citefind: a colon inside the title, and an editor's credit after it

Kymlicka, Will, Politics in the Vernacular: Nationalism,
    Multiculturalism and Citizenship, Oxford, 2001

The imprint is found by counting commas, and the last comma-part carrying a
colon is taken to be it — which is right when the colon is `Place:
Publisher` and wrong when it is the title's own. Here it is the only colon in
the entry, and looking for it before the names have been read meant giving up
rather than falling through to the year at the end. The search starts after
the names now, so a colon that is part of the title is not mistaken for one.

    Juvenal, Saturae XIV, ed. J. D. Duff (Cambridge, 1962), p. 9.

A title that ran on into an editor's credit was refused outright. Trimming
the credit off is the repair that was wanted: the credit belongs to the
entry, not to the work's name, and the work is `Saturae XIV`.

crates/citefind/src/parse.rs

@@ -126,7 +126,6 @@ static NOT_A_TITLE: LazyLock<Regex> = LazyLock::new(|| {
     | ;                                # two entries run together
     | \bibid\b
     | \\                               # an escape conversion left behind
-    | ,\s+(?:eds?|trans)\.\s+\p{Lu}     # ran past the title into a credit
     "#,
     )
     .expect("a valid pattern")
@@ -672,29 +671,6 @@ fn comma_form(text: &str) -> Option<(Vec<String>, Option<String>, Option<String>
     if parts.len() < 4 {
         return None;
     }
-    // The imprint is whichever comma-part carries the colon. Found by
-    // counting commas rather than by where a pattern matched: a pattern for
-    // `Place: Publisher` starts at the comma before the place, which is
-    // exactly where the title ends, and the two cannot both have it.
-    // The last part with a colon, not the first: `Theories of Nationalism: A
-    // Critical Introduction, Basingstoke: Palgrave` has two, and the imprint
-    // is the one at the end.
-    // The imprint is whichever comma-part carries the colon — the last one,
-    // because a title has a colon in it as often as an imprint does. Failing
-    // that, an entry ending on a year has its imprint at the end: `Cheyne,
-    // T. K., Founders of Old Testament Criticism, London, 1893.`
-    let at = match parts.iter().rposition(|part| part.contains(':')) {
-        Some(at) => at,
-        None => {
-            let last = parts.last()?.trim().trim_end_matches('.');
-            if !looks_like_a_year(last) || parts.len() < 4 {
-                return None;
-            }
-            // The part before the year is where it was published, which is
-            // not a publisher and not kept.
-            parts.len() - 2
-        }
-    };
 
     // `Miller, David` is the usual opening, and `Procopius` on its own is the
     // other one. Whichever reads as a name, the title starts after it.
@@ -704,7 +680,37 @@ fn comma_form(text: &str) -> Option<(Vec<String>, Option<String>, Option<String>
         false if reads_as_names(parts[0]) => (authors(parts[0]), 1),
         false => return None,
     };
-    if names.is_empty() || at <= used {
+    if names.is_empty() {
+        return None;
+    }
+
+    // The imprint is whichever comma-part carries the colon, counted in
+    // commas rather than matched by a pattern: a pattern for `Place:
+    // Publisher` starts at the comma before the place, which is exactly
+    // where the title ends, and the two cannot both have it. The last such
+    // part, because a title has a colon in it as often as an imprint does —
+    // and if the only one is inside the title, the colon says nothing about
+    // where the imprint is and the year does.
+    let colon = parts
+        .iter()
+        .enumerate()
+        .rev()
+        .find(|(at, part)| *at > used && part.contains(':'))
+        .map(|(at, _)| at);
+    let at = match colon {
+        Some(at) => at,
+        None => {
+            // `Cheyne, T. K., Founders of Old Testament Criticism, London,
+            // 1893.` — the year is last and the place before it, which is
+            // not a publisher and is not kept.
+            let last = parts.last()?.trim().trim_end_matches('.');
+            if !looks_like_a_year(last) {
+                return None;
+            }
+            parts.len() - 2
+        }
+    };
+    if at <= used {
         return None;
     }
 
@@ -712,8 +718,7 @@ fn comma_form(text: &str) -> Option<(Vec<String>, Option<String>, Option<String>
     if title.chars().count() < 4 {
         return None;
     }
-    // Where the colon was is a publisher; where the year was is a place, and
-    // a place is not kept.
+    // Where the colon was is a publisher; where the year was is a place.
     let publisher = match parts[at].contains(':') {
         true => tidy_container(parts[at]),
         false => None,
@@ -994,7 +999,9 @@ fn tidy_marked_title(raw: &str) -> String {
     let text = text.trim_end_matches('\\');
     let text =
         text.trim_matches(|c: char| ",.;:'\u{2018}\u{2019}".contains(c) || c.is_whitespace());
-    let text = TRAILING_VOLUME.replace(text, "");
+    let without_volume = TRAILING_VOLUME.replace(text, "").into_owned();
+    let text = TRAILING_CREDIT.replace(&without_volume, "").into_owned();
+    let text = text.as_str();
     let text = text.trim_end_matches(|c: char| ",.;: ".contains(c));
     match MARKED_NOT_A_TITLE.is_match(text) || !quotes_balance(text) {
         true => String::new(),
@@ -1005,7 +1012,13 @@ fn tidy_marked_title(raw: &str) -> String {
 /// What a marked-up title still may not be: two entries run together, or a
 /// reference to where a work is rather than what it is.
 static MARKED_NOT_A_TITLE: LazyLock<Regex> = LazyLock::new(|| {
-    Regex::new(r"(?ix) \bpp?\.\s*\d | ; | \bibid\b | \\ | ,\s+(?:eds?|trans)\.\s+\p{Lu}")
+    Regex::new(r"(?ix) \bpp?\.\s*\d | ; | \bibid\b | \\").expect("a valid pattern")
+});
+
+/// An editor's or translator's credit following a title: `Saturae XIV, ed.
+/// J. D. Duff`. It belongs to the entry, not to the work's name.
+static TRAILING_CREDIT: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(?i),\s+(?:eds?|trans|translated\s+by|comp)\.?\s+\p{Lu}.*$")
         .expect("a valid pattern")
 });
 
@@ -1028,7 +1041,9 @@ fn tidy_title(raw: &str) -> String {
         text.trim_matches(|c: char| ",.;:'\u{2018}\u{2019}".contains(c) || c.is_whitespace());
     // `Geschichte des Papsttums, Vol. 2` names the work and then which volume
     // of it. The volume is where in the work, not what the work is called.
-    let text = TRAILING_VOLUME.replace(text, "");
+    let without_volume = TRAILING_VOLUME.replace(text, "").into_owned();
+    let text = TRAILING_CREDIT.replace(&without_volume, "").into_owned();
+    let text = text.as_str();
     let text = text.trim_end_matches(|c: char| ",.;: ".contains(c));
     // A run that opens on a page number, or carries one, is two entries run
     // together or a reference to a work rather than the name of one. Better
@@ -1775,6 +1790,30 @@ mod tests {
         assert!(bare.missing().is_empty());
     }
 
+    #[test]
+    fn a_colon_inside_the_title_does_not_hide_the_imprint() {
+        // The only colon here is the title's own, so it says nothing about
+        // where the imprint is — and the year at the end does.
+        let found = read(
+            "Kymlicka, Will, Politics in the Vernacular: Nationalism, \
+             Multiculturalism and Citizenship, Oxford, 2001",
+        );
+        assert_eq!(found.authors, ["Kymlicka, Will"]);
+        assert_eq!(
+            found.title.as_deref(),
+            Some("Politics in the Vernacular: Nationalism, Multiculturalism and Citizenship")
+        );
+        assert_eq!(found.year, Some(2001));
+    }
+
+    #[test]
+    fn an_editors_credit_is_trimmed_off_a_title_not_grounds_to_drop_it() {
+        let found = read("Juvenal, Saturae XIV, ed. J. D. Duff (Cambridge, 1962), p. 9.");
+        assert_eq!(found.authors, ["Juvenal"]);
+        assert_eq!(found.title.as_deref(), Some("Saturae XIV"));
+        assert_eq!(found.pages.as_deref(), Some("9"));
+    }
+
     #[test]
     fn a_continuation_is_flagged_and_not_resolved() {
         let found = read("Ibid., pp. 23-24, 39.");