parses xjustiz and orders attachments

src/lib/zip-processing.ts

@@ -10,9 +10,8 @@ export type ZipAttachment = {
 };
 
 export type ProcessedZipArchive = {
-	name: string;
-	attachments: ZipAttachment[];
-	xjustizNachrichtXml: Uint8Array | null;
+	name: string
+	attachments: ZipAttachment[]
 };
 
 const IMAGE_EXTENSIONS = new Set([
@@ -30,6 +29,7 @@ const IMAGE_EXTENSIONS = new Set([
 ]);
 
 const ZIP_META_FILE_NAME = "xjustiz_nachricht.xml";
+const XML_TEXT_DECODER = new TextDecoder();
 
 const unzipArchive = (data: Uint8Array) =>
 	new Promise<Record<string, Uint8Array>>((resolve, reject) => {
@@ -56,6 +56,126 @@ const isPdf = (path: string) => getExtension(path) === ".pdf";
 
 const isImage = (path: string) => IMAGE_EXTENSIONS.has(getExtension(path));
 
+const normalizeKey = (value: string) => value.replaceAll("\\", "/").toLowerCase().trim();
+
+const matchesElementName = (element: Element, requestedName: string) => {
+	const requestedLocalName = requestedName.split(".").pop() ?? requestedName;
+
+	return (
+		element.localName === requestedName ||
+		element.localName === requestedLocalName ||
+		element.tagName === requestedName ||
+		element.tagName === requestedLocalName ||
+		element.tagName.endsWith(`:${requestedLocalName}`)
+	);
+};
+
+const getDirectChildByLocalName = (element: Element, localName: string) =>
+	Array.from(element.children).find((child) => matchesElementName(child, localName)) ?? null;
+
+const getFirstDescendantByLocalName = (root: Document | Element, localName: string) =>
+	Array.from(root.querySelectorAll("*")).find((element) => matchesElementName(element, localName)) ?? null;
+
+const getTextContentByPath = (root: Element, path: string[]) => {
+	let current: Element | null = root;
+
+	for (const localName of path) {
+		current = current ? getDirectChildByLocalName(current, localName) : null;
+
+		if (!current) {
+			return null;
+		}
+	}
+
+	const textContent = current.textContent?.trim();
+
+	return textContent ? textContent : null;
+};
+
+const parseXmlDocument = (xmlBytes: Uint8Array) => {
+	const parser = new DOMParser();
+	const document = parser.parseFromString(XML_TEXT_DECODER.decode(xmlBytes), "application/xml");
+
+	if (document.querySelector("parsererror")) {
+		return null;
+	}
+
+	return document;
+};
+
+const getXJustizDocumentNames = (xmlBytes: Uint8Array) => {
+	const document = parseXmlDocument(xmlBytes);
+
+	if (!document) {
+		return [];
+	}
+
+	const schriftgutobjekte = getFirstDescendantByLocalName(document, "schriftgutobjekte");
+
+	if (!schriftgutobjekte) {
+		return [];
+	}
+
+	const dokumentNodes = Array.from(schriftgutobjekte.children).filter((child) =>
+		matchesElementName(child, "dokument"),
+	);
+
+	return dokumentNodes
+		.map((dokument) => getTextContentByPath(dokument, ["xjustiz.fachspezifischeDaten", "datei", "dateiname"]))
+		.filter((name): name is string => Boolean(name));
+};
+
+const orderAttachmentsByDocumentNames = (attachments: ZipAttachment[], documentNames: string[]) => {
+	const buckets = new Map<string, ZipAttachment[]>();
+
+	for (const attachment of attachments) {
+		const keys = new Set([normalizeKey(attachment.path), normalizeKey(attachment.name)]);
+
+		for (const key of keys) {
+			const bucket = buckets.get(key) ?? [];
+			bucket.push(attachment);
+			buckets.set(key, bucket);
+		}
+	}
+
+	const usedAttachments = new Set<ZipAttachment>();
+	const orderedAttachments: ZipAttachment[] = [];
+
+	const takeAttachment = (key: string) => {
+		const bucket = buckets.get(key);
+
+		if (!bucket) {
+			return null;
+		}
+
+		const nextAttachment = bucket.find((attachment) => !usedAttachments.has(attachment)) ?? null;
+
+		if (nextAttachment) {
+			usedAttachments.add(nextAttachment);
+		}
+
+		return nextAttachment;
+	};
+
+	for (const documentName of documentNames) {
+		const normalizedName = normalizeKey(documentName);
+		const normalizedBaseName = normalizeKey(getBaseName(documentName));
+		const matchedAttachment = takeAttachment(normalizedName) ?? takeAttachment(normalizedBaseName);
+
+		if (matchedAttachment) {
+			orderedAttachments.push(matchedAttachment);
+		}
+	}
+
+	for (const attachment of attachments) {
+		if (!usedAttachments.has(attachment)) {
+			orderedAttachments.push(attachment);
+		}
+	}
+
+	return orderedAttachments;
+};
+
 export const extractZipArchive = async (file: File): Promise<ProcessedZipArchive> => {
 	const archiveBytes = new Uint8Array(await file.arrayBuffer());
 	const files = await unzipArchive(archiveBytes);
@@ -91,9 +211,12 @@ export const extractZipArchive = async (file: File): Promise<ProcessedZipArchive
 		}
 	}
 
+	const documentNames = xjustizNachrichtXml ? getXJustizDocumentNames(xjustizNachrichtXml) : [];
+	const orderedAttachments = orderAttachmentsByDocumentNames(attachments, documentNames);
+	const archiveName = documentNames[0] ? getBaseName(documentNames[0]) : file.name;
+
 	return {
-		name: file.name,
-		attachments,
-		xjustizNachrichtXml,
+		name: archiveName,
+		attachments: orderedAttachments
 	};
 };