bibclean: record faults, weighed, with the repairs that need no lookup

parity/src/bin/regen.rs

@@ -19,7 +19,7 @@ type Written = Result<Vec<String>, String>;
 type Job = (&'static str, fn() -> Written);
 
 fn main() -> ExitCode {
-    let jobs: [Job; 10] = [
+    let jobs: [Job; 11] = [
         ("tokens", tokens),
         ("unescape", unescape),
         ("chunkmd", chunkmd),
@@ -30,6 +30,7 @@ fn main() -> ExitCode {
         ("rrf", rrf),
         ("cuepack", cuepack),
         ("booksniff", booksniff),
+        ("bibclean", bibclean),
     ];
 
     for (name, job) in jobs {
@@ -73,6 +74,18 @@ fn unescape() -> Written {
     Ok(vec![path.display().to_string()])
 }
 
+fn bibclean() -> Written {
+    let dir = parity::golden("bibclean");
+    std::fs::create_dir_all(&dir).map_err(|e| e.to_string())?;
+
+    let cases = parity::dir().join("cases/bibclean/records.ndjson");
+    let cases = std::fs::read_to_string(&cases).map_err(|e| e.to_string())?;
+    let judgements = parity::run_oracle("bibclean.py", &cases)?;
+    let path = dir.join("records.ndjson");
+    std::fs::write(&path, judgements).map_err(|e| e.to_string())?;
+    Ok(vec![path.display().to_string()])
+}
+
 fn booksniff() -> Written {
     let dir = parity::golden("booksniff");
     std::fs::create_dir_all(&dir).map_err(|e| e.to_string())?;