Identity: a set's number before yEnc is no size; names after an arrow
- The unit-less size before the closing yEnc ('"x.mp3" - 8123 yEnc') took
the set number in 'Newstar Destiny I set 1 [1 of 60] "destiny-001-001.jpg"
yEnc' once the quotes were gone: 97 sets of 60 were one release. A number
of fewer than 4 digits after a word stays.
- unquotedName also cuts after "-> " ("… -> [0001 of 0076] ->
Exclusive-tgirls_000000-lg.jpg"): the arrow was part of the file name.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
18a146e130
commit
e1d21d1e7c
6 files changed
+37
-5
No files matched your search
+4
-1
@@ -338,7 +338,9 @@ pub fn unquoted_name(before_counter: &str) -> Option<String> {
|
||||
let bracket = text.rfind("] ").map_or(0, |i| i + 2);
|
||||
// A track or part number (/ - \d{1,3} - /): the name starts there; else after the last " - ".
|
||||
let dash = first_track(text).or_else(|| text.rfind(" - ")).map_or(0, |i| i + 3);
|
||||
let name = js::trim(&text[bracket.max(dash)..]);
|
||||
// Or after an arrow ("… -> Exclusive-tgirls_000000-lg.jpg").
|
||||
let arrow = text.rfind("-> ").map_or(0, |i| i + 3);
|
||||
let name = js::trim(&text[bracket.max(dash).max(arrow)..]);
|
||||
return (!name.is_empty() && has_extension(name)).then(|| name.to_string());
|
||||
}
|
||||
text.split(' ').find(|w| has_extension(w)).map(str::to_string)
|
||||
@@ -910,6 +912,7 @@ mod unquoted_tests {
|
||||
("A Strange Light from the East - Eastern Influenced Western Music (1965-72) - CD2 - 07 - Rascals - Sattva.mp3 ", Some("07 - Rascals - Sattva.mp3")),
|
||||
("pal! [Mivina high(1).wmv\" yEnc] ", Some("Mivina high(1).wmv")),
|
||||
("[03/21 Geroellheimer - S03E16.part01.rar\" yEnc ", Some("Geroellheimer - S03E16.part01.rar")),
|
||||
("<Group> FP -> {EXL01~P~004} -> Alexia (3) -> VF -> [0001 of 0076] -> Exclusive-tgirls_000000-lg.jpg ", Some("Exclusive-tgirls_000000-lg.jpg")),
|
||||
("[7w] - 9BbEV2MrTi59GBTEFlVrssBrXNwXf2SK9cQEeDjg yEnc ", None),
|
||||
("Re: Will somebody answer the question? ", None),
|
||||
];
|
||||
|
||||
+13
-1
@@ -360,6 +360,11 @@ fn strip_number_before_yenc(s: &str) -> String {
|
||||
if digits_start > 0 && (b[digits_start - 1].is_ascii_alphanumeric() || b[digits_start - 1] == b'_') {
|
||||
return s.to_string();
|
||||
}
|
||||
// A short number after a word is a set's ("… set 1 [1 of 60] "destiny-001-001.jpg" yEnc"), not a size.
|
||||
let word_before = b[..digits_start].iter().rev().find(|&&c| !js::is_space(c as char)).is_some_and(|c| c.is_ascii_alphanumeric());
|
||||
if digits_end - digits_start < 4 && word_before {
|
||||
return s.to_string();
|
||||
}
|
||||
format!("{}{}", &s[..digits_start], &s[digits_end..])
|
||||
}
|
||||
|
||||
@@ -1273,7 +1278,14 @@ mod image_set_tests {
|
||||
let a = release_identity(mb, r#"Herb Ernst@256 - NMR -[07/17] "Herb Ernst - Dreamflight II - 05 - The Light of Forever.mp3" - 8123 yEnc"#, "Herb Ernst - Dreamflight II - 05 - The Light of Forever.mp3");
|
||||
let b = release_identity(mb, r#"Herb Ernst@256 - NMR -[08/17] "Herb Ernst - Dreamflight II - 06 - Other.mp3" - 9001 yEnc"#, "Herb Ernst - Dreamflight II - 06 - Other.mp3");
|
||||
assert_eq!(a.key, b.key);
|
||||
for s in [r#"x - "a.mp3" - 8123 yEnc"#, "a8123 yEnc", "-8123 yEnc ", "yEnc"] {
|
||||
let destiny = |set: u32, i: u32| {
|
||||
let name = format!("destiny-{set:03}-{i:03}.jpg");
|
||||
release_identity("p", &format!(r#"Newstar Destiny I set {set} [{i} of 60] "{name}" yEnc"#), &name)
|
||||
};
|
||||
assert_eq!(destiny(1, 1).key, destiny(1, 2).key);
|
||||
assert_ne!(destiny(1, 1).key, destiny(63, 34).key);
|
||||
assert_eq!(destiny(63, 34).name, "Newstar Destiny I set 63");
|
||||
for s in [r#"x - "a.mp3" - 8123 yEnc"#, "a8123 yEnc", "-8123 yEnc ", "yEnc", r#"set 1 [1 of 60] "x.jpg" yEnc"#, "Album 12 yEnc", "Album 1234 yEnc"] {
|
||||
assert_eq!(super::subject_remainder(s), super::subject_remainder_re(s), "{s}");
|
||||
}
|
||||
assert!(random_poster("SOFZRw1-OTEd@EEFUbh.srG"));
|
||||
|
||||
@@ -63,7 +63,9 @@ export function unquotedName(beforeCounter: string): string | null {
|
||||
// is the album, the same for every track (LOG.md 2026-10-11 (10)); else after the last " - ".
|
||||
const track = / - \d{1,3} - /.exec(text)
|
||||
const dash = track ? track.index : text.lastIndexOf(' - ')
|
||||
const cut = Math.max(bracket === -1 ? 0 : bracket + 2, dash === -1 ? 0 : dash + 3)
|
||||
// Or after an arrow ("… -> [0001 of 0076] -> Exclusive-tgirls_000000-lg.jpg").
|
||||
const arrow = text.lastIndexOf('-> ')
|
||||
const cut = Math.max(bracket === -1 ? 0 : bracket + 2, dash === -1 ? 0 : dash + 3, arrow === -1 ? 0 : arrow + 3)
|
||||
const name = text.slice(cut).trim()
|
||||
return name && EXTENSION.test(name) ? name : null
|
||||
}
|
||||
|
||||
@@ -77,8 +77,11 @@ export function subjectRemainder(subject: string): string {
|
||||
// File sizes ("76.88 MB", "335555 Bytes", "336573 byt" cut off, "[347K]") differ per file.
|
||||
.replace(/\b\d+(?:[.,]\d+)?\s*(?:[kmgt]i?b|bytes?|byt)\b/gi, ' ')
|
||||
.replace(/\[\s*\d+(?:[.,]\d+)?\s*[kmgt]\s*\]/gi, ' ')
|
||||
// A size without a unit right before the closing yEnc ("…mp3" - 8123 yEnc").
|
||||
.replace(/\b\d+(\s+yEnc\s*)$/i, '$1')
|
||||
// A size without a unit right before the closing yEnc ("…mp3" - 8123 yEnc"); a short number after a word is a
|
||||
// set's ("… set 1 [1 of 60] "destiny-001-001.jpg" yEnc": 97 sets were one release; LOG.md 2026-10-12 (8)).
|
||||
.replace(/\b(\d+)(\s+yEnc\s*)$/i, (text, n: string, end: string, at: number, s: string) =>
|
||||
n.length >= 4 || !/[A-Za-z0-9]\s*$/.test(s.slice(0, at)) ? end : text
|
||||
)
|
||||
// File sizes follow "yEnc" at the end; other numbers (years, post ids) stay.
|
||||
.replace(/\byEnc\b(\s+\d+)?\s*$/i, ' ')
|
||||
.replace(/\byEnc\b/gi, ' ')
|
||||
|
||||
@@ -64,6 +64,7 @@ test.group('subject / parseSubject', () => {
|
||||
['A Strange Light from the East - Eastern Influenced Western Music (1965-72) - CD2 - 07 - Rascals - Sattva.mp3 (1/43)', '07 - Rascals - Sattva.mp3', 1, 43],
|
||||
["pal! [Mivina high(1).wmv\" yEnc] (323/349)", 'Mivina high(1).wmv', 323, 349],
|
||||
['[03/21 Geroellheimer - S03E16.part01.rar" yEnc (1/50)', 'Geroellheimer - S03E16.part01.rar', 1, 50],
|
||||
['<Group> FP -> {EXL01~P~004} -> Alexia (3) -> VF -> [0001 of 0076] -> Exclusive-tgirls_000000-lg.jpg (1/1)', 'Exclusive-tgirls_000000-lg.jpg', 1, 1],
|
||||
]
|
||||
for (const [subject, name, segment, total, fileIndex] of cases) {
|
||||
const parsed = parseSubject(subject)
|
||||
|
||||
@@ -251,6 +251,17 @@ test.group('release / par2 names', () => {
|
||||
]) assert.equal(imageSet(stem), set, stem)
|
||||
})
|
||||
|
||||
test("a set's number before the closing yEnc is no size", ({ assert }) => {
|
||||
const id = (subject: string, name: string) => releaseIdentity({ poster: 'p', subject, name, weak: false })!
|
||||
const destiny = (set: number, i: number) => {
|
||||
const name = `destiny-${String(set).padStart(3, '0')}-${String(i).padStart(3, '0')}.jpg`
|
||||
return id(`Newstar Destiny I set ${set} [${i} of 60] "${name}" yEnc`, name)
|
||||
}
|
||||
assert.equal(destiny(1, 1).key, destiny(1, 2).key)
|
||||
assert.notEqual(destiny(1, 1).key, destiny(63, 34).key)
|
||||
assert.equal(destiny(63, 34).name, 'Newstar Destiny I set 63')
|
||||
})
|
||||
|
||||
test('an image without words around it is keyed by its set', ({ assert }) => {
|
||||
const id = (name: string, counter: string) => releaseIdentity({ poster: 'p', subject: `##[ "${name}" (${counter})`, name, weak: false })!
|
||||
assert.equal(id('diana-040-001.jpg', '01/68').key, id('diana-040-002.jpg', '02/68').key)
|
||||
|
||||
Reference in new issue
Block a user