@stll/folio-core 0.54.2 → 0.56.0

package/dist/docx/canonicalStoryRepack.js

@@ -0,0 +1,74 @@
+import { withTrackedSectionEndpointRemoval } from "../internal/sectionEndpointResolution.js";
+import { withSectionReferenceResolution } from "../internal/sectionReferenceResolution.js";
+import { readDocumentSectionFacts } from "./documentSectionFacts.js";
+import { parseDocx } from "./parser.js";
+import { assertSectionCarriersMatchModel } from "./rezip.js";
+import { serializeDocument } from "./serializer/documentSerializer.js";
+import JSZip from "jszip";
+//#region src/docx/canonicalStoryRepack.ts
+/** Canonical snapshots own explicit section references; the PM projection does not. */
+const referenceKey = ({ element, type, rId }) => `${element}:${type}:${rId}`;
+const xmlFingerprint = async (xml) => {
+	const digest = await crypto.subtle.digest("SHA-256", new TextEncoder().encode(xml));
+	let fingerprint = "";
+	for (const byte of new Uint8Array(digest)) fingerprint += byte.toString(16).padStart(2, "0");
+	return fingerprint;
+};
+/** Authorize exact canonical section removals and references missing from the committed snapshot. */
+const repackWithCanonicalStoryRemovals = async ({ document, repack }) => {
+	const originalBuffer = document.originalBuffer;
+	if (!originalBuffer) return repack();
+	const xml = await (await JSZip.loadAsync(originalBuffer)).file("word/document.xml")?.async("text");
+	if (!xml) return repack();
+	const before = readDocumentSectionFacts(xml);
+	const serializedXml = serializeDocument(document);
+	const after = readDocumentSectionFacts(serializedXml);
+	const remaining = /* @__PURE__ */ new Map();
+	for (const reference of after.headerFooterReferences) {
+		const key = referenceKey(reference);
+		remaining.set(key, (remaining.get(key) ?? 0) + 1);
+	}
+	const removedReferences = [];
+	for (const reference of before.headerFooterReferences) {
+		const key = referenceKey(reference);
+		const count = remaining.get(key) ?? 0;
+		if (count > 0) {
+			remaining.set(key, count - 1);
+			continue;
+		}
+		removedReferences.push({
+			part: reference.element === "headerReference" ? "header" : "footer",
+			type: reference.type,
+			relationshipId: reference.rId
+		});
+	}
+	return withSectionReferenceResolution({
+		document,
+		removedReferences,
+		repack: async () => {
+			if (after.sectionCount >= before.sectionCount) return repack();
+			const baseline = await parseDocx(originalBuffer, { preloadFonts: false });
+			const baselineXml = serializeDocument(baseline);
+			const baselineFacts = readDocumentSectionFacts(baselineXml);
+			if (baselineFacts.sectionCount !== before.sectionCount) return repack();
+			assertSectionCarriersMatchModel({
+				doc: baseline,
+				serializedSectionCount: baselineFacts.sectionCount
+			});
+			return withTrackedSectionEndpointRemoval({
+				document,
+				resolution: {
+					type: "tracked-section-endpoint-removal",
+					sourceParagraphEndpointCount: before.sectionCount,
+					expectedParagraphEndpointCount: after.sectionCount,
+					sourceEndpointFingerprint: await xmlFingerprint(baselineXml),
+					expectedEndpointFingerprint: await xmlFingerprint(serializedXml),
+					removedReferences
+				},
+				repack
+			});
+		}
+	});
+};
+//#endregion
+export { repackWithCanonicalStoryRemovals };

package/dist/docx/contentTypeRegistry.d.ts

@@ -0,0 +1,8 @@
+//#region src/docx/contentTypeRegistry.d.ts
+type ContentTypePart = {
+  partName: string;
+  contentType: string;
+};
+declare const registerContentTypeParts: (xml: string, parts: readonly ContentTypePart[]) => string;
+//#endregion
+export { registerContentTypeParts };

package/dist/docx/contentTypeRegistry.js

@@ -0,0 +1,64 @@
+import { resolvePackageRelationshipTarget } from "./packageParts.js";
+import { spliceXml } from "./selectiveXmlPatch.js";
+import { getAttribute, getLocalName, getNamespacePrefix, getNamespaceUri, parseXmlDocument } from "./xmlParser.js";
+import { panic } from "better-result";
+import { escapeXmlAttribute } from "@stll/docx-core";
+import { OOXML_NS } from "@stll/docx-utils";
+//#region src/docx/contentTypeRegistry.ts
+/** Register a package part without rewriting authored content-type entries. */
+const registerContentTypeParts = (xml, parts) => {
+	const root = parseXmlDocument(xml);
+	if (root?.name === void 0 || getLocalName(root.name) !== "Types" || getNamespaceUri(root) !== OOXML_NS.ct) return panic("The package content-type registry must have a valid Types root.");
+	const overrides = /* @__PURE__ */ new Map();
+	const defaults = /* @__PURE__ */ new Map();
+	for (const element of root.elements ?? []) {
+		if (getNamespaceUri(element) !== OOXML_NS.ct) continue;
+		const type = getAttribute(element, null, "ContentType");
+		if (type === null) continue;
+		if (getLocalName(element.name) === "Override") {
+			const name = getAttribute(element, null, "PartName");
+			const path = name === null ? void 0 : resolvePackageRelationshipTarget(name, "_rels/.rels");
+			if (path !== void 0) overrides.set(path, type);
+		}
+		if (getLocalName(element.name) === "Default") {
+			const extension = getAttribute(element, null, "Extension");
+			if (extension !== null) defaults.set(extension.toLowerCase(), type);
+		}
+	}
+	const prefix = getNamespacePrefix(root.name);
+	const name = prefix ? `${prefix}:Override` : "Override";
+	const additions = [];
+	for (const part of parts) {
+		const path = resolvePackageRelationshipTarget(part.partName, "_rels/.rels");
+		if (path === void 0) return panic("A registered part must have a safe package URI.");
+		const type = overrides.get(path);
+		if (type !== void 0) {
+			if (type !== part.contentType) return panic("A part's registered content type cannot change.");
+			continue;
+		}
+		const extension = path.slice(path.lastIndexOf(".") + 1);
+		if (defaults.get(extension) === part.contentType) continue;
+		additions.push(`<${name} PartName="${escapeXmlAttribute(part.partName)}" ContentType="${escapeXmlAttribute(part.contentType)}"/>`);
+		overrides.set(path, part.contentType);
+	}
+	if (additions.length === 0) return xml;
+	for (const token of xml.matchAll(/<!--[\s\S]*?-->|<\?[\s\S]*?\?>|<!\[CDATA\[[\s\S]*?\]\]>|<(?:"[^"]*"|'[^']*'|[^'">])*>/gu)) {
+		const markup = token[0];
+		if (markup.startsWith("</") && markup.slice(2, -1).trim() === root.name) return spliceXml(xml, [{
+			start: token.index,
+			end: token.index,
+			newXml: additions.join("")
+		}]) ?? panic("A registry insertion must preserve XML ranges.");
+		if (markup.startsWith(`<${root.name}`) && /[\s/>]/u.test(markup.charAt(root.name.length + 1)) && markup.endsWith("/>")) {
+			const end = token.index + markup.length - 2;
+			return spliceXml(xml, [{
+				start: end,
+				end: end + 2,
+				newXml: ">" + additions.join("") + `</${root.name}>`
+			}]) ?? panic("A registry insertion must preserve XML ranges.");
+		}
+	}
+	return panic("A content-type registry needs a closing root boundary.");
+};
+//#endregion
+export { registerContentTypeParts };

package/dist/docx/documentParser.d.ts

@@ -54,22 +54,10 @@ declare function getAllParagraphs(body: document_d_exports.DocumentBody): docume
  * Get all tables from document body
  */
 declare function getAllTables(body: document_d_exports.DocumentBody): document_d_exports.Table[];
-/**
- * Get plain text from entire document body
- */
-declare function getDocumentText(body: document_d_exports.DocumentBody): string;
 /**
  * Count total paragraphs in document
  */
 declare function getParagraphCount(body: document_d_exports.DocumentBody): number;
-/**
- * Count total words in document (approximate)
- */
-declare function getWordCount(body: document_d_exports.DocumentBody): number;
-/**
- * Count total characters in document
- */
-declare function getCharacterCount(body: document_d_exports.DocumentBody): number;
 /**
  * Get section count
  */
@@ -79,4 +67,4 @@ declare function getSectionCount(body: document_d_exports.DocumentBody): number;
  */
 declare function hasTemplateVariables(body: document_d_exports.DocumentBody): boolean;
 //#endregion
-export { extractAllTemplateVariables, extractTemplateVariables, getAllParagraphs, getAllTables, getCharacterCount, getDocumentText, getParagraphCount, getSectionCount, getWordCount, hasTemplateVariables, parseDocumentBody, parseDocumentBodyTree };
+export { extractAllTemplateVariables, extractTemplateVariables, getAllParagraphs, getAllTables, getParagraphCount, getSectionCount, hasTemplateVariables, parseDocumentBody, parseDocumentBodyTree };

package/dist/docx/documentParser.js

@@ -239,58 +239,12 @@ function getNestedTables(table) {
 	return tables;
 }
 /**
-* Get plain text from entire document body
-*/
-function getDocumentText(body) {
-	const lines = [];
-	for (const block of body.content) if (block.type === "paragraph") lines.push(getParagraphText(block));
-	else if (block.type === "table") lines.push(getTableText(block));
-	else if (block.type === "blockSdt" || block.type === "blockCustomXml") lines.push(getTextFromBlocks(block.content));
-	return lines.join("\n");
-}
-const getTextFromBlocks = (blocks) => blocks.flatMap((block) => {
-	if (block.type === "paragraph") return [getParagraphText(block)];
-	if (block.type === "table") return [getTableText(block)];
-	return block.type === "blockSdt" || block.type === "blockCustomXml" ? [getTextFromBlocks(block.content)] : [];
-}).join("\n");
-/**
-* Get plain text from a table
-*/
-function getTableText(table) {
-	const lines = [];
-	for (const row of table.rows) {
-		const rowTexts = [];
-		for (const cell of row.cells) {
-			const cellTexts = [];
-			for (const content of cell.content) if (content.type === "paragraph") cellTexts.push(getParagraphText(content));
-			else if (content.type === "table") cellTexts.push(getTableText(content));
-			else if (content.type === "blockSdt" || content.type === "blockCustomXml") cellTexts.push(getTextFromBlocks(content.content));
-			rowTexts.push(cellTexts.join("\n"));
-		}
-		lines.push(rowTexts.join("	"));
-	}
-	return lines.join("\n");
-}
-/**
 * Count total paragraphs in document
 */
 function getParagraphCount(body) {
 	return getAllParagraphs(body).length;
 }
 /**
-* Count total words in document (approximate)
-*/
-function getWordCount(body) {
-	const words = getDocumentText(body).trim().split(/\s+/u);
-	return words.length > 0 && words[0] !== "" ? words.length : 0;
-}
-/**
-* Count total characters in document
-*/
-function getCharacterCount(body) {
-	return getDocumentText(body).length;
-}
-/**
 * Get section count
 */
 function getSectionCount(body) {
@@ -303,4 +257,4 @@ function hasTemplateVariables(body) {
 	return extractAllTemplateVariables(body.content).length > 0;
 }
 //#endregion
-export { extractAllTemplateVariables, extractTemplateVariables, getAllParagraphs, getAllTables, getCharacterCount, getDocumentText, getParagraphCount, getSectionCount, getWordCount, hasTemplateVariables, parseDocumentBody, parseDocumentBodyTree };
+export { extractAllTemplateVariables, extractTemplateVariables, getAllParagraphs, getAllTables, getParagraphCount, getSectionCount, hasTemplateVariables, parseDocumentBody, parseDocumentBodyTree };

package/dist/docx/ensureParaIds.d.ts

@@ -1,8 +1,14 @@
 //#region src/docx/ensureParaIds.d.ts
+declare const ENSURE_PARA_IDS_REASONS: {
+  readonly NORMALIZATION_FAILED: "normalization-failed";
+  readonly NAMESPACE_INVALID: "namespace-invalid";
+  readonly SIGNED_PACKAGE: "signed-package";
+};
 declare const EnsureParaIdsError_base: import("better-result").TaggedErrorClass<"EnsureParaIdsError">;
 /** A malformed or unsupported package prevented paragraph-ID normalization. */
 declare class EnsureParaIdsError extends EnsureParaIdsError_base<{
   message: string;
+  reason: (typeof ENSURE_PARA_IDS_REASONS)[keyof typeof ENSURE_PARA_IDS_REASONS];
   cause?: unknown;
 }> {}
 /** Counts and normalized bytes returned by {@link ensureParaIds}. */
@@ -35,4 +41,4 @@ type EnsureParaIdsOptions = {
  */
 declare const ensureParaIds: (docx: Uint8Array | ArrayBuffer, options?: EnsureParaIdsOptions) => Promise<EnsureParaIdsResult>;
 //#endregion
-export { EnsureParaIdsError, EnsureParaIdsOptions, EnsureParaIdsResult, ensureParaIds };
+export { ENSURE_PARA_IDS_REASONS, EnsureParaIdsError, EnsureParaIdsOptions, EnsureParaIdsResult, ensureParaIds };