@stll/folio-core 0.54.1 → 0.55.0

package/dist/docx/structuralXmlPatch.js

@@ -1,7 +1,7 @@
 import { canonicalJson } from "../utils/canonicalJson.js";
 import { parseDocumentBody } from "./documentParser.js";
 import { paraIdAttribute } from "./paraIdAttribute.js";
-import { spliceXml } from "./selectiveXmlPatch.js";
+import { spliceSupportsRootNamespaces, spliceXml } from "./selectiveXmlPatch.js";
 import { readRootNamespaceBindings, serializePartElement } from "./serializer/partNamespaces.js";
 import { NAMESPACES, WORDPROCESSINGML_NAMESPACE_URIS, getChildElements, getLocalName, getNamespaceUri, parseXmlDocument } from "./xmlParser.js";
 //#region src/docx/structuralXmlPatch.ts
@@ -203,6 +203,13 @@ const buildStructuralDocumentPatch = ({ originalXml, serializedXml, changedIds }
 		end: source.bodyEnd,
 		newXml: pending.join("")
 	});
+	if (!spliceSupportsRootNamespaces({
+		originalXml,
+		serializedXml,
+		fragments: splices.map(({ newXml }) => newXml).filter(Boolean),
+		originalRoot: source.root,
+		serializedRoot: current.root
+	})) return null;
 	return spliceXml(originalXml, splices);
 };
 //#endregion

package/dist/docx/unzip.d.ts

@@ -99,6 +99,8 @@ type UnzipDocxBehavior = {
  * @returns Promise resolving to extracted content
  */
 declare function unzipDocx(buffer: ArrayBuffer, options?: DocxUnzipOptions, { verifyUnreadEntries }?: UnzipDocxBehavior): Promise<RawDocxContent>;
+/** Keep extracted projections and the source ZIP aligned after parser normalization. */
+declare const replaceRawDocxXmlParts: (content: RawDocxContent, parts: ReadonlyMap<string, string>) => Promise<void>;
 declare function getEntryUncompressedSize(file: JSZip.JSZipObject): number | null;
 /**
  * Get a list of all files in the DOCX
@@ -160,4 +162,4 @@ declare function getContentSummary(content: RawDocxContent): {
   totalFiles: number;
 };
 //#endregion
-export { DocxSecurityError, DocxUnzipLimits, DocxUnzipOptions, RawDocxContent, UnzipDocxBehavior, extractFile, getContentSummary, getEntryUncompressedSize, getFileList, getMediaMimeType, hasFile, mediaToDataUrl, unzipDocx };
+export { DocxSecurityError, DocxUnzipLimits, DocxUnzipOptions, RawDocxContent, UnzipDocxBehavior, extractFile, getContentSummary, getEntryUncompressedSize, getFileList, getMediaMimeType, hasFile, mediaToDataUrl, replaceRawDocxXmlParts, unzipDocx };

package/dist/docx/unzip.js

@@ -2,6 +2,7 @@ import { bytesToDataUrl } from "../utils/base64.js";
 import { compressionRatioLimitFor, countCentralDirectoryRecords, createInflationBudget, exceedsCompressionRatio, getZipEntrySizes, inflateEntryWithinLimits, isStoredZipEntry } from "./archiveInflation.js";
 import { DOCX_CONTAINER_TYPES, detectDocxContainerType } from "./encryption/containerFormat.js";
 import { openDocxBuffer } from "./encryption/openEncryptedDocx.js";
+import { RASTER_MIME_TYPES, detectRasterMimeType } from "./rasterMime.js";
 import { decodeXmlBytes } from "./xmlEncoding.js";
 import { FOLIO_XML_RESOURCE_LIMITS, assertXmlResourceLimits, createXmlPackageBudget } from "./xmlResourceLimits.js";
 import JSZip from "jszip";
@@ -38,12 +39,7 @@ var DocxSecurityError = class extends Error {
 	}
 };
 const DEFAULT_ALLOWED_MEDIA_MIME_TYPES = /* @__PURE__ */ new Set([
-	"image/png",
-	"image/jpeg",
-	"image/gif",
-	"image/bmp",
-	"image/tiff",
-	"image/webp",
+	...RASTER_MIME_TYPES,
 	"image/x-emf",
 	"image/x-wmf"
 ]);
@@ -227,11 +223,12 @@ async function unzipDocx(buffer, options = {}, { verifyUnreadEntries = true } =
 					return null;
 				}
 				const binaryContent = result.bytes.buffer;
-				if (!isMediaContentAllowed(binaryContent, mimeType)) return null;
+				const effectiveMimeType = RASTER_MIME_TYPES.has(mimeType) ? detectRasterMimeType(binaryContent) : mimeType;
+				if (!effectiveMimeType || !limits.allowedMediaMimeTypes.has(effectiveMimeType) || !isMediaContentAllowed(binaryContent, effectiveMimeType)) return null;
 				return {
 					type: "media",
 					path,
-					mimeType,
+					mimeType: effectiveMimeType,
 					content: binaryContent
 				};
 			});
@@ -274,6 +271,26 @@ async function unzipDocx(buffer, options = {}, { verifyUnreadEntries = true } =
 	}
 	return content;
 }
+/** Keep extracted projections and the source ZIP aligned after parser normalization. */
+const replaceRawDocxXmlParts = async (content, parts) => {
+	let identitiesChanged = false;
+	for (const [path, xmlContent] of parts) {
+		if (content.allXml.get(path) === xmlContent) continue;
+		indexXmlContent(content, {
+			type: "xml",
+			path,
+			lowerPath: path.toLowerCase(),
+			content: xmlContent
+		});
+		content.originalZip.file(path, xmlContent);
+		identitiesChanged = true;
+	}
+	if (identitiesChanged) content.originalBuffer = await content.originalZip.generateAsync({
+		type: "arraybuffer",
+		compression: "DEFLATE",
+		compressionOptions: { level: 1 }
+	});
+};
 /**
 * Preflight every XML part the unzip retains, not a named few.
 *
@@ -291,6 +308,14 @@ function assignXmlContent(content, { path, lowerPath, content: xmlContent }, lim
 		partPath: path,
 		budget
 	});
+	indexXmlContent(content, {
+		type: "xml",
+		path,
+		lowerPath,
+		content: xmlContent
+	});
+}
+const indexXmlContent = (content, { path, lowerPath, content: xmlContent }) => {
 	content.allXml.set(path, xmlContent);
 	if (lowerPath === "word/document.xml") content.documentXml = xmlContent;
 	else if (lowerPath === "word/styles.xml") content.stylesXml = xmlContent;
@@ -317,7 +342,7 @@ function assignXmlContent(content, { path, lowerPath, content: xmlContent }, lim
 		const filename = path.split("/").pop() || path;
 		content.footers.set(filename, xmlContent);
 	}
-}
+};
 function inflationLimitError(limit, path) {
 	switch (limit) {
 		case "declared-size": return new DocxSecurityError(`DOCX entry inflates past its declared size: ${path}`);
@@ -499,14 +524,9 @@ function isEntryTooLarge(declaredSize, maxBytes) {
 	return declaredSize !== null && declaredSize > maxBytes;
 }
 function isMediaContentAllowed(data, mimeType) {
+	if (RASTER_MIME_TYPES.has(mimeType)) return detectRasterMimeType(data) === mimeType;
 	const bytes = new Uint8Array(data);
 	switch (mimeType) {
-		case "image/png": return bytes[0] === 137 && bytes[1] === 80 && bytes[2] === 78 && bytes[3] === 71;
-		case "image/jpeg": return bytes[0] === 255 && bytes[1] === 216;
-		case "image/gif": return bytes[0] === 71 && bytes[1] === 73 && bytes[2] === 70 && bytes[3] === 56;
-		case "image/bmp": return bytes[0] === 66 && bytes[1] === 77;
-		case "image/webp": return bytes[0] === 82 && bytes[1] === 73 && bytes[2] === 70 && bytes[3] === 70 && bytes[8] === 87 && bytes[9] === 69 && bytes[10] === 66 && bytes[11] === 80;
-		case "image/tiff": return bytes[0] === 73 && bytes[1] === 73 || bytes[0] === 77 && bytes[1] === 77;
 		case "image/x-emf":
 		case "image/emf": return bytes.length >= 44 && bytes[0] === 1 && bytes[40] === 32 && bytes[41] === 69 && bytes[42] === 77 && bytes[43] === 70;
 		case "image/x-wmf":
@@ -605,4 +625,4 @@ function getContentSummary(content) {
 	};
 }
 //#endregion
-export { DocxSecurityError, extractFile, getContentSummary, getEntryUncompressedSize, getFileList, getMediaMimeType, hasFile, mediaToDataUrl, unzipDocx };
+export { DocxSecurityError, extractFile, getContentSummary, getEntryUncompressedSize, getFileList, getMediaMimeType, hasFile, mediaToDataUrl, replaceRawDocxXmlParts, unzipDocx };

package/dist/docx/verbatimCapture.d.ts

@@ -1,3 +1,4 @@
+import { document_d_exports } from "../types/document.js";
 import { XmlElement, XmlNamespaceScope } from "./xmlParser.js";
 //#region src/docx/verbatimCapture.d.ts
 declare const UntranslatableStrictNamespaceError_base: import("better-result").TaggedErrorClass<"UntranslatableStrictNamespaceError">;
@@ -29,5 +30,6 @@ declare const isSafeCapturedXmlDocument: (xml: string) => boolean;
  * must therefore depend on nothing but the element they are given.
  */
 declare const createCapturedXmlSanitizer: (options: SanitizeCapturedXmlOptions) => ((xml: string | undefined) => string | null);
+declare const captureSourceProfileXml: (element: XmlElement, conformance: document_d_exports.DocxConformanceClass | undefined) => string;
 //#endregion
-export { UntranslatableStrictNamespaceError, captureVerbatimXml, createCapturedXmlSanitizer, isSafeCapturedXmlDocument };
+export { UntranslatableStrictNamespaceError, captureSourceProfileXml, captureVerbatimXml, createCapturedXmlSanitizer, isSafeCapturedXmlDocument };

package/dist/docx/verbatimCapture.js

@@ -1,10 +1,11 @@
 import { TRANSITIONAL_NAMESPACE_BY_STRICT_URI } from "./strictValueEncodings.gen.js";
 import { NUMBERS_PER_PERCENT, isStrictNamespaceUri, percentageSpelling, toTransitionalNamespaceUri, transitionalSlotEncoding } from "./transitionalSpelling.js";
 import { roundHalfAwayFromZero, universalMeasureAs } from "./universalMeasure.js";
-import { NAMESPACES, OOXML_NAMESPACE_SCOPE, cloneElement, elementToXml, getChildElements, getLocalName, getNamespacePrefix, parseXml } from "./xmlParser.js";
+import { NAMESPACES, OOXML_NAMESPACE_SCOPE, WORDPROCESSINGML_NAMESPACE_URIS, cloneElement, elementToXml, getAttributeByNamespaceUri, getChildElements, getLocalName, getNamespacePrefix, parseXml, resolveAttributeNamespaceUri } from "./xmlParser.js";
 import { assertXmlResourceLimits } from "./xmlResourceLimits.js";
 import { getDocxXmlSafetyIssue } from "./xmlSafety.js";
-import { TaggedError } from "better-result";
+import { TaggedError, panic } from "better-result";
+import { DOCX_CONFORMANCE_CLASSES } from "@stll/docx-core/model";
 //#region src/docx/verbatimCapture.ts
 /**
 * The one way to turn parsed markup into a string a rebuilt part replays.
@@ -75,14 +76,22 @@ const originOf = (element, inherited) => {
 	});
 	return TRANSITIONAL_URIS.has(uri) ? "transitional" : inherited;
 };
-const convertAttributes = (element, namespaceUri, localName, convertValues) => {
+/** A same-local-name extension attribute is not a slot in the element's vocabulary. */
+const attributeSlotEncoding = (element, name) => {
+	if (getNamespacePrefix(name) !== null) {
+		const attributeUri = resolveAttributeNamespaceUri(element, name);
+		if (attributeUri === void 0 || element.namespaceUri === void 0 || toTransitionalNamespaceUri(attributeUri) !== toTransitionalNamespaceUri(element.namespaceUri)) return void 0;
+	}
+	return transitionalSlotEncoding(element.namespaceUri, getLocalName(element.name), getLocalName(name));
+};
+const convertAttributes = (element, namespaceUri, convertValues) => {
 	const attributes = element.attributes ?? {};
 	let converted;
 	for (const [name, raw] of Object.entries(attributes)) {
 		if (typeof raw !== "string") continue;
 		let next = raw;
 		if (name === "xmlns" || name.startsWith("xmlns:")) next = toTransitionalNamespaceUri(raw);
-		else if (convertValues) next = transitionalValue(raw, namespaceUri, transitionalSlotEncoding(element.namespaceUri, localName, getLocalName(name)));
+		else if (convertValues) next = transitionalValue(raw, namespaceUri, attributeSlotEncoding(element, name));
 		if (next === raw) continue;
 		converted ??= { ...attributes };
 		converted[name] = next;
@@ -99,7 +108,7 @@ const toTransitional = (element, inherited) => {
 	const namespaceUri = element.namespaceUri === void 0 ? "" : toTransitionalNamespaceUri(element.namespaceUri);
 	const localName = getLocalName(element.name);
 	const textEncoding = transitionalSlotEncoding(element.namespaceUri, localName);
-	const attributes = element.attributes ? convertAttributes(element, namespaceUri, localName, origin === "strict") : void 0;
+	const attributes = element.attributes ? convertAttributes(element, namespaceUri, origin === "strict") : void 0;
 	let elements;
 	for (const [index, child] of (element.elements ?? []).entries()) {
 		const next = origin === "strict" && child.type === "text" && typeof child.text === "string" ? childWithValue(child, transitionalValue(child.text, namespaceUri, textEncoding)) : toTransitional(child, origin);
@@ -363,5 +372,39 @@ const createCapturedXmlSanitizer = (options) => {
 		return sanitized;
 	};
 };
+const STRICT_BY_TRANSITIONAL = new Map([...TRANSITIONAL_NAMESPACE_BY_STRICT_URI].map(([strict, transitional]) => [transitional, strict]));
+const strictPercentage = (value, unit) => /^-?\d+(?:\.\d+)?$/u.test(value) ? `${Number(value) / NUMBERS_PER_PERCENT[unit]}%` : value;
+/** Translate generated markup by namespace/slot, never by a user-text replacement. */
+const strictElement = (element) => {
+	const localName = getLocalName(element.name);
+	for (const [name, value] of Object.entries(element.attributes ?? {})) {
+		if (typeof value !== "string") continue;
+		if (name === "xmlns" || name.startsWith("xmlns:")) {
+			element.attributes[name] = STRICT_BY_TRANSITIONAL.get(value) ?? value;
+			continue;
+		}
+		if (localName === "graphicData" && toTransitionalNamespaceUri(element.namespaceUri ?? "") === NAMESPACES.a && name === "uri") {
+			element.attributes[name] = STRICT_BY_TRANSITIONAL.get(value) ?? value;
+			continue;
+		}
+		const encoding = attributeSlotEncoding(element, name);
+		if (encoding?.percent && (!encoding.measure || getAttributeByNamespaceUri(element, WORDPROCESSINGML_NAMESPACE_URIS, "type") === "pct")) element.attributes[name] = strictPercentage(value, encoding.percent);
+	}
+	const encoding = transitionalSlotEncoding(element.namespaceUri, localName);
+	for (const child of element.elements ?? []) if (child.type === "element") strictElement(child);
+	else if (child.type === "text" && typeof child.text === "string" && encoding?.percent) child.text = strictPercentage(child.text, encoding.percent);
+};
+/** Serialize a generated, scope-bound block in the source part's profile. */
+const cloneSourceProfileTree = (element) => cloneElement(element, {
+	...element.attributes === void 0 ? {} : { attributes: { ...element.attributes } },
+	...element.elements === void 0 ? {} : { elements: element.elements.map(cloneSourceProfileTree) }
+});
+const captureSourceProfileXml = (element, conformance) => {
+	const owned = cloneSourceProfileTree(element);
+	if (conformance === DOCX_CONFORMANCE_CLASSES.STRICT) strictElement(owned);
+	const xml = elementToXml(owned);
+	if (!isSafeCapturedXmlDocument(xml)) panic("Generated source-profile markup is unsafe.");
+	return xml;
+};
 //#endregion
-export { UntranslatableStrictNamespaceError, captureVerbatimXml, createCapturedXmlSanitizer, isSafeCapturedXmlDocument };
+export { UntranslatableStrictNamespaceError, captureSourceProfileXml, captureVerbatimXml, createCapturedXmlSanitizer, isSafeCapturedXmlDocument };

package/dist/i18n/messages/ar.gen.d.ts

@@ -0,0 +1,5 @@
+import { FolioMessages } from "../messages.js";
+//#region src/i18n/messages/ar.gen.d.ts
+declare const messages: FolioMessages;
+//#endregion
+export { messages as default, messages };