colophon: six more things the front matter of a book is not

Read out of the review queue a real run produced, where each of these was sitting as a proposed title waiting for a person to say no: 1. General Remarks a numbered heading 2 James W. McAllister a numbered heading 5222 What Is This a folio that ran into the line 827 North Fourth Street an address Page xxxvi / Chapter I a marker title a form label, printed literally Sidebottom, Harry. Ancient a bibliography entry To Mackenzie, a dedication The numbers are the delicate part. A leading number that could be a year is left alone, because 1984 Revisited is a book and 5222 What Is This is a folio, and nothing else distinguishes them. A dedication has to end in a comma, so To Kill a Mockingbird survives. The copyright notices in the same pile are already fixed: those proposals were written before 0.2.2 and are stale rather than wrong.
0.2.2 — a subtitle begins with a capital
2026-10-22 18:57:14 +00:00 · 2026-09-21 15:47:24 -04:00 · 2026-09-21 11:30:26 -04:00 · 2026-09-21 11:19:54 -04:00
5 changed files with 205 additions and 56 deletions
@@ -101,7 +101,7 @@ checksum = "f2032f911046de80f0a198e0901378627c33f59ea0ac00e363d481118bd70a53"

 [[package]]
 name = "avplan"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -126,7 +126,7 @@ checksum = "ac07cdecf99051d9a5238b80f35af32cdeba5b336e55d957b318b50137e18da5"

 [[package]]
 name = "bibclean"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -143,7 +143,7 @@ dependencies = [

 [[package]]
 name = "bibfetch"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -196,7 +196,7 @@ dependencies = [

 [[package]]
 name = "bookmeta"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -210,7 +210,7 @@ dependencies = [

 [[package]]
 name = "booksniff"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -250,7 +250,7 @@ checksum = "fc652a48c352aef3ea3aed32080501cf3ef6ed5da78602a020c991775b0aff04"

 [[package]]
 name = "catena"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "bibclean",
 "bibfetch",
@@ -277,7 +277,7 @@ dependencies = [

 [[package]]
 name = "catena-cli"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "clap",
 "clap_mangen",
@@ -312,7 +312,7 @@ checksum = "4e7648175b45a9a48536d676f68d918270699102aa8dab5496df06904c914600"

 [[package]]
 name = "chunkmd"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -334,7 +334,7 @@ dependencies = [

 [[package]]
 name = "citefind"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "chunkmd",
@@ -399,7 +399,7 @@ dependencies = [

 [[package]]
 name = "colophon"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "bibclean",
 "catena-cli",
@@ -448,7 +448,7 @@ dependencies = [

 [[package]]
 name = "corpus-core"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "schemars",
 "serde",
@@ -492,7 +492,7 @@ dependencies = [

 [[package]]
 name = "cuekit"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "corpus-core",
 "regex",
@@ -504,7 +504,7 @@ dependencies = [

 [[package]]
 name = "cuenorm"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -514,7 +514,7 @@ dependencies = [

 [[package]]
 name = "cuepack"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -525,7 +525,7 @@ dependencies = [

 [[package]]
 name = "cueseg"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -612,7 +612,7 @@ dependencies = [

 [[package]]
 name = "emb"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -876,7 +876,7 @@ dependencies = [

 [[package]]
 name = "idnorm"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "bibclean",
 "catena-cli",
@@ -942,7 +942,7 @@ dependencies = [

 [[package]]
 name = "langid"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1022,7 +1022,7 @@ dependencies = [

 [[package]]
 name = "mdscrub"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1034,7 +1034,7 @@ dependencies = [

 [[package]]
 name = "mdsplit"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1113,7 +1113,7 @@ checksum = "384b8ab6d37215f3c5301a95a4accb5d64aa607f1fcb26a11b5303878451b4fe"

 [[package]]
 name = "pdfgutter"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1126,7 +1126,7 @@ dependencies = [

 [[package]]
 name = "pdftriage"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1370,7 +1370,7 @@ checksum = "323c417e1d9665a65b263ec744ba09030cfb277e9daa0b018a4ab62e57bc8189"

 [[package]]
 name = "rrf"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1517,7 +1517,7 @@ dependencies = [

 [[package]]
 name = "seriesof"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -1634,7 +1634,7 @@ checksum = "adb6935a6f5c20170eeceb1a3835a49e12e19d792f6dd344ccc76a985ca5a6ca"

 [[package]]
 name = "textkit"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "regex",
 "tiktoken-rs",
@@ -1724,7 +1724,7 @@ checksum = "fd3ca314f692efd6c868f8408f53fe444634a845f96c028b97d35f6a1f79f0ee"

 [[package]]
 name = "tocfind"
-version = "0.2.0"
+version = "0.2.3"
 dependencies = [
 "catena-cli",
 "clap",
@@ -3,39 +3,39 @@ resolver = "3"
 members = ["crates/*"]

 [workspace.package]
-version = "0.2.0"
+version = "0.2.3"
 edition = "2024"
 rust-version = "1.85"
 license = "MPL-2.0"
 repository = "https://git.pumpin.online/funkpower/catena"

 [workspace.dependencies]
-catena-cli = { path = "crates/catena-cli", version = "0.2.0" }
-corpus-core = { path = "crates/corpus-core", version = "0.2.0" }
-cuekit = { path = "crates/cuekit", version = "0.2.0" }
-textkit = { path = "crates/textkit", version = "0.2.0" }
-chunkmd = { path = "crates/chunkmd", version = "0.2.0" }
-mdscrub = { path = "crates/mdscrub", version = "0.2.0" }
-mdsplit = { path = "crates/mdsplit", version = "0.2.0" }
-pdftriage = { path = "crates/pdftriage", version = "0.2.0" }
-cuenorm = { path = "crates/cuenorm", version = "0.2.0" }
-cueseg = { path = "crates/cueseg", version = "0.2.0" }
-rrf = { path = "crates/rrf", version = "0.2.0" }
-cuepack = { path = "crates/cuepack", version = "0.2.0" }
-bibclean = { path = "crates/bibclean", version = "0.2.0" }
-avplan = { path = "crates/avplan", version = "0.2.0" }
-pdfgutter = { path = "crates/pdfgutter", version = "0.2.0" }
-catena = { path = "crates/bindings", version = "0.2.0" }
-seriesof = { path = "crates/seriesof", version = "0.2.0" }
-idnorm = { path = "crates/idnorm", version = "0.2.0" }
-bibfetch = { path = "crates/bibfetch", version = "0.2.0" }
-bookmeta = { path = "crates/bookmeta", version = "0.2.0" }
-langid = { path = "crates/langid", version = "0.2.0" }
-tocfind = { path = "crates/tocfind", version = "0.2.0" }
-booksniff = { path = "crates/booksniff", version = "0.2.0" }
-citefind = { path = "crates/citefind", version = "0.2.0" }
-colophon = { path = "crates/colophon", version = "0.2.0" }
-emb = { path = "crates/emb", version = "0.2.0" }
+catena-cli = { path = "crates/catena-cli", version = "0.2.3" }
+corpus-core = { path = "crates/corpus-core", version = "0.2.3" }
+cuekit = { path = "crates/cuekit", version = "0.2.3" }
+textkit = { path = "crates/textkit", version = "0.2.3" }
+chunkmd = { path = "crates/chunkmd", version = "0.2.3" }
+mdscrub = { path = "crates/mdscrub", version = "0.2.3" }
+mdsplit = { path = "crates/mdsplit", version = "0.2.3" }
+pdftriage = { path = "crates/pdftriage", version = "0.2.3" }
+cuenorm = { path = "crates/cuenorm", version = "0.2.3" }
+cueseg = { path = "crates/cueseg", version = "0.2.3" }
+rrf = { path = "crates/rrf", version = "0.2.3" }
+cuepack = { path = "crates/cuepack", version = "0.2.3" }
+bibclean = { path = "crates/bibclean", version = "0.2.3" }
+avplan = { path = "crates/avplan", version = "0.2.3" }
+pdfgutter = { path = "crates/pdfgutter", version = "0.2.3" }
+catena = { path = "crates/bindings", version = "0.2.3" }
+seriesof = { path = "crates/seriesof", version = "0.2.3" }
+idnorm = { path = "crates/idnorm", version = "0.2.3" }
+bibfetch = { path = "crates/bibfetch", version = "0.2.3" }
+bookmeta = { path = "crates/bookmeta", version = "0.2.3" }
+langid = { path = "crates/langid", version = "0.2.3" }
+tocfind = { path = "crates/tocfind", version = "0.2.3" }
+booksniff = { path = "crates/booksniff", version = "0.2.3" }
+citefind = { path = "crates/citefind", version = "0.2.3" }
+colophon = { path = "crates/colophon", version = "0.2.3" }
+emb = { path = "crates/emb", version = "0.2.3" }

 clap = { version = "4", features = ["derive"] }
 clap_mangen = "0.3"
@@ -67,7 +67,19 @@ pub fn title_case(title: &str) -> String {
                .to_lowercase()
                .trim_end_matches(['.', ',', ':', ';', '!', '?', ')', ']', '"', '\''])
                .to_string();
-            if SMALL_WORDS.binary_search(&core.as_str()).is_ok() && at > 0 && at < last {
+            // A small word stays small in the middle of a title, but not
+            // where a title actually starts. A colon starts a subtitle, and
+            // "Aesthetic in Kant: a Critique" is the giveaway that the rule
+            // was only ever counting from the beginning of the whole string.
+            let opens_a_subtitle = at > 0
+                && words[at - 1]
+                    .trim_end_matches(['"', '\'', ')', ']'])
+                    .ends_with([':', ';', '?', '!', '.']);
+            if SMALL_WORDS.binary_search(&core.as_str()).is_ok()
+                && at > 0
+                && at < last
+                && !opens_a_subtitle
+            {
                return word.to_lowercase();
            }
            let lowered = word.to_lowercase();
@@ -218,3 +230,35 @@ mod tests {
        );
    }
 }
+
+#[cfg(test)]
+mod subtitle_case_tests {
+    use super::title_case;
+
+    #[test]
+    fn a_subtitle_starts_with_a_capital() {
+        // The rule counted small words from the start of the string, so every
+        // subtitle beginning "A", "The" or "Of" was set lower case.
+        assert_eq!(
+            title_case("AESTHETIC IN KANT: A CRITIQUE"),
+            "Aesthetic in Kant: A Critique"
+        );
+        assert_eq!(
+            title_case("BORN FIGHTING: THE SCOTS-IRISH"),
+            "Born Fighting: The Scots-Irish"
+        );
+        assert_eq!(
+            title_case("A HISTORY: AN INTRODUCTION TO IT"),
+            "A History: An Introduction to It"
+        );
+    }
+
+    #[test]
+    fn a_small_word_in_the_middle_is_still_small() {
+        assert_eq!(
+            title_case("THE RISE AND FALL OF THE THIRD REICH"),
+            "The Rise and Fall of the Third Reich"
+        );
+        assert_eq!(title_case("WAR OF THE WORLDS"), "War of the Worlds");
+    }
+}
@@ -260,9 +260,14 @@ mod tests {
    #[test]
    fn an_underscored_title_is_cleaned_before_it_is_cased() {
        let cleaned = clean(&book("THE FALL_ AND THE RISE_txt", "Arendt, Hannah"));
+        // "And" is capitalised because the colon starts a subtitle, and the
+        // first word of one is capitalised however small it is. This asserted
+        // "and" until the rule was found to be counting small words from the
+        // start of the whole string, which set every subtitle opening "A",
+        // "The" or "An" in lower case.
        assert_eq!(
            cleaned.book.title.as_deref(),
-            Some("The Fall: and the Rise")
+            Some("The Fall: And the Rise")
        );
        assert_eq!(cleaned.changes.len(), 2);
    }
@@ -35,8 +35,33 @@ static LIFESPAN: LazyLock<Regex> = LazyLock::new(|| {
 static BARE_DATE: LazyLock<Regex> =
    LazyLock::new(|| Regex::new(r"(?i)^\d{1,4}\s*(bc|ad|bce|ce)?\.?$").expect("a valid pattern"));
 /// `Copyright © 2013 by`, `\u{a9} 2013 Richard Swinburne`, `Copyright 1971 by`.
+/// `1. General Remarks`, `2   James W. McAllister` — a numbered heading.
+static NUMBERED_HEADING: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\s*\d+\s*[.)]\s+\S|^\s*\d+\s{2,}\S").expect("a valid pattern"));
+/// `5222 What Is This`, `827 North Fourth Street` — a number that is not a
+/// year, leading a line.
+static LEADING_NUMBER: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\s*(\d{1,5})\s+\S").expect("a valid pattern"));
+/// `Page xxxvi`, `Chapter I`, `Fig. 3` — a marker, not a name.
+static MARKER: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(
+        r"(?i)^\s*(page|pages|p\.|pp\.|fig\.|figure|plate|chapter|ch\.|section|part)\s+[\divxlcm]",
+    )
+    .expect("a valid pattern")
+});
+/// `Sidebottom, Harry. Ancient warfare` — a bibliography entry, which front
+/// matter prints in its cataloguing block and in its further reading.
+static CATALOGUE_ENTRY: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"^[A-Z][A-Za-z'\-]+,\s+[A-Z][A-Za-z'.\-]*\.\s+\S").expect("a valid pattern")
+});
+/// `To Mackenzie,` — a dedication, set alone on its own page.
+static DEDICATION: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(?i)^\s*(to|for|in memory of)\s+\S.*,\s*$").expect("a valid pattern")
+});
 static COPYRIGHT_NOTICE: LazyLock<Regex> = LazyLock::new(|| {
-    Regex::new(r"(?i)(\u{a9}|\(c\)|copyright)\s*(\u{a9}|\(c\))?\s*(1[0-9]|20)\d{2}")
+    // The year need not follow the symbol directly: "(c) Alfredo Ferrarin
+    // 2004" puts the holder between them, and was read as a title.
+    Regex::new(r"(?i)(\u{a9}|\(c\)|copyright)[^\n]{0,60}?(1[0-9]|20)\d{2}")
        .expect("a valid pattern")
 });
 /// `Edited by`, with whoever it was set on the line beneath.
@@ -77,10 +102,13 @@ copenhagen|stockholm|moscow|beijing|tokyo|delhi|nairobi)\b[\s,·•|and]*",
 static STRUCTURAL: &[&str] = &[
    "abbreviations",
    "about the author",
+    "abstract",
    "acknowledgements",
    "acknowledgments",
    "afterword",
    "appendix",
+    "author",
+    "authors",
    "bibliography",
    "contents",
    "dedication",
@@ -93,6 +121,7 @@ static STRUCTURAL: &[&str] = &[
    "illustrations",
    "index",
    "introduction",
+    "keywords",
    "list of contributors",
    "list of illustrations",
    "maps",
@@ -100,7 +129,9 @@ static STRUCTURAL: &[&str] = &[
    "postscript",
    "preface",
    "prologue",
+    "subtitle",
    "table of contents",
+    "title",
    "title page",
    "translator's note",
 ];
@@ -180,6 +211,22 @@ fn is_furniture(line: &str) -> bool {
    }
    // Sentences a copyright page sets, which are not notices and so not caught
    // by the notice pattern.
+    if NUMBERED_HEADING.is_match(t) || MARKER.is_match(t) || CATALOGUE_ENTRY.is_match(t) {
+        return true;
+    }
+    // A line opening with a number that is not a year. "5222 What Is This" and
+    // "827 North Fourth Street" are a stray folio and an address; "1984
+    // Revisited" is a book, and the difference is whether the number could be
+    // the year of one.
+    if let Some(c) = LEADING_NUMBER.captures(t) {
+        let n: u32 = c[1].parse().unwrap_or(0);
+        if !(1000..=2099).contains(&n) {
+            return true;
+        }
+    }
+    if DEDICATION.is_match(t) {
+        return true;
+    }
    let lower = t.to_lowercase();
    for phrase in [
        "moral rights of the author",
@@ -413,6 +460,8 @@ mod furniture_tests {
            "12 INDEX",
            "Translated from the French by Willard R. Trask",
            "Copyright © 2013 by Barry Davies, BEM",
+            "© Alfredo Ferrarin 2004",
+            "\u{a9} Cambridge University Press 2011",
            "[PART ONE:",
            "Piper Verlag GmbH, Munich<span style=\"font-size\">",
            "Vol. 11. HENRY JAMES, JR.",
@@ -492,3 +541,54 @@ mod conversion_tests {
        }
    }
 }
+
+#[cfg(test)]
+mod review_pile_tests {
+    use super::*;
+
+    /// Everything here was proposed as a title by a real run over this
+    /// library, and sat in the review queue waiting for a person to say no.
+    #[test]
+    fn what_the_review_pile_was_full_of() {
+        for line in [
+            "1. General Remarks",
+            "2   James W. McAllister",
+            "5222 What Is This",
+            "827 North Fourth Street",
+            "Page xxxvi",
+            "Page xxxviii",
+            "Chapter I",
+            "title",
+            "Sidebottom, Harry. Ancient warfare",
+            "To Mackenzie,",
+        ] {
+            assert!(
+                is_structural(line) || is_furniture(line),
+                "still read as a title: {line:?}"
+            );
+        }
+    }
+
+    /// And none of it catches a book. The numbers are the delicate part: a
+    /// year can open a title, and a folio cannot.
+    #[test]
+    fn a_book_is_still_a_book() {
+        for line in [
+            "1984 Revisited",
+            "1914: The Year the World Ended",
+            "Chapter and Verse",
+            "Part of the Furniture",
+            "To Kill a Mockingbird",
+            "For Whom the Bell Tolls",
+            "The Title of This Book",
+            "Pages from a Cold Island",
+            "Sections of a Broken Whole",
+            "In Memory of Bread",
+        ] {
+            assert!(
+                !(is_structural(line) || is_furniture(line)),
+                "rejected a real title: {line:?}"
+            );
+        }
+    }
+}