@heroiclands/package-build 23.12.6 → 23.12.7

package/engine/content-charset.mjs

@@ -162,6 +162,19 @@ export const NOTATION = Object.freeze(
     ]),
 );
 
+/**
+ * Tier 4 — the blocks the HeroicLands script fonts are registered in, so a
+ * mapping note or a `font=` run never draws a warning for a script character.
+ *
+ * @type {ReadonlyArray<readonly [number, number]>}
+ */
+const SCRIPT_BLOCKS = Object.freeze([
+    [0xf0000, 0xf2fff], // plane 15: Pirzath, Thuravark, Semaron, the reserved alphabets, Teng Chou
+    [0x1e960, 0x1ec6f], // Zethu overflow, Qalezu, Zethu
+    [0x1ecc0, 0x1ecff], // Qalezu catch-before forms
+    [0x13430, 0x1345f], // the Egyptian format controls Zethu uses
+]);
+
 /**
  * Whether the charset admits a code point anywhere in a note.
  *
@@ -169,7 +182,12 @@ export const NOTATION = Object.freeze(
  * @returns {boolean} Whether it is allowed outside a code fence.
  */
 export function isAllowedCodePoint(cp) {
-    return isLetterTier(cp) || TYPOGRAPHY.has(cp) || NOTATION.has(cp);
+    return (
+        isLetterTier(cp) ||
+        TYPOGRAPHY.has(cp) ||
+        NOTATION.has(cp) ||
+        SCRIPT_BLOCKS.some(([from, to]) => cp >= from && cp <= to)
+    );
 }
 
 /**
@@ -362,7 +380,8 @@ export function checkText(text, file) {
                 line: i + 1,
                 column,
                 severity: /** @type {const} */ ("warning"),
-                message: `\`${ch}\` ${hex(cp)} is ${refusalFor(cp)}`,
+                // A Private Use character draws as nothing, so quoting it leaves an empty span.
+                message: `${/\p{Co}/u.test(ch) ? "" : `\`${ch}\` `}${hex(cp)} is ${refusalFor(cp)}`,
             });
         }
     }

package/engine/content-format.yaml

@@ -363,6 +363,7 @@ types:
     doc:
         data:
             - culture
+            - map
         subTypes:
             - rules
             - userguide

package/engine/content-index.mjs

@@ -106,6 +106,7 @@ import { AddressLink } from "./address-values.mjs";
 import { scanSpans } from "./content-spans.mjs";
 import { rebaseNameSpans } from "./name-markup.mjs";
 import { analyzeNameSpans } from "./names.mjs";
+import { recordRuns } from "./run-rows.mjs";
 import fs from "node:fs";
 import path from "node:path";
 
@@ -448,6 +449,8 @@ function assertNoDerivedKeys(frontmatter, relPath, absPath, contentPackage, raw)
  *   `name` reads `Map 1` rather than `Figure 1` where it is due — see
  *   {@link module:engine/content-figures.scanFigures}.
  * @param {import("./content-session.mjs").ContentSyntax} [options.syntax] - Session scanner projections.
+ * @param {import("./content-scripts.mjs").RunContext} [options.runContext] -
+ *   The configured languages, which give each run's `lang=` its tag.
  * @returns {Record<string, any>} The record, keys sorted at every depth. A
  *   **stub** — a note with an empty body, on a type an empty body suppresses —
  *   carries `address` and `anchors` as `null`: it publishes no page, so it
@@ -474,6 +477,7 @@ export function buildIndexRecord({
     dateContext,
     resolveRole,
     syntax,
+    runContext,
 }) {
     assertNoDerivedKeys(frontmatter, relPath, absPath, contentPackage, raw);
 
@@ -541,6 +545,7 @@ export function buildIndexRecord({
                 ...(Object.keys(resolvedDates).length ? { resolvedDates } : {}),
                 package: contentPackage,
                 names: marked.names,
+                runs: recordRuns(source, { context: runContext, bodyLine, bodyColumn, syntax }),
                 address: stub ? null : address,
                 nameAscii: asciiName(frontmatter?.name?.full),
                 aliasesAscii: asciiAliases(frontmatter?.name?.aliases),
@@ -695,6 +700,7 @@ export function indexRecordsForNote({
     dateContext,
     resolveRole,
     syntax,
+    runContext,
 }) {
     resolveNoteId(frontmatter, { pkg: contentPackage });
     const record = buildIndexRecord({
@@ -712,6 +718,7 @@ export function indexRecordsForNote({
         dateContext,
         resolveRole,
         syntax,
+        runContext,
     });
     const records = [decodeIndexAddresses(record, addressContext ?? { package: contentPackage })];
     const address = noteAddress(frontmatter, contentPackage);
@@ -1067,6 +1074,8 @@ export function resolveCultures(
  *   calendar note's shortcode; see {@link module:engine/note-calendars}.
  * @param {string} [options.configFile] - The configuration file, where an
  *   unknown `defaultCalendar` is reported.
+ * @param {Readonly<Record<string, {tag: string}>>} [options.languages] - The
+ *   configured languages, which give each run's `lang=` its tag.
  * @param {(key: string) => ({culture?: unknown}|undefined)} [options.foreignRecord] -
  *   Another package's entry for a canonical address, read for a parent that
  *   package publishes; see {@link resolveCultures}.
@@ -1086,6 +1095,7 @@ export function collectContentIndex(
         currentDate,
         defaultCalendar,
         configFile,
+        languages,
         skipDirectories,
         assetsBase,
         manifest,
@@ -1192,6 +1202,7 @@ export function collectContentIndex(
                 dateContext: dates,
                 resolveRole,
                 syntax,
+                runContext: { languages },
             });
             records.push(...made);
             bodies.push({ record: made[0], body });
@@ -1478,6 +1489,7 @@ export function indexRecordsFor({
         currentDate: resolved.currentDate,
         defaultCalendar: resolved.defaultCalendar ?? undefined,
         configFile: configFileIn(resolved.rootDir, CONFIG_FILENAMES),
+        languages: resolved.languages,
         addressContext: {
             ...noteAddressContext(resolved),
             declaresAddress: dependencyAddresses(resolved),

package/engine/content-poetry.mjs

@@ -77,6 +77,24 @@ import { parseExtensionAttributes, refusedAttributes } from "./extension-attribu
 
 const structureParser = new MarkdownIt({ html: true });
 
+/**
+ * The keys a poetry fence's attribute block takes. `lang`, `font` and `map`
+ * are read by `scanRuns` in content-scripts.mjs; `dir`, and on the site
+ * `data-pagefind-ignore`, are written by its rewrite of a fence set in one of
+ * the faces, and reach the element as they are written.
+ */
+export const POETRY_KEYS = Object.freeze([
+    "form",
+    "meter",
+    "rhyme",
+    "syllables",
+    "lang",
+    "font",
+    "map",
+    "dir",
+    "data-pagefind-ignore",
+]);
+
 /** Read poetry code fences; start/end are inclusive zero-based source lines.
  * @param {string} source - Authored Markdown.
  */
@@ -157,15 +175,13 @@ export function scanPoetry(source) {
         for (const message of [...parsed.problems, ...refusedAttributes(parsed.values)])
             errors.push({ ...at, message });
         for (const [key, value] of Object.entries(parsed.values)) {
-            if (!["form", "meter", "rhyme", "syllables", "lang"].includes(key))
+            if (!POETRY_KEYS.includes(key))
                 errors.push({ ...at, message: `${key}= is not a poetry attribute` });
             else if (key === "syllables" && !/^[1-9]\d*(?:,[1-9]\d*)*$/.test(value))
                 errors.push({
                     ...at,
                     message: "poetry syllables= needs positive comma-separated counts",
                 });
-            else if (key === "lang" && !/^[A-Za-z]{2,8}(?:-[A-Za-z0-9]{1,8})*$/.test(value))
-                errors.push({ ...at, message: "poetry lang= needs a language tag" });
         }
     }
     if (opening?.name === "poetry")
@@ -197,7 +213,9 @@ export function renderPoetry(source) {
         if (block.id) attrs.push(`id="${escape(slugify(block.id))}"`);
         for (const [key, value] of Object.entries(block.attributes))
             if (!refusedAttributes({ [key]: value }).length)
-                attrs.push(`${key === "lang" ? key : `data-${key}`}="${escape(value)}"`);
+                attrs.push(
+                    `${["lang", "dir", "data-pagefind-ignore"].includes(key) ? key : `data-${key}`}="${escape(value)}"`,
+                );
         let indent = block.indent;
         if (
             indent.length >= 4 &&