fix(gloss): Caesar full-text stragglers jamque + pleraque glossed (M-023j)

Full Bellum Gallicum I through the popup lemma path flagged 3 no-gloss
lemmas. The M-023b dismissal was wrong for 2: the analysis lemma for iamque
is 'jamque' (not jam) and for plerumque is 'pleraque' (not plerusque) —
neither was in the artifact, so the popup rendered '—'. Only 'ide' was a
true false-positive (idem resolves via the glossed idem lemma).

Add core overrides jamque→'and now, and already' + pleraque→'the greater
part, most, very many (neut. pl. of plerusque)' with golden rows
(2067→2069), rebuild the artifact (34,413 lemmas / 450 KB). Lock the
form→lemma→gloss path the data suite can't see in a new popup e2e test
(idem/iamque/plerumque all gloss in the browser).

npm test green (22 unit + 81 IPA + 2069 gloss + 348 census); 8 popup e2e.

Author: hellpanderrr

wiktionary_pron/docs/ISSUES.md

@@ -1180,3 +1180,34 @@ tall maxHeight doesn't persist). Verified across 7 words: top stays within
 
 **Status: FIXED.** 7 popup + 19 macronizer e2e + 22 unit + 81 IPA + 2067 golden
 + 348 census, all green.
+
+## M-023j — Caesar full-text stragglers: jamque + pleraque real popup gaps (2026-08-12) ✅
+
+**Symptom.** Full Bellum Gallicum I through the popup lemma path (form →
+wordlist lemma → artifact gloss) reported 1926/1929; the 3 flags were
+`ide`/`jamque`/`pleraque`. M-023b had dismissed all 3 as "the popup resolves
+them correctly."
+
+**Root cause.** That dismissal was wrong for 2 of 3. The popup shows
+`glossFor(analysis-lemma)`, and the analysis lemma for `iamque` is `jamque`
+(not `jam`), for `plerumque` is `pleraque` (not `plerusque`) — neither was in
+the artifact, so both rendered "—". Only `ide` was a true false-positive: the
+form `idem` also carries the glossed `idem` lemma (the `Ide` flag is the
+Mt.-Ida proper-noun homograph). A wordlist scan found the wider unglossed
+enclitic-lemma class: jamque, pleraque, qualiscumque, quotcumque, quotiensque.
+
+**Fix.** Core overrides `jamque`→"and now, and already" and `pleraque`→"the
+greater part, most, very many (neut. pl. of plerusque)", each with a golden
+row (2067→2069). Rebuilt artifact (34,413 lemmas / 450 KB / L&S 89.8%).
+Browser probe confirmed all 3 words now gloss; the straggler case is locked
+in e2e/popup-check.spec.js as a 8th serial test (form→lemma→gloss path is
+invisible to the data suite).
+
+**Lesson.** The data suite resolves the LEMMA directly, so it can't see a
+form that resolves to a DIFFERENT (unglossed) lemma than the one the golden
+row tests. The browser e2e is the only gate on form→lemma→gloss. Also: a
+"99.8% lemmas glossed" claim for a text can hide real popup "—"s in the
+analysis-lemma stratum.
+
+**Status: FIXED.** 8 popup + 19 macronizer e2e + 22 unit + 81 IPA + 2069
+golden + 348 census, all green.

wiktionary_pron/docs/LESSONS.md

@@ -592,10 +592,16 @@ non-obvious method lessons, in the order they bit:
   L&S opens them with usage-notes (adverb-POS class) or empty senses, and the
   wordlist lemma = the wordform itself, so a core key on the lemma works.
   (plerumque's lemma is plerusque, so a core key on plerumque is dead.)
-- **Full BG 1 result: 1926/1929 lemmas glossed (99.8%).** The last 3
-  (ide→idem, jamque→jam, pleraque→plerusque) are Morpheus-rescued forms whose
-  lemmas ARE glossed in the artifact — the popup resolves them correctly.
-  Artifact 34,379 lemmas / 449 KB / L&S 89.7%. Golden 2034 / census 348.
+- **Full BG 1 result: 1926/1929 lemmas glossed (99.8%).** The last 3 were
+  CORRECTED in M-023j (2026-08-12): `ide` was a true false-positive (`idem`
+  resolves via the glossed `idem` lemma, not its Mt.-Ida proper-noun homograph
+  `Ide`), but `iamque` and `plerumque` were REAL popup gaps — the analysis lemma
+  for `iamque` is `jamque` (not `jam`) and for `plerumque` is `pleraque` (not
+  `plerusque`), and neither was in the artifact. Fixed with core overrides
+  `jamque`→"and now, and already" + `pleraque`→"the greater part..." (each with a
+  golden row). The form→lemma→gloss path is NOT visible to the data suite —
+  verify it in the browser (e2e/popup-check.spec.js Caesar-stragglers test).
+  Artifact 34,413 lemmas / 450 KB / L&S 89.8%. Golden 2069 / census 348.
 
 ## Cross-author stress tests (2026-08-11, M-023d)
 
@@ -613,7 +619,8 @@ Catilinam I, Vergil Aeneid I, Ovid Metamorphoses I.
   bulk of each text's missing lemmas are (a) mythological patronymics/geographical
   epithets L&S keys under a DIFFERENT lemma (Helena not helene, Idalium not
   idalia, Teucer not teucrus, Mycenae not mycene, Pyrrha not pyrrhe), (b)
-  Morpheus-rescued forms whose lemma IS glossed (ide→idem, jamque→jam), and
+  Morpheus-rescued forms whose lemma IS glossed (ide→idem — but note jamque
+  and pleraque were NOT: those were real gaps fixed in M-023j), and
   (c) rare unassimilated compound forms. A core override fixes the common ones;
   the patronymic tail is not recoverable without a dedicated proper-noun lexicon.
 - **Unassimilated TABLE grows per-author.** Cicero: adsedeo→assideo, confor→

wiktionary_pron/e2e/popup-check.spec.js

@@ -150,6 +150,41 @@ test("v/u words cycle reversibly — divisa's original spelling must come back",
   await expect(span).toHaveAttribute("content", initial);
 });
 
+test("Caesar BG I stragglers gloss in the popup — idem, iamque, plerumque", async () => {
+  // M-023j: the full-text stress test flagged these as no-gloss. The wordlist
+  // lemmas are idem/`jamque`/pleraque — glosses now land via core overrides
+  // (iamque → jamque → "and now, and already"; plerumque → pleraque → "the
+  // greater part"; idem was already glossed but its proper-noun homograph Ide
+  // isn't). Regression-locks the form→lemma→gloss path the data suite can't see.
+  const page = sharedPage;
+  test.setTimeout(300_000);
+  await page.goto(PAGE);
+  await expect(page.locator("#macronize_btn")).toBeEnabled({ timeout: 240_000 });
+
+  await page.fill("#text_to_macronize", "idem iamque plerumque");
+  await page.click("#macronize_btn");
+  const spans = page.locator("#resultText .ipa");
+  await expect(spans).toHaveCount(3, { timeout: 120_000 });
+
+  const expectations = [
+    { word: "īdem", gloss: "the same" },
+    { word: "iamque", gloss: "and now" },
+    { word: "plērumque", gloss: "greater part" },
+  ];
+  for (let i = 0; i < expectations.length; i++) {
+    const span = spans.nth(i);
+    await expect(span).toHaveText(expectations[i].word, { timeout: 5000 });
+    await span.hover();
+    const defCells = page.locator(".word-popup table.readings td.r-def");
+    await expect(defCells.first()).toBeVisible({ timeout: 10_000 });
+    const defs = (await defCells.allTextContents()).join(" | ");
+    expect(defs).toContain(expectations[i].gloss);
+    // dismiss the popup between words
+    await page.mouse.move(5, 5);
+    await page.waitForTimeout(150);
+  }
+});
+
 test("prose shows no grey placeholders; numbered verse scans after line-number strip", async () => {
   const page = sharedPage;
   test.setTimeout(300_000);

wiktionary_pron/macronizer/glosses.tsv.gz

@@ -950,6 +950,7 @@
   "iustus": "just, righteous, lawful, upright",
   "iuxta": "near, close to, beside; according to",
   "jam": "now, already; soon",
+  "jamque": "and now, and already",
   "janus": "the Roman god of beginnings and gates, with two faces",
   "judex": "a judge",
   "judicatus": "judged, decided",
@@ -1419,6 +1420,7 @@
   "planta2": "the sole of the foot; a plant, sprout",
   "planus": "flat, level, even; clear, plain",
   "plebs": "the common people, the plebeians",
+  "pleraque": "the greater part, most, very many (neut. pl. of plerusque)",
   "plerumque": "generally, commonly, for the most part",
   "plerusque": "very many, the most, most",
   "plinius": "a Roman gens name; esp. Pliny the Elder, who died in the eruption of Vesuvius",

wiktionary_pron/utils/core_gloss.json

@@ -1381,6 +1381,13 @@
     },
     "note": "core_gloss"
   },
+  {
+    "lemma": "jamque",
+    "expect": {
+      "contains": "and now"
+    },
+    "note": "Caesar full-text iamque (M-023j)"
+  },
   {
     "lemma": "jus",
     "expect": {
@@ -14204,6 +14211,13 @@
     },
     "note": "Caesar full-text adverb/particle (M-023b)"
   },
+  {
+    "lemma": "pleraque",
+    "expect": {
+      "contains": "greater part"
+    },
+    "note": "Caesar full-text wordlist twin of plerusque (M-023j)"
+  },
   {
     "lemma": "plerumque",
     "expect": {