@cj-tech-master/excelts 3.0.1-canary.20251230173853.c3c79fa → 3.1.0-canary.20260102041828.82ed571

package/dist/cjs/modules/excel/stream/sheet-rels-writer.js

@@ -0,0 +1,94 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.SheetRelsWriter = void 0;
+const utils_1 = require("@utils/utils");
+const rel_type_1 = require("@excel/xlsx/rel-type");
+const ooxml_paths_1 = require("@excel/utils/ooxml-paths");
+class HyperlinksProxy {
+    constructor(sheetRelsWriter) {
+        this.writer = sheetRelsWriter;
+    }
+    push(hyperlink) {
+        this.writer.addHyperlink(hyperlink);
+    }
+}
+class SheetRelsWriter {
+    constructor(options) {
+        // in a workbook, each sheet will have a number
+        this.id = options.id;
+        // count of all relationships
+        this.count = 0;
+        // keep record of all hyperlinks
+        this._hyperlinks = [];
+        this._workbook = options.workbook;
+    }
+    get stream() {
+        if (!this._stream) {
+            this._stream = this._workbook._openStream((0, ooxml_paths_1.worksheetRelsPath)(this.id));
+        }
+        return this._stream;
+    }
+    get length() {
+        return this._hyperlinks.length;
+    }
+    each(fn) {
+        return this._hyperlinks.forEach(fn);
+    }
+    get hyperlinksProxy() {
+        return this._hyperlinksProxy || (this._hyperlinksProxy = new HyperlinksProxy(this));
+    }
+    addHyperlink(hyperlink) {
+        // Write to stream
+        const relationship = {
+            Target: hyperlink.target,
+            Type: rel_type_1.RelType.Hyperlink,
+            TargetMode: "External"
+        };
+        const rId = this._writeRelationship(relationship);
+        // store sheet stuff for later
+        this._hyperlinks.push({
+            rId,
+            address: hyperlink.address
+        });
+    }
+    addMedia(media) {
+        return this._writeRelationship(media);
+    }
+    addRelationship(rel) {
+        return this._writeRelationship(rel);
+    }
+    commit() {
+        if (this.count) {
+            // write xml utro
+            this._writeClose();
+            // and close stream
+            this.stream.end();
+        }
+    }
+    // ================================================================================
+    _writeOpen() {
+        this.stream.write(`<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+       <Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">`);
+    }
+    _writeRelationship(relationship) {
+        if (!this.count) {
+            this._writeOpen();
+        }
+        const rId = `rId${++this.count}`;
+        if (relationship.TargetMode) {
+            this.stream.write(`<Relationship Id="${rId}"` +
+                ` Type="${relationship.Type}"` +
+                ` Target="${(0, utils_1.xmlEncode)(relationship.Target)}"` +
+                ` TargetMode="${relationship.TargetMode}"` +
+                "/>");
+        }
+        else {
+            this.stream.write(`<Relationship Id="${rId}" Type="${relationship.Type}" Target="${relationship.Target}"/>`);
+        }
+        return rId;
+    }
+    _writeClose() {
+        this.stream.write("</Relationships>");
+    }
+}
+exports.SheetRelsWriter = SheetRelsWriter;

package/dist/cjs/modules/excel/stream/workbook-reader.browser.js

@@ -0,0 +1,447 @@
+"use strict";
+/**
+ * WorkbookReader - Browser Streaming Workbook Reader
+ *
+ * This module contains the full shared implementation for the streaming
+ * workbook reader and a browser-compatible `WorkbookReader` that buffers
+ * waiting worksheets in memory.
+ *
+ * Node.js uses `workbook-reader.ts`, which extends the same base implementation
+ * with filesystem-specific features (filename input + temp-file buffering).
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.WorkbookReader = exports.WorkbookReaderOptionsSchema = exports.WorkbookReaderBase = void 0;
+const _stream_1 = require("@stream");
+const parse_1 = require("@archive/parse");
+const iterate_stream_1 = require("@excel/utils/iterate-stream");
+const ooxml_paths_1 = require("@excel/utils/ooxml-paths");
+const parse_sax_1 = require("@excel/utils/parse-sax");
+const styles_xform_1 = require("@excel/xlsx/xform/style/styles-xform");
+const workbook_xform_1 = require("@excel/xlsx/xform/book/workbook-xform");
+const relationships_xform_1 = require("@excel/xlsx/xform/core/relationships-xform");
+const worksheet_reader_1 = require("@excel/stream/worksheet-reader");
+const hyperlink_reader_1 = require("@excel/stream/hyperlink-reader");
+// ============================================================================
+// Base Class
+// ============================================================================
+class WorkbookReaderBase extends _stream_1.EventEmitter {
+    getHyperlinkReader(sheetNo) {
+        return this._hyperlinkReadersBySheetNo?.[String(sheetNo)];
+    }
+    getHyperlink(sheetNo, rId) {
+        return this.getHyperlinkReader(sheetNo)?.hyperlinks?.[rId];
+    }
+    getHyperlinkTarget(sheetNo, rId) {
+        return this.getHyperlink(sheetNo, rId)?.target;
+    }
+    constructor(input, options, WorksheetReaderClass, HyperlinkReaderClass) {
+        super();
+        this.input = input;
+        this.WorksheetReaderClass = WorksheetReaderClass;
+        this.HyperlinkReaderClass = HyperlinkReaderClass;
+        this.options = {
+            worksheets: "emit",
+            sharedStrings: "cache",
+            hyperlinks: "ignore",
+            styles: "ignore",
+            entries: "ignore",
+            ...options
+        };
+        this.styles = new styles_xform_1.StylesXform();
+        this.styles.init();
+    }
+    // Default implementation for CommonInput types
+    _getStream(input) {
+        if (input instanceof _stream_1.Readable) {
+            return input;
+        }
+        // Accept Web ReadableStream (browser fetch() body, Node 18+ fetch(), etc.)
+        if (input &&
+            typeof input === "object" &&
+            typeof input.getReader === "function") {
+            const fromWeb = _stream_1.Readable.fromWeb;
+            if (typeof fromWeb === "function") {
+                return fromWeb(input);
+            }
+            // Browser wrapper supports `{ stream }` constructor option.
+            // Node's Readable does not, so this is best-effort.
+            try {
+                return new _stream_1.Readable({
+                    stream: input
+                });
+            }
+            catch {
+                throw new Error("Could not recognise input: ReadableStream");
+            }
+        }
+        let data = input;
+        if (data instanceof ArrayBuffer) {
+            data = new Uint8Array(data);
+        }
+        if (data instanceof Uint8Array) {
+            // Cross-platform: both Node's Readable and our browser Readable implement `.from()`.
+            return _stream_1.Readable.from([data]);
+        }
+        throw new Error(`Could not recognise input: ${input}`);
+    }
+    _cleanupWaitingWorksheets(_waitingWorksheets) {
+        // Default: attempt best-effort cleanup if the stored object provides it.
+        for (const ws of _waitingWorksheets) {
+            if (ws && typeof ws.cleanup === "function") {
+                ws.cleanup();
+            }
+        }
+    }
+    // Unified implementations using passed-in classes
+    _createWorksheetReader(params) {
+        return new this.WorksheetReaderClass({
+            workbook: this,
+            id: params.id,
+            iterator: params.iterator,
+            options: params.options
+        });
+    }
+    _createHyperlinkReader(params) {
+        return new this.HyperlinkReaderClass({
+            workbook: this,
+            id: params.id,
+            iterator: params.iterator,
+            options: params.options
+        });
+    }
+    async read(input, options) {
+        try {
+            for await (const { eventType, value } of this.parse(input, options)) {
+                switch (eventType) {
+                    case "shared-strings":
+                        this.emit(eventType, value);
+                        break;
+                    case "worksheet":
+                        this.emit(eventType, value);
+                        await value.read();
+                        break;
+                    case "hyperlinks":
+                        this.emit(eventType, value);
+                        break;
+                }
+            }
+            this.emit("end");
+            this.emit("finished");
+        }
+        catch (error) {
+            this.emit("error", error);
+        }
+    }
+    async *[Symbol.asyncIterator]() {
+        for await (const { eventType, value } of this.parse()) {
+            if (eventType === "worksheet") {
+                yield value;
+            }
+        }
+    }
+    async *parse(input, options) {
+        if (options) {
+            this.options = options;
+        }
+        const stream = (this.stream = this._getStream(input || this.input));
+        const waitingWorksheets = [];
+        try {
+            for await (const event of this._parseEntries(stream)) {
+                if (event.eventType === "waiting-worksheet") {
+                    const stored = await this._storeWaitingWorksheet(event.sheetNo, event.entry);
+                    waitingWorksheets.push(stored);
+                }
+                else {
+                    yield event;
+                }
+            }
+            yield* this._processWaitingWorksheets(waitingWorksheets);
+        }
+        catch (error) {
+            this._cleanupWaitingWorksheets(waitingWorksheets);
+            throw error;
+        }
+    }
+    _emitEntry(payload) {
+        if (this.options.entries === "emit") {
+            this.emit("entry", payload);
+        }
+    }
+    async _parseRels(entry) {
+        const xform = new relationships_xform_1.RelationshipsXform();
+        this.workbookRels = await xform.parseStream((0, iterate_stream_1.iterateStream)(entry));
+        // Build fast lookup for worksheet relationship ids.
+        this._workbookRelIdByTarget = Object.create(null);
+        for (const rel of this.workbookRels || []) {
+            if (rel?.Target && rel?.Id) {
+                this._workbookRelIdByTarget[rel.Target] = rel.Id;
+            }
+        }
+    }
+    async _parseWorkbook(entry) {
+        this._emitEntry({ type: "workbook" });
+        const workbook = new workbook_xform_1.WorkbookXform();
+        this.model = await workbook.parseStream((0, iterate_stream_1.iterateStream)(entry));
+        this.properties = workbook.map?.workbookPr;
+        // Build fast lookup for sheet metadata by relationship id.
+        this._sheetByRelId = Object.create(null);
+        for (const sheet of this.model?.sheets || []) {
+            this._sheetByRelId[sheet.rId] = sheet;
+        }
+    }
+    async *_parseSharedStrings(entry) {
+        this._emitEntry({ type: "shared-strings" });
+        switch (this.options.sharedStrings) {
+            case "cache":
+                this.sharedStrings = [];
+                break;
+            case "emit":
+                break;
+            default:
+                return;
+        }
+        let text = null;
+        let richText = [];
+        let index = 0;
+        let font = null;
+        let inRichText = false;
+        for await (const events of (0, parse_sax_1.parseSax)((0, iterate_stream_1.iterateStream)(entry))) {
+            for (const { eventType, value } of events) {
+                if (eventType === "opentag") {
+                    const node = value;
+                    switch (node.name) {
+                        case "b":
+                            font = font || {};
+                            font.bold = true;
+                            break;
+                        case "charset":
+                            font = font || {};
+                            font.charset = parseInt(node.attributes.charset, 10);
+                            break;
+                        case "color":
+                            font = font || {};
+                            font.color = {};
+                            if (node.attributes.rgb) {
+                                font.color.argb = node.attributes.rgb;
+                            }
+                            if (node.attributes.val) {
+                                font.color.argb = node.attributes.val;
+                            }
+                            if (node.attributes.theme) {
+                                font.color.theme = node.attributes.theme;
+                            }
+                            break;
+                        case "family":
+                            font = font || {};
+                            font.family = parseInt(node.attributes.val, 10);
+                            break;
+                        case "i":
+                            font = font || {};
+                            font.italic = true;
+                            break;
+                        case "outline":
+                            font = font || {};
+                            font.outline = true;
+                            break;
+                        case "rFont":
+                            font = font || {};
+                            font.name = node.attributes.val;
+                            break;
+                        case "r":
+                            inRichText = true;
+                            break;
+                        case "si":
+                            font = null;
+                            richText = [];
+                            text = null;
+                            inRichText = false;
+                            break;
+                        case "sz":
+                            font = font || {};
+                            font.size = parseInt(node.attributes.val, 10);
+                            break;
+                        case "strike":
+                            font = font || {};
+                            font.strike = true;
+                            break;
+                        case "t":
+                            text = null;
+                            break;
+                        case "u":
+                            font = font || {};
+                            font.underline = true;
+                            break;
+                        case "vertAlign":
+                            font = font || {};
+                            font.vertAlign = node.attributes.val;
+                            break;
+                    }
+                }
+                else if (eventType === "text") {
+                    text = text ? text + value : value;
+                }
+                else if (eventType === "closetag") {
+                    const node = value;
+                    switch (node.name) {
+                        case "r":
+                            if (inRichText) {
+                                richText.push({ font, text });
+                                font = null;
+                                text = null;
+                            }
+                            break;
+                        case "si":
+                            if (this.options.sharedStrings === "cache") {
+                                this.sharedStrings.push(richText.length ? { richText } : text || "");
+                            }
+                            else if (this.options.sharedStrings === "emit") {
+                                yield { index: index++, text: richText.length ? { richText } : text || "" };
+                            }
+                            richText = [];
+                            font = null;
+                            text = null;
+                            inRichText = false;
+                            break;
+                    }
+                }
+            }
+        }
+    }
+    async _parseStyles(entry) {
+        this._emitEntry({ type: "styles" });
+        if (this.options.styles === "cache") {
+            this.styles = new styles_xform_1.StylesXform();
+            await this.styles.parseStream((0, iterate_stream_1.iterateStream)(entry));
+        }
+    }
+    *_parseWorksheet(iterator, sheetNo) {
+        this._emitEntry({ type: "worksheet", id: sheetNo });
+        const sheetNoNumber = parseInt(sheetNo, 10);
+        const worksheetReader = this._createWorksheetReader({
+            id: sheetNoNumber,
+            iterator,
+            options: this.options
+        });
+        // Preserve original sheet index from the zip path. `worksheetReader.id` may be remapped
+        // later using workbook metadata.
+        worksheetReader.sheetNo = sheetNoNumber;
+        const relId = this._workbookRelIdByTarget?.[(0, ooxml_paths_1.worksheetRelTarget)(sheetNo)];
+        const matchingSheet = relId ? this._sheetByRelId?.[relId] : undefined;
+        if (matchingSheet) {
+            worksheetReader.id = matchingSheet.id;
+            worksheetReader.name = matchingSheet.name;
+            worksheetReader.state = matchingSheet.state;
+        }
+        if (this.options.worksheets === "emit") {
+            yield { eventType: "worksheet", value: worksheetReader };
+        }
+    }
+    async *_parseHyperlinks(iterator, sheetNo) {
+        this._emitEntry({ type: "hyperlinks", id: sheetNo });
+        const hyperlinksReader = this._createHyperlinkReader({
+            id: parseInt(sheetNo, 10),
+            iterator,
+            options: this.options
+        });
+        if (this.options.hyperlinks === "cache") {
+            if (!this._hyperlinkReadersBySheetNo) {
+                this._hyperlinkReadersBySheetNo = Object.create(null);
+            }
+            this._hyperlinkReadersBySheetNo[sheetNo] = hyperlinksReader;
+            const readFn = hyperlinksReader.read;
+            if (typeof readFn === "function") {
+                await readFn.call(hyperlinksReader);
+            }
+            return;
+        }
+        if (this.options.hyperlinks === "emit") {
+            yield { eventType: "hyperlinks", value: hyperlinksReader };
+        }
+    }
+    async *_parseEntries(stream) {
+        const zip = (0, parse_1.createParse)({ forceStream: true });
+        stream.on("error", (err) => zip.emit("error", err));
+        stream.pipe(zip);
+        for await (const entry of (0, iterate_stream_1.iterateStream)(zip)) {
+            let sheetNo;
+            const normalizedPath = (0, ooxml_paths_1.normalizeZipPath)(entry.path);
+            switch (normalizedPath) {
+                case ooxml_paths_1.OOXML_PATHS.rootRels:
+                    break;
+                case ooxml_paths_1.OOXML_PATHS.xlWorkbookRels:
+                    await this._parseRels(entry);
+                    break;
+                case ooxml_paths_1.OOXML_PATHS.xlWorkbook:
+                    await this._parseWorkbook(entry);
+                    break;
+                case ooxml_paths_1.OOXML_PATHS.xlSharedStrings:
+                    for await (const item of this._parseSharedStrings(entry)) {
+                        yield { eventType: "shared-strings", value: item };
+                    }
+                    break;
+                case ooxml_paths_1.OOXML_PATHS.xlStyles:
+                    await this._parseStyles(entry);
+                    break;
+                default:
+                    sheetNo = (0, ooxml_paths_1.getWorksheetNoFromWorksheetPath)(normalizedPath)?.toString();
+                    if (sheetNo) {
+                        // Performance: only wait for sharedStrings when they are actually needed.
+                        const hasPrerequisites = !!this.workbookRels &&
+                            (this.options.sharedStrings !== "cache" || !!this.sharedStrings);
+                        if (hasPrerequisites) {
+                            yield* this._parseWorksheet((0, iterate_stream_1.iterateStream)(entry), sheetNo);
+                            continue;
+                        }
+                        else {
+                            yield { eventType: "waiting-worksheet", sheetNo, entry };
+                            continue;
+                        }
+                    }
+                    sheetNo = (0, ooxml_paths_1.getWorksheetNoFromWorksheetRelsPath)(normalizedPath)?.toString();
+                    if (sheetNo) {
+                        yield* this._parseHyperlinks((0, iterate_stream_1.iterateStream)(entry), sheetNo);
+                        continue;
+                    }
+                    break;
+            }
+            entry.autodrain();
+        }
+    }
+}
+exports.WorkbookReaderBase = WorkbookReaderBase;
+exports.WorkbookReaderOptionsSchema = {
+    worksheets: ["emit", "ignore"],
+    sharedStrings: ["cache", "emit", "ignore"],
+    hyperlinks: ["cache", "emit", "ignore"],
+    styles: ["cache", "ignore"],
+    entries: ["emit", "ignore"]
+};
+class WorkbookReader extends WorkbookReaderBase {
+    constructor(input, options = {}) {
+        super(input, options, worksheet_reader_1.WorksheetReader, hyperlink_reader_1.HyperlinkReader);
+    }
+    async _storeWaitingWorksheet(sheetNo, entry) {
+        const chunks = [];
+        const encoder = new TextEncoder();
+        for await (const chunk of (0, iterate_stream_1.iterateStream)(entry)) {
+            if (chunk instanceof Uint8Array) {
+                chunks.push(chunk);
+            }
+            else if (typeof chunk === "string") {
+                chunks.push(encoder.encode(chunk));
+            }
+        }
+        return { sheetNo, data: chunks };
+    }
+    async *_processWaitingWorksheets(waitingWorksheets) {
+        for (const ws of waitingWorksheets) {
+            const iterator = (async function* () {
+                for (const chunk of ws.data) {
+                    yield chunk;
+                }
+            })();
+            yield* this._parseWorksheet(iterator, ws.sheetNo);
+        }
+    }
+}
+exports.WorkbookReader = WorkbookReader;

package/dist/cjs/modules/excel/stream/workbook-reader.js

@@ -0,0 +1,57 @@
+"use strict";
+/**
+ * WorkbookReader - Node.js Streaming Workbook Reader
+ *
+ * Extends base with file path support and temp file storage for large files.
+ */
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.WorkbookReaderOptionsSchema = exports.WorkbookReader = void 0;
+const fs_1 = __importDefault(require("fs"));
+const os_1 = __importDefault(require("os"));
+const path_1 = require("path");
+const iterate_stream_1 = require("@excel/utils/iterate-stream");
+const worksheet_reader_1 = require("@excel/stream/worksheet-reader");
+const hyperlink_reader_1 = require("@excel/stream/hyperlink-reader");
+const workbook_reader_browser_1 = require("@excel/stream/workbook-reader.browser");
+Object.defineProperty(exports, "WorkbookReaderOptionsSchema", { enumerable: true, get: function () { return workbook_reader_browser_1.WorkbookReaderOptionsSchema; } });
+class WorkbookReader extends workbook_reader_browser_1.WorkbookReaderBase {
+    constructor(input, options = {}) {
+        super(input, options, worksheet_reader_1.WorksheetReader, hyperlink_reader_1.HyperlinkReader);
+        this.input = input;
+    }
+    _getStream(input) {
+        if (typeof input === "string") {
+            return fs_1.default.createReadStream(input);
+        }
+        return super._getStream(input);
+    }
+    async _storeWaitingWorksheet(sheetNo, entry) {
+        const tmpDir = fs_1.default.mkdtempSync((0, path_1.join)(os_1.default.tmpdir(), "excelts-"));
+        const path = (0, path_1.join)(tmpDir, `sheet${sheetNo}.xml`);
+        const cleanup = () => fs_1.default.rm(tmpDir, { recursive: true, force: true }, () => { });
+        const writePromise = new Promise((resolve, reject) => {
+            const tempStream = fs_1.default.createWriteStream(path);
+            tempStream.on("error", reject);
+            tempStream.on("finish", resolve);
+            entry.pipe(tempStream);
+        });
+        return { sheetNo, path, cleanup, writePromise };
+    }
+    async *_processWaitingWorksheets(waitingWorksheets) {
+        for (const ws of waitingWorksheets) {
+            await ws.writePromise;
+            const fileStream = fs_1.default.createReadStream(ws.path);
+            try {
+                yield* this._parseWorksheet((0, iterate_stream_1.iterateStream)(fileStream), ws.sheetNo);
+            }
+            finally {
+                fileStream.close();
+                ws.cleanup();
+            }
+        }
+    }
+}
+exports.WorkbookReader = WorkbookReader;