citefind: a note the translator signed is not a reference

Both volumes of *The Decline of the West* went from zero citations to 415
once the arrow marker was recognised — and 167 of those were the editor's and
the translator's own glosses:

    Latin: "nature of the thing." — Ed.
    Reference to the wars between Athens and Sparta. Their dates are
    431-404 BC and 31 BC. — Ed.

An em dash and a role at the very end is how a translator signs a note of
their own. It cites nothing, and it is a convention every academic
translation uses.

391 of the 415 also carried a stray `<!-- -->`: pandoc's list separator, left
in the middle of the reference text. HTML comments are conversion furniture
and are cut with the marker.

And `See R. Hirzel, Die Person (1914), p. 7.` read as an author-date entry
whose authors were "R. Hirzel, Die Person" and whose title was nothing — the
head is all capitalised words and the year is bracketed, which is exactly
what an author-date opening looks like. Where that reading yields no title
and reading it as a note yields one, the note wins. Having a title is how you
can tell which was meant.

crates/citefind/src/parse.rs

@@ -231,6 +231,14 @@ static ACKNOWLEDGEMENT: LazyLock<Regex> = LazyLock::new(|| {
     .expect("a valid pattern")
 });
 
+/// How a translator or an editor signs a note of their own: an em dash and
+/// their role, at the very end. `Latin: "nature of the thing." — Ed.` is a
+/// gloss on the text in front of it and cites nothing.
+static GLOSS_SIGNATURE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(?i)[—–-]{1,2}\s*(?:ed|tr|trans|transl|a\.\s*n|au)\.?\s*$")
+        .expect("a valid pattern")
+});
+
 /// A page reference or a bracketed imprint: what a citation has and a remark
 /// about the argument does not.
 static CITATION_ANCHOR: LazyLock<Regex> = LazyLock::new(|| {
@@ -358,6 +366,7 @@ pub fn one(text: &str, marker: Option<&str>) -> Option<Reference> {
         || BIOGRAPHY.is_match(text)
         || PRINT_FURNITURE.is_match(text)
         || ACKNOWLEDGEMENT.is_match(text)
+        || GLOSS_SIGNATURE.is_match(text)
     {
         return None;
     }
@@ -481,6 +490,21 @@ pub fn one(text: &str, marker: Option<&str>) -> Option<Reference> {
     if out.title.as_deref().is_some_and(str::is_empty) {
         out.title = None;
     }
+
+    // `See R. Hirzel, Die Person (1914), p. 7.` reads as an author-date entry
+    // — the head is all capitalised words and the year is in brackets — and
+    // then yields no title, because the title is sitting inside what was
+    // taken for the names. A note that names a work is the better reading of
+    // the same line, and having a title is how you can tell.
+    if out.title.is_none() {
+        if let Some((names, title, container)) = note_form(text) {
+            if title.is_some() {
+                out.authors = names;
+                out.title = title;
+                out.container = container.or(out.container.take());
+            }
+        }
+    }
     out.kind = kind(text, &out);
 
     Some(out)
@@ -1483,6 +1507,22 @@ mod tests {
         assert_eq!(found.year, Some(2000));
     }
 
+    #[test]
+    fn a_translators_note_signs_itself_and_cites_nothing() {
+        for gloss in [
+            "Latin: \"nature of the thing.\" — Ed.",
+            "Reference to the wars between Athens and Sparta, 431-404 BC. — Ed.",
+            "Meaning here names, dates and numbers, in the usual sense. — Tr.",
+        ] {
+            assert!(one(gloss, None).is_none(), "{gloss}");
+        }
+        // A note that cites a work is still a citation, sign-off or not.
+        let real = read("See R. Hirzel, Die Person (1914), p. 7.");
+        assert_eq!(real.authors, ["R. Hirzel"]);
+        assert_eq!(real.title.as_deref(), Some("Die Person"));
+        assert_eq!(real.year, Some(1914));
+    }
+
     #[test]
     fn a_short_note_names_a_work_without_a_year_or_a_publisher() {
         // The form a history book uses for every citation after the first:

crates/citefind/src/split.rs

@@ -112,6 +112,11 @@ static ARROW_MARKER: LazyLock<Regex> = LazyLock::new(|| {
     .expect("a valid pattern")
 });
 
+/// An HTML comment. Conversion leaves `<!-- -->` between list items and
+/// `<!-- Page 41 -->` where a page turned; neither is part of a reference.
+static COMMENT: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"<!--.*?-->").expect("a valid pattern"));
+
 /// A bullet or a blockquote marker, which conversion puts in front of a note
 /// often enough to matter. A leading `*` is left alone: it opens an italic
 /// title as often as it opens a list.
@@ -276,6 +281,7 @@ fn unnumbered(text: &str) -> Vec<Entry> {
 /// further along survives.
 fn strip_marker(raw: &str) -> Entry {
     let text = cut_footer(&tidy(raw));
+    let text = tidy(&COMMENT.replace_all(&text, " "));
     let text = BULLET.replace(&text, "").to_string();
     let (bare, from) = visible(&text, 160);
 
@@ -517,6 +523,24 @@ Another Anonymous Work. Cambridge, 2001.";
         assert_eq!(entry.text, "Holquist, Making War, pp. 211-12.");
     }
 
+    #[test]
+    fn the_back_reference_marker_an_epub_conversion_leaves() {
+        let plain = strip_marker("\\[←788 \\]\u{a0}  Krumbacher, Byz. Lit.-Gesch., p. 969.");
+        assert_eq!(plain.marker.as_deref(), Some("788"));
+        assert_eq!(plain.text, "Krumbacher, Byz. Lit.-Gesch., p. 969.");
+
+        let linked =
+            strip_marker("\\[[←193](index_split_011.xhtml#citation-193) \\]\u{a0}  Œd. Rex., 642.");
+        assert_eq!(linked.marker.as_deref(), Some("193"));
+        assert_eq!(linked.text, "Œd. Rex., 642.");
+    }
+
+    #[test]
+    fn a_comment_conversion_left_behind_is_not_part_of_a_reference() {
+        let entry = strip_marker("1. Smith, *A Title* (Oxford, 1998). <!-- -->");
+        assert_eq!(entry.text, "Smith, *A Title* (Oxford, 1998).");
+    }
+
     #[test]
     fn an_entry_with_no_marker_keeps_all_of_itself() {
         let entry = strip_marker("Smith, John. The Title. Oxford, 1998.");