Import pipeline: probe, hardlink, rename, layout (#82)

crates/arr-parse/src/lib.rs

@@ -332,6 +332,27 @@ pub fn parse(name: &str) -> NameClaims {
     claims
 }
 
+/// Collapse a release name to the blacklist key (`DESIGN.md` §6.3): lowercase,
+/// every run of non-alphanumerics as one dot. The same release reappearing
+/// with different separators or a different infohash still matches.
+#[must_use]
+pub fn normalise(name: &str) -> String {
+    let mut out = String::with_capacity(name.len());
+    let mut pending_gap = false;
+    for character in name.chars() {
+        if character.is_alphanumeric() {
+            if pending_gap && !out.is_empty() {
+                out.push('.');
+            }
+            pending_gap = false;
+            out.extend(character.to_lowercase());
+        } else {
+            pending_gap = true;
+        }
+    }
+    out
+}
+
 fn apply_marker(claims: &mut NameClaims, sources: &mut Vec<Source>, m: Marker) {
     match m {
         Marker::Resolution(r) => {

crates/arr-parse/tests/parse.rs

@@ -423,3 +423,15 @@ fn non_tv_names_claim_no_episode() {
         assert_eq!(parse(name).episode, None, "{name}");
     }
 }
+
+#[test]
+fn normalisation_is_separator_and_case_blind() {
+    let key = arr_parse::normalise("Dune.Part.Two.2024.2160p.WEB-DL.DDP5.1.Atmos");
+    assert_eq!(key, "dune.part.two.2024.2160p.web.dl.ddp5.1.atmos");
+    assert_eq!(
+        arr_parse::normalise("dune part two [2024] 2160p WEB DL DDP5 1 atmos "),
+        key,
+        "the same release under different separators is the same blacklist key"
+    );
+    assert_eq!(arr_parse::normalise("!!!"), "");
+}