbnail(io::repeat(0)).is_err());
|
|
for id in ["../secrets", "abcdefghij/", "abcdefghijé", "--bad"] {
|
|
assert!(matches!(
|
|
ThumbnailJob::start(id),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
for url in [
|
|
"http://example.test/cover.jpg",
|
|
"file:///tmp/cover.jpg",
|
|
"ftp://example.test/cover.jpg",
|
|
"//example.test/cover.jpg",
|
|
"https://",
|
|
"https://user:secret@example.test/cover.jpg",
|
|
"https://example.test/\ncover.jpg",
|
|
] {
|
|
assert!(matches!(
|
|
ThumbnailJob::from_url(url),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn bad_video_entries_do_not_discard_valid_results() {
|
|
for feed in [false, true] {
|
|
for bad in [
|
|
r#"{"id":"UC4QobU6STFB0P71PMvOGN5A","title":"Channel result"}"#,
|
|
r#"{"id":"--bad","title":"Unsafe ID"}"#,
|
|
r#"{"id":"jNQXAC9IVRw","title":"Bad channel","channel_id":"malformed"}"#,
|
|
r#"{"id":"jNQXAC9IVRw","title":"Bad handle","uploader_id":"@jawed\u001b"}"#,
|
|
r#"{"id":42,"title":"Wrong ID type"}"#,
|
|
r#"{"id":"jNQXAC9IVRw","title":null}"#,
|
|
r#"{"id":"jNQXAC9IVRw","title":"Bad date type","timestamp":"yesterday"}"#,
|
|
r#"{"id":"jNQXAC9IVRw","title":"Bad channel type","channel":[]}"#,
|
|
r#"{"id":"jNQXAC9IVRw"}"#,
|
|
"null",
|
|
"42",
|
|
"[]",
|
|
"{}",
|
|
] {
|
|
let output = format!(
|
|
r#"{{"entries":[{{"id":"jNQXAC9IVRw","title":"Before"}},{bad},{{"id":"rQ_J9WH6CGk","title":"After"}}]}}"#,
|
|
);
|
|
let page = if feed {
|
|
feed_results(&output, &ONE_CHANNEL)
|
|
} else {
|
|
search_results(&output)
|
|
}
|
|
.unwrap_or_else(|error| panic!("bad entry {bad} discarded the entire page: {error}"));
|
|
assert_eq!(
|
|
page.videos
|
|
.iter()
|
|
.map(|video| video.title.as_str())
|
|
.collect::<Vec<_>>(),
|
|
["Before", "After"]
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn members_only_rows_are_hidden_without_exhausting_search_or_feed_pages() {
|
|
let members = serde_json::json!({
|
|
"id": "jNQXAC9IVRw", "title": "Members only", "availability": "subscriber_only"
|
|
});
|
|
let mut entries = vec![members; PAGE_SIZE];
|
|
let output = serde_json::json!({"entries": entries}).to_string();
|
|
let search = search_results(&output).unwrap();
|
|
assert!(search.videos.is_empty());
|
|
assert!(search.has_more);
|
|
let mut cursor = FeedCursor::default();
|
|
let requests = cursor.plan(1).unwrap();
|
|
let feed = feed_results(&output, &requests).unwrap();
|
|
assert!(feed.videos.is_empty());
|
|
assert!(feed.has_more);
|
|
cursor.accept(&feed.feed_progress).unwrap();
|
|
assert_eq!(cursor.plan(1).unwrap()[0].start, PAGE_SIZE + 1);
|
|
|
|
for (index, availability) in [
|
|
Some("public"),
|
|
Some("unlisted"),
|
|
None,
|
|
Some("new_availability"),
|
|
]
|
|
.into_iter()
|
|
.enumerate()
|
|
{
|
|
entries[index] = serde_json::json!({
|
|
"id": format!("{index:011}"),
|
|
"title": "Join my channel membership",
|
|
"availability": availability,
|
|
});
|
|
}
|
|
entries[4] = serde_json::json!({"id": "rQ_J9WH6CGk", "title": "No availability"});
|
|
let output = serde_json::json!({"entries": entries}).to_string();
|
|
for page in [
|
|
search_results(&output).unwrap(),
|
|
feed_results(&output, &cursor.plan(1).unwrap()).unwrap(),
|
|
] {
|
|
assert_eq!(
|
|
page.videos
|
|
.iter()
|
|
.map(|video| video.id.as_str())
|
|
.collect::<Vec<_>>(),
|
|
[
|
|
"00000000000",
|
|
"00000000001",
|
|
"00000000002",
|
|
"00000000003",
|
|
"rQ_J9WH6CGk"
|
|
]
|
|
);
|
|
assert!(page.has_more);
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn discarded_raw_entries_advance_the_cursor_only_after_acceptance() {
|
|
let bad = serde_json::json!({"id":"invalid", "title":"Bad", "timestamp":-500});
|
|
let mut entries = vec![bad; PAGE_SIZE];
|
|
let output = serde_json::json!({"entries": entries}).to_string();
|
|
let search = search_results(&output).unwrap();
|
|
assert!(search.videos.is_empty());
|
|
assert!(search.has_more);
|
|
assert!(search.feed_progress.is_empty());
|
|
let mut cursor = FeedCursor::default();
|
|
let requests = cursor.plan(1).unwrap();
|
|
let saved = cursor.clone();
|
|
let page = feed_results(&output, &requests).unwrap();
|
|
assert!(page.videos.is_empty());
|
|
assert!(page.has_more);
|
|
assert_eq!(page.feed_progress, full_progress(&requests));
|
|
assert_eq!(cursor, saved, "parsing alone must not advance the cursor");
|
|
cursor.accept(&page.feed_progress).unwrap();
|
|
assert_eq!(cursor.plan(1).unwrap()[0].start, PAGE_SIZE + 1);
|
|
|
|
entries[0] = serde_json::json!({"id":"jNQXAC9IVRw","title":"Valid","timestamp":100});
|
|
let page = feed_results(
|
|
&serde_json::json!({"entries": entries}).to_string(),
|
|
&cursor.plan(1).unwrap(),
|
|
)
|
|
.unwrap();
|
|
assert!(page.has_more);
|
|
assert_eq!(page.feed_progress[0].count, PAGE_SIZE);
|
|
assert_eq!(page.videos[0].id, "jNQXAC9IVRw");
|
|
}
|
|
|
|
#[test]
|
|
fn search_sanitizes_display_metadata_and_rejects_invalid_envelopes() {
|
|
let videos = search_results(
|
|
r#"{"entries":[null,{"id":"rQ_J9WH6CGk","title":"Rust\u001b","channel":null,"uploader_id":"@RustVideos"}]}"#,
|
|
)
|
|
.unwrap()
|
|
.videos;
|
|
assert_eq!(videos[0].title, "Rust");
|
|
assert_eq!(videos[0].channel, None);
|
|
assert_eq!(videos[0].channel_id, None);
|
|
assert_eq!(videos[0].uploader_id.as_deref(), Some("@RustVideos"));
|
|
for output in [
|
|
"not JSON",
|
|
"{}",
|
|
r#"{"entries":null}"#,
|
|
r#"{"entries":{}}"#,
|
|
r#"{"entries":[{"id":} ]}"#,
|
|
] {
|
|
assert!(search_results(output).is_err());
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn feeds_merge_newest_first_deduplicate_and_inherit_channel_identity() {
|
|
let output = r#"{"channel_id":"UCaYhcUwRBNscFNUKTjgPFiA","uploader_id":"@RustVideos","channel":"Rust\u001b","entries":[null,{"id":"l-GaaivVJCE","title":"Older","timestamp":100},{"id":"jNQXAC9IVRw","title":"Unknown","timestamp":null}]}
|
|
{"channel_id":"UCaYhcUwRBNscFNUKTjgPFiA","uploader_id":"@RustVideos","channel":"Rust","entries":[{"id":"SLV9AJxnb1U","title":"Newest","timestamp":300},{"id":"l-GaaivVJCE","title":"Newer","timestamp":200},{"id":"rQ_J9WH6CGk","title":"No metadata","channel_id":null}]}"#;
|
|
let requests = FeedCursor::default().plan(2).unwrap();
|
|
let page = feed_results(output, &requests).unwrap();
|
|
assert!(!page.has_more);
|
|
assert_eq!(
|
|
page.feed_progress
|
|
.iter()
|
|
.map(|entry| entry.count)
|
|
.collect::<Vec<_>>(),
|
|
[3, 3]
|
|
);
|
|
let videos = page.videos;
|
|
assert_eq!(
|
|
videos
|
|
.iter()
|
|
.map(|video| video.id.as_str())
|
|
.collect::<Vec<_>>(),
|
|
["SLV9AJxnb1U", "l-GaaivVJCE", "jNQXAC9IVRw", "rQ_J9WH6CGk"]
|
|
);
|
|
assert_eq!(videos[1].title, "Newer");
|
|
assert_eq!(videos[1].timestamp, Some(200));
|
|
assert_eq!(videos[2].timestamp, None);
|
|
assert_eq!(videos[2].channel.as_deref(), Some("Rust"));
|
|
assert!(
|
|
videos
|
|
.iter()
|
|
.all(|video| video.channel_id.as_deref() == Some("UCaYhcUwRBNscFNUKTjgPFiA"))
|
|
);
|
|
assert!(
|
|
videos
|
|
.iter()
|
|
.all(|video| video.uploader_id.as_deref() == Some("@RustVideos"))
|
|
);
|
|
let unavailable = feed_results(
|
|
r#"{"entries":[null,{"id":"jNQXAC9IVRw","title":"Unknown","channel_id":null}]}"#,
|
|
&ONE_CHANNEL,
|
|
)
|
|
.unwrap()
|
|
.videos;
|
|
assert_eq!(unavailable[0].channel_id, None);
|
|
assert!(feed_results(r#"{"channel_id":"malformed","entries":[]}"#, &ONE_CHANNEL).is_err());
|
|
assert!(feed_results("", &ONE_CHANNEL).is_err());
|
|
assert!(feed_results("{}", &ONE_CHANNEL).is_err());
|
|
assert!(feed_results("{not JSON}", &ONE_CHANNEL).is_err());
|
|
let exhausted = feed_results(r#"{"entries":[]}"#, &ONE_CHANNEL).unwrap();
|
|
assert!(exhausted.videos.is_empty());
|
|
assert!(!exhausted.has_more);
|
|
assert_eq!(exhausted.feed_progress[0].count, 0);
|
|
}
|
|
|
|
#[test]
|
|
fn channel_search_query_preserves_utf8_and_isolates_url_parameters() {
|
|
assert_eq!(
|
|
encode_query("café 🦀 &next=1#?/+%20\t\n\0 AZaz09-._~"),
|
|
"caf%C3%A9%20%F0%9F%A6%80%20%26next%3D1%23%3F%2F%2B%2520%09%0A%00%20AZaz09-._~"
|
|
);
|
|
}
|
|
|
|
#[test]
|
|
fn page_offsets_reject_zero_and_overflow_before_spawning() {
|
|
assert_eq!(page_end(1).unwrap(), PAGE_SIZE);
|
|
assert_eq!(page_end(PAGE_SIZE + 1).unwrap(), PAGE_SIZE * 2);
|
|
assert_eq!(page_end(usize::MAX - (PAGE_SIZE - 1)).unwrap(), usize::MAX);
|
|
let channels = vec!["https://www.youtube.com/@RustVideos".to_owned()];
|
|
for start in [0, usize::MAX, usize::MAX - (PAGE_SIZE - 2)] {
|
|
assert!(matches!(
|
|
Job::search("rust", start),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
assert!(matches!(
|
|
Job::feed(&channels, &[FeedRequest { start, ..ONE_CHANNEL[0] }], ""),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
assert!(matches!(
|
|
Job::feed(&[], &ONE_CHANNEL, ""),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
|
|
#[test]
|
|
fn feeds_reject_invalid_request_budgets_and_envelope_or_row_count_mismatches() {
|
|
let channels = vec!["https://www.youtube.com/@RustVideos".to_owned(); 2];
|
|
for requests in [
|
|
Vec::new(),
|
|
vec![FeedRequest {
|
|
count: 0,
|
|
..ONE_CHANNEL[0]
|
|
}],
|
|
vec![FeedRequest {
|
|
count: PAGE_SIZE + 1,
|
|
..ONE_CHANNEL[0]
|
|
}],
|
|
vec![FeedRequest {
|
|
channel: 2,
|
|
..ONE_CHANNEL[0]
|
|
}],
|
|
vec![
|
|
FeedRequest {
|
|
channel: 0,
|
|
start: 1,
|
|
count: 30,
|
|
},
|
|
FeedRequest {
|
|
channel: 0,
|
|
start: 2,
|
|
count: 30,
|
|
},
|
|
],
|
|
vec![
|
|
FeedRequest {
|
|
channel: 0,
|
|
start: 1,
|
|
count: 30,
|
|
},
|
|
FeedRequest {
|
|
channel: 1,
|
|
start: 1,
|
|
count: 30,
|
|
},
|
|
],
|
|
vec![
|
|
FeedRequest {
|
|
channel: 0,
|
|
start: 1,
|
|
count: 31,
|
|
},
|
|
FeedRequest {
|
|
channel: 1,
|
|
start: 1,
|
|
count: 30,
|
|
},
|
|
],
|
|
] {
|
|
assert!(matches!(
|
|
Job::feed(&channels, &requests, ""),
|
|
Err(error) if error.kind() == io::ErrorKind::InvalidInput
|
|
));
|
|
}
|
|
let requests = FeedCursor::default().plan(2).unwrap();
|
|
assert!(feed_results(r#"{"entries":[]}"#, &requests).is_err());
|
|
for output in [
|
|
r#"{"entries":[]} {"entries":[]}"#,
|
|
r#"{"entries":[]} garbage"#,
|
|
r#"{"entries":[null]} {"entries":null}"#,
|
|
r#"{"entries":null}"#,
|
|
r#"{"entries":{}}"#,
|
|
r#"{"entries":[{"id":}]}"#,
|
|
] {
|
|
assert!(feed_results(output, &ONE_CHANNEL).is_err());
|
|
}
|
|
let request = [FeedRequest {
|
|
channel: 0,
|
|
start: 1,
|
|
count: 2,
|
|
}];
|
|
assert!(feed_results(r#"{"entries":[null,null,null]}"#, &request).is_err());
|
|
assert!(feed_results(r#"{"entries":[]}"#, &[]).is_err());
|
|
assert!(
|
|
feed_results(
|
|
r#"{"entries":[]}"#,
|
|
&[FeedRequest {
|
|
count: 0,
|
|
..ONE_CHANNEL[0]
|
|
}]
|
|
)
|
|
.is_err()
|
|
);
|
|
assert!(
|
|
feed_results(
|
|
r#"{"entries":[]}"#,
|
|
&[FeedRequest {
|
|
count: PAGE_SIZE + 1,
|
|
..ONE_CHANNEL[0]
|
|
}]
|
|
)
|
|
.is_err()
|
|
);
|
|
}
|
|
|
|
#[test]
|
|
fn search_pagination_counts_raw_rows_and_preserves_relevance_order() {
|
|
for count in [0, PAGE_SIZE - 1, PAGE_SIZE, PAGE_SIZE + 1] {
|
|
let output = serde_json::json!({
|
|
"entries": vec![serde_json::Value::Null; count]
|
|
})
|
|
.to_string();
|
|
let page = search_results(&output).unwrap();
|
|
assert_eq!(page.has_more, count >= PAGE_SIZE);
|
|
assert!(page.videos.is_empty());
|
|
assert!(page.feed_progress.is_empty());
|
|
}
|
|
let mut entries = vec![serde_json::Value::Null; PAGE_SIZE];
|
|
entries[0] = serde_json::json!({
|
|
"id": "aaaaaaaaaaa", "title": "Relevant older video", "timestamp": 100
|
|
});
|
|
entries[PAGE_SIZE - 1] = serde_json::json!({
|
|
"id": "bbbbbbbbbbb", "title": "Less relevant newer video", "timestamp": 300
|
|
});
|
|
let page = search_results(&serde_json::json!({"entries": entries}).to_string()).unwrap();
|
|
assert!(page.has_more);
|
|
assert!(page.feed_progress.is_empty());
|
|
assert_eq!(page.videos.len(), 2);
|
|
assert_eq!(page.videos[0].id, "aaaaaaaaaaa");
|
|
assert_eq!(page.videos[1].id, "bbbbbbbbbbb");
|
|
}
|
|
|
|
#[test]
|
|
fn feed_pagination_counts_nulls_and_duplicates_before_filtering() {
|
|
for count in [0, PAGE_SIZE - 1, PAGE_SIZE] {
|
|
let output = serde_json::json!({
|
|
"entries": vec![serde_json::Value::Null; count]
|
|
})
|
|
.to_string();
|
|
let page = feed_results(&output, &ONE_CHANNEL).unwrap();
|
|
assert_eq!(page.has_more, count == PAGE_SIZE);
|
|
assert!(page.videos.is_empty());
|
|
assert_eq!(
|
|
page.feed_progress,
|
|
[FeedProgress {
|
|
request: ONE_CHANNEL[0],
|
|
count
|
|
}]
|
|
);
|
|
}
|
|
let entry = serde_json::json!({
|
|
"id": "aaaaaaaaaaa", "title": "Repeated", "timestamp": 123
|
|
});
|
|
let page = feed_results(
|
|
&serde_json::json!({"entries": vec![entry; PAGE_SIZE]}).to_string(),
|
|
&ONE_CHANNEL,
|
|
)
|
|
.unwrap();
|
|
assert!(page.has_more);
|
|
assert_eq!(page.feed_progress[0].count, PAGE_SIZE);
|
|
assert_eq!(page.videos.len(), 1);
|
|
|
|
let mut entries = vec![serde_json::Value::Null; PAGE_SIZE];
|
|
entries[0] = serde_json::json!({"id": "bbbbbbbbbbb", "title": "Unknown"});
|
|
let page = feed_results(
|
|
&serde_json::json!({"entries": entries}).to_string(),
|
|
&ONE_CHANNEL,
|
|
)
|
|
.unwrap();
|
|
assert!(page.has_more);
|
|
assert_eq!(page.feed_progress[0].count, PAGE_SIZE);
|
|
assert_eq!(page.videos.len(), 1);
|
|
assert_eq!(page.videos[0].timestamp, None);
|
|
}
|
|
|
|
#[test]
|
|
fn feeds_retain_short_and_undated_rows_without_a_global_frontier() {
|
|
let requests = FeedCursor::default().plan(5).unwrap();
|
|
let mut first = vec![serde_json::Value::Null; requests[0].count];
|
|
first[0] = serde_json::json!({"id":"aaaaaaaaaaa","title":"Older duplicate","timestamp":500});
|
|
first[1] = serde_json::json!({"id":"bbbbbbbbbbb","title":"First dated row","timestamp":300});
|
|
first[2] = serde_json::json!({"id":"ccccccccccc","title":"Unknown"});
|
|
let mut second = vec![serde_json::Value::Null; requests[1].count];
|
|
second[0] = serde_json::json!({"id":"aaaaaaaaaaa","title":"Newest duplicate","timestamp":600});
|
|
second[1] = serde_json::json!({"id":"ddddddddddd","title":"Second dated row","timestamp":400});
|
|
let short = serde_json::json!({"entries":[
|
|
{"id":"eeeeeeeeeee","title":"Short newer","timestamp":999}
|
|
]});
|
|
let older_short = serde_json::json!({"entries":[
|
|
{"id":"fffffffffff","title":"Short older","timestamp":50}
|
|
]});
|
|
let output = format!(
|
|
"{}\n{}\n{}\n{}\n{}",
|
|
serde_json::json!({"entries": first}),
|
|
serde_json::json!({"entries": second}),
|
|
short,
|
|
older_short,
|
|
serde_json::json!({"entries":[]})
|
|
);
|
|
let page = feed_results(&output, &requests).unwrap();
|
|
assert!(page.has_more);
|
|
assert_eq!(
|
|
page.feed_progress
|
|
.iter()
|
|
.map(|entry| entry.count)
|
|
.collect::<Vec<_>>(),
|
|
[12, 12, 1, 1, 0]
|
|
);
|
|
assert_eq!(
|
|
page.videos
|
|
.iter()
|
|
.map(|video| video.id.as_str())
|
|
.collect::<Vec<_>>(),
|
|
[
|
|
"eeeeeeeeeee",
|
|
"aaaaaaaaaaa",
|
|
"ddddddddddd",
|
|
"bbbbbbbbbbb",
|
|
"fffffffffff",
|
|
"ccccccccccc"
|
|
]
|
|
);
|
|
assert_eq!(page.videos[1].title, "Newest duplicate");
|
|
assert_eq!(page.videos.last().unwrap().timestamp, None);
|
|
|
|
let requests = FeedCursor::default().plan(3).unwrap();
|
|
let exhausted = feed_results(
|
|
&format!(
|
|
"{}\n{}\n{}",
|
|
short,
|
|
older_short,
|
|
serde_json::json!({"entries":[]})
|
|
),
|
|
&requests,
|
|
)
|
|
.unwrap();
|
|
assert!(!exhausted.has_more);
|
|
assert_eq!(exhausted.videos.len(), 2);
|
|
}
|
|
|
|
#[test]
|
|
fn channel_handles_require_the_requested_identity_and_a_valid_handle() {
|
|
let id = "UChAQQ9at2MDlXjvQX3vL3Yw";
|
|
for handle in ["@ch.1676", "@敷嶋てとら", "@hontame_"] {
|
|
let metadata = serde_json::json!({"channel_id": id, "uploader_id": handle}).to_string();
|
|
assert_eq!(channel_handle(&metadata, id).unwrap(), handle);
|
|
assert!(channel_handle(&metadata, "UCaYhcUwRBNscFNUKTjgPFiA").is_err());
|
|
}
|
|
for handle in [
|
|
None,
|
|
Some("ch.1676"),
|
|
Some("@bad/handle"),
|
|
Some("@bad\u{1b}"),
|
|
] {
|
|
let metadata = serde_json::json!({"channel_id": id, "uploader_id": handle}).to_string();
|
|
assert!(channel_handle(&metadata, id).is_err());
|
|
}
|
|
assert!(channel_handle("not JSON", id).is_err());
|
|
}
|
|
|
|
#[test]
|
|
fn full_descriptions_preserve_paragraphs_and_reject_mismatched_videos() {
|
|
let metadata = r#"{"id":"jNQXAC9IVRw","description":"First paragraph.\n\n日本語\tsecond paragraph.\u001b\nLast line."}"#;
|
|
assert_eq!(
|
|
full_description(metadata, "jNQXAC9IVRw").unwrap(),
|
|
"First paragraph.\n\n日本語\tsecond paragraph.\nLast line."
|
|
);
|
|
assert!(full_description(metadata, "rQ_J9WH6CGk").is_err());
|
|
assert!(full_description("not JSON", "jNQXAC9IVRw").is_err());
|
|
assert_eq!(
|
|
full_description(r#"{"id":"jNQXAC9IVRw","description":null}"#, "jNQXAC9IVRw").unwrap(),
|
|
""
|
|
);
|
|
}
|