@llamaindex/liteparse 1.0.0

package/dist/src/conversion/convertToPdf.js

@@ -0,0 +1,337 @@
+import { promises as fs, constants as fsConstants } from "fs";
+import { spawn } from "child_process";
+import path from "path";
+import os from "os";
+// File extension categories
+export const officeExtensions = [
+    ".doc",
+    ".docx",
+    ".docm",
+    ".dot",
+    ".dotm",
+    ".dotx",
+    ".odt",
+    ".ott",
+    ".ppt",
+    ".pptx",
+    ".pptm",
+    ".pot",
+    ".potm",
+    ".potx",
+    ".odp",
+    ".otp",
+    ".rtf",
+    ".pages",
+    ".key",
+];
+export const spreadsheetExtensions = [
+    ".xls",
+    ".xlsx",
+    ".xlsm",
+    ".xlsb",
+    ".ods",
+    ".ots",
+    ".csv",
+    ".tsv",
+    ".numbers",
+];
+export const imageExtensions = [
+    ".jpg",
+    ".jpeg",
+    ".png",
+    ".gif",
+    ".bmp",
+    ".tiff",
+    ".tif",
+    ".webp",
+    ".svg",
+];
+export const htmlExtensions = [".htm", ".html", ".xhtml"];
+/**
+ * Guess file extension from file content
+ */
+export async function guessFileExtension(filePath) {
+    const ext = path.extname(filePath).toLowerCase();
+    if (ext) {
+        return ext;
+    }
+    // Read first few bytes to detect file type
+    const buffer = Buffer.alloc(16);
+    const fd = await fs.open(filePath, "r");
+    await fd.read(buffer, 0, 16, 0);
+    await fd.close();
+    // PDF: %PDF
+    if (buffer.toString("utf-8", 0, 4) === "%PDF") {
+        return ".pdf";
+    }
+    // PNG: 89 50 4E 47
+    if (buffer[0] === 0x89 && buffer[1] === 0x50 && buffer[2] === 0x4e && buffer[3] === 0x47) {
+        return ".png";
+    }
+    // JPEG: FF D8 FF
+    if (buffer[0] === 0xff && buffer[1] === 0xd8 && buffer[2] === 0xff) {
+        return ".jpg";
+    }
+    // ZIP-based formats (docx, xlsx, etc): PK
+    if (buffer[0] === 0x50 && buffer[1] === 0x4b) {
+        // Could be docx, xlsx, pptx, odt, etc.
+        return ".docx"; // Default to docx for now
+    }
+    return ext || ".pdf";
+}
+/**
+ * Execute command with timeout
+ */
+async function executeCommand(command, args, timeoutMs = 60000) {
+    return new Promise((resolve, reject) => {
+        const proc = spawn(command, args, {
+            stdio: ["ignore", "pipe", "pipe"],
+        });
+        let stdout = "";
+        let stderr = "";
+        proc.stdout?.on("data", (data) => {
+            stdout += data.toString();
+        });
+        proc.stderr?.on("data", (data) => {
+            stderr += data.toString();
+        });
+        const timeout = setTimeout(() => {
+            proc.kill();
+            reject(new Error(`Command timeout after ${timeoutMs}ms`));
+        }, timeoutMs);
+        proc.on("close", (code) => {
+            clearTimeout(timeout);
+            if (code === 0) {
+                resolve(stdout);
+            }
+            else {
+                reject(new Error(`Command failed with code ${code}: ${stderr}`));
+            }
+        });
+        proc.on("error", (err) => {
+            clearTimeout(timeout);
+            reject(err);
+        });
+    });
+}
+/* Execute a command for PowerShell */
+async function executePowerShell(command, timeoutMs = 60000) {
+    return executeCommand("powershell", ["-NoProfile", "-Command", command], timeoutMs);
+}
+/**
+ * Check if a command is available
+ */
+async function isCommandAvailable(command) {
+    try {
+        await executeCommand("which", [command], 5000);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
+async function isCommandAvailableWindows(command) {
+    try {
+        await executePowerShell(`Get-Command ${command}`, 5000);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
+/**
+ * Check if a file path exists and is executable
+ */
+async function isPathExecutable(filePath) {
+    try {
+        await fs.access(filePath, fsConstants.X_OK);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
+/**
+ * Find LibreOffice command - handles different installation methods
+ */
+export async function findLibreOfficeCommand() {
+    // Check for 'libreoffice' in PATH (Linux, some macOS setups)
+    if ((await isCommandAvailable("libreoffice")) ||
+        (await isCommandAvailableWindows("libreoffice"))) {
+        return "libreoffice";
+    }
+    // Check for 'soffice' in PATH
+    if ((await isCommandAvailable("soffice")) || (await isCommandAvailableWindows("soffice"))) {
+        return "soffice";
+    }
+    // macOS: Check standard application paths
+    const macOSPaths = [
+        "/Applications/LibreOffice.app/Contents/MacOS/soffice",
+        "/Applications/LibreOffice.app/Contents/MacOS/libreoffice",
+    ];
+    const windowsPaths = ["C:\\Program Files\\Libreoffice\\program\\soffice.exe"];
+    for (const libPath of macOSPaths) {
+        if (await isPathExecutable(libPath)) {
+            return libPath;
+        }
+    }
+    for (const libPath of windowsPaths) {
+        if (await isPathExecutable(libPath)) {
+            return libPath;
+        }
+    }
+    return null;
+}
+/**
+ * Find ImageMagick command - handles v6 (convert) and v7 (magick)
+ */
+export async function findImageMagickCommand() {
+    // ImageMagick v7 uses 'magick' command
+    if ((await isCommandAvailable("magick")) || (await isCommandAvailableWindows("magick"))) {
+        return { command: "magick", args: [] };
+    }
+    // ImageMagick v6 uses 'convert' command
+    if ((await isCommandAvailable("convert")) || (await isCommandAvailableWindows("convert"))) {
+        return { command: "convert", args: [] };
+    }
+    return null;
+}
+/**
+ * Convert office documents using LibreOffice
+ */
+export async function convertOfficeDocument(filePath, outputDir) {
+    const libreOfficeCmd = await findLibreOfficeCommand();
+    if (!libreOfficeCmd) {
+        throw new Error("LibreOffice is not installed. Please install LibreOffice to convert office documents. On macOS: brew install --cask libreoffice, On Ubuntu: apt-get install libreoffice, On Windows: choco install libreoffice-fresh");
+    }
+    await executeCommand(libreOfficeCmd, ["--headless", "--invisible", "--convert-to", "pdf", "--outdir", outputDir, filePath], 120000 // 2 minutes timeout
+    );
+    // LibreOffice creates output with same name but .pdf extension
+    const baseName = path.basename(filePath, path.extname(filePath));
+    const pdfPath = path.join(outputDir, `${baseName}.pdf`);
+    // Verify the PDF was created
+    try {
+        await fs.access(pdfPath);
+        return pdfPath;
+    }
+    catch {
+        throw new Error("LibreOffice conversion succeeded but output PDF not found");
+    }
+}
+// Extensions that require Ghostscript for ImageMagick conversion
+const ghostscriptRequiredExtensions = [".svg", ".eps", ".ps", ".ai"];
+/**
+ * Convert images to PDF using ImageMagick
+ */
+export async function convertImageToPdf(filePath, outputDir) {
+    const imageMagick = await findImageMagickCommand();
+    if (!imageMagick) {
+        throw new Error("ImageMagick is not installed. Please install ImageMagick to convert images. On macOS: brew install imagemagick, On Ubuntu: apt-get install imagemagick, On Windows: choco install imagemagick.app");
+    }
+    const ext = path.extname(filePath).toLowerCase();
+    const needsGhostscript = ghostscriptRequiredExtensions.includes(ext);
+    // Check for Ghostscript if needed for this file type
+    if (needsGhostscript) {
+        const hasGhostscript = (await isCommandAvailable("gs")) || (await isCommandAvailableWindows("gs"));
+        if (!hasGhostscript) {
+            throw new Error(`Ghostscript is required to convert ${ext.toUpperCase().slice(1)} files but is not installed. ` +
+                "On macOS: brew install ghostscript, On Ubuntu: apt-get install ghostscript, On Windows: choco install ghostscript");
+        }
+    }
+    const baseName = path.basename(filePath, path.extname(filePath));
+    const pdfPath = path.join(outputDir, `${baseName}.pdf`);
+    try {
+        await executeCommand(imageMagick.command, [...imageMagick.args, filePath, "-density", "150", "-units", "PixelsPerInch", pdfPath], 60000);
+    }
+    catch (error) {
+        const errorMsg = error instanceof Error ? error.message : String(error);
+        // Provide better error messages for common issues
+        if (errorMsg.includes("gs") && errorMsg.includes("command not found")) {
+            throw new Error(`Ghostscript is required to convert ${ext.toUpperCase().slice(1)} files but is not installed. ` +
+                "On macOS: brew install ghostscript, On Ubuntu: apt-get install ghostscript, On Windows: choco install ghostscript", {
+                cause: error,
+            });
+        }
+        if (errorMsg.includes("FailedToExecuteCommand") && errorMsg.includes("gs")) {
+            throw new Error(`Ghostscript failed during ${ext.toUpperCase().slice(1)} conversion. ` +
+                "Ensure Ghostscript is properly installed: brew install ghostscript", {
+                cause: error,
+            });
+        }
+        throw error;
+    }
+    return pdfPath;
+}
+/**
+ * Main conversion function
+ */
+export async function convertToPdf(filePath) {
+    try {
+        // Check if file exists
+        try {
+            await fs.access(filePath);
+        }
+        catch {
+            return {
+                message: `File not found: ${filePath}`,
+                code: "FILE_NOT_FOUND",
+            };
+        }
+        // Get file extension
+        const extension = await guessFileExtension(filePath);
+        // If already PDF, return as-is
+        if (extension === ".pdf") {
+            return {
+                pdfPath: filePath,
+                originalExtension: extension,
+            };
+        }
+        // Create temp directory for output
+        const tmpDir = await fs.mkdtemp(path.join(os.tmpdir(), "liteparse-"));
+        // Convert based on file type
+        let pdfPath;
+        if (officeExtensions.includes(extension)) {
+            pdfPath = await convertOfficeDocument(filePath, tmpDir);
+        }
+        else if (spreadsheetExtensions.includes(extension)) {
+            pdfPath = await convertOfficeDocument(filePath, tmpDir);
+        }
+        else if (imageExtensions.includes(extension)) {
+            pdfPath = await convertImageToPdf(filePath, tmpDir);
+        }
+        else {
+            // Unsupported format
+            // Assume its a text-based format and pass through text directly
+            const content = await fs.readFile(filePath, "utf-8");
+            return {
+                content,
+            };
+        }
+        return {
+            pdfPath,
+            originalExtension: extension,
+        };
+    }
+    catch (error) {
+        return {
+            message: error instanceof Error ? error.message : String(error),
+            code: "CONVERSION_ERROR",
+        };
+    }
+}
+/**
+ * Clean up temporary conversion files
+ */
+export async function cleanupConversionFiles(pdfPath) {
+    try {
+        // Only delete if in temp directory
+        if (pdfPath.includes(os.tmpdir())) {
+            const dir = path.dirname(pdfPath);
+            await fs.rm(dir, { recursive: true, force: true });
+        }
+    }
+    catch {
+        // Ignore cleanup errors
+    }
+}
+//# sourceMappingURL=convertToPdf.js.map

package/dist/src/conversion/convertToPdf.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"convertToPdf.js","sourceRoot":"","sources":["../../../src/conversion/convertToPdf.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,IAAI,EAAE,EAAE,SAAS,IAAI,WAAW,EAAE,MAAM,IAAI,CAAC;AAC9D,OAAO,EAAE,KAAK,EAAE,MAAM,eAAe,CAAC;AACtC,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,MAAM,IAAI,CAAC;AAgBpB,4BAA4B;AAC5B,MAAM,CAAC,MAAM,gBAAgB,GAAG;IAC9B,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;IACN,MAAM;IACN,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;IACN,MAAM;IACN,MAAM;IACN,QAAQ;IACR,MAAM;CACP,CAAC;AAEF,MAAM,CAAC,MAAM,qBAAqB,GAAG;IACnC,MAAM;IACN,OAAO;IACP,OAAO;IACP,OAAO;IACP,MAAM;IACN,MAAM;IACN,MAAM;IACN,MAAM;IACN,UAAU;CACX,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG;IAC7B,MAAM;IACN,OAAO;IACP,MAAM;IACN,MAAM;IACN,MAAM;IACN,OAAO;IACP,MAAM;IACN,OAAO;IACP,MAAM;CACP,CAAC;AAEF,MAAM,CAAC,MAAM,cAAc,GAAG,CAAC,MAAM,EAAE,OAAO,EAAE,QAAQ,CAAC,CAAC;AAE1D;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IACjD,IAAI,GAAG,EAAE,CAAC;QACR,OAAO,GAAG,CAAC;IACb,CAAC;IAED,2CAA2C;IAC3C,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;IAChC,MAAM,EAAE,GAAG,MAAM,EAAE,CAAC,IAAI,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;IACxC,MAAM,EAAE,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC,CAAC;IAChC,MAAM,EAAE,CAAC,KAAK,EAAE,CAAC;IAEjB,YAAY;IACZ,IAAI,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,CAAC,EAAE,CAAC,CAAC,KAAK,MAAM,EAAE,CAAC;QAC9C,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,mBAAmB;IACnB,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACzF,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,iBAAiB;IACjB,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACnE,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,0CAA0C;IAC1C,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC7C,uCAAuC;QACvC,OAAO,OAAO,CAAC,CAAC,0BAA0B;IAC5C,CAAC;IAED,OAAO,GAAG,IAAI,MAAM,CAAC;AACvB,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,cAAc,CAAC,OAAe,EAAE,IAAc,EAAE,SAAS,GAAG,KAAK;IAC9E,OAAO,IAAI,OAAO,CAAC,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;QACrC,MAAM,IAAI,GAAG,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE;YAChC,KAAK,EAAE,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,CAAC;SAClC,CAAC,CAAC;QAEH,IAAI,MAAM,GAAG,EAAE,CAAC;QAChB,IAAI,MAAM,GAAG,EAAE,CAAC;QAEhB,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAI,EAAE,EAAE;YAC/B,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;QAC5B,CAAC,CAAC,CAAC;QAEH,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAI,EAAE,EAAE;YAC/B,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;QAC5B,CAAC,CAAC,CAAC;QAEH,MAAM,OAAO,GAAG,UAAU,CAAC,GAAG,EAAE;YAC9B,IAAI,CAAC,IAAI,EAAE,CAAC;YACZ,MAAM,CAAC,IAAI,KAAK,CAAC,yBAAyB,SAAS,IAAI,CAAC,CAAC,CAAC;QAC5D,CAAC,EAAE,SAAS,CAAC,CAAC;QAEd,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE;YACxB,YAAY,CAAC,OAAO,CAAC,CAAC;YACtB,IAAI,IAAI,KAAK,CAAC,EAAE,CAAC;gBACf,OAAO,CAAC,MAAM,CAAC,CAAC;YAClB,CAAC;iBAAM,CAAC;gBACN,MAAM,CAAC,IAAI,KAAK,CAAC,4BAA4B,IAAI,KAAK,MAAM,EAAE,CAAC,CAAC,CAAC;YACnE,CAAC;QACH,CAAC,CAAC,CAAC;QAEH,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE;YACvB,YAAY,CAAC,OAAO,CAAC,CAAC;YACtB,MAAM,CAAC,GAAG,CAAC,CAAC;QACd,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC;AAED,sCAAsC;AACtC,KAAK,UAAU,iBAAiB,CAAC,OAAe,EAAE,SAAS,GAAG,KAAK;IACjE,OAAO,cAAc,CAAC,YAAY,EAAE,CAAC,YAAY,EAAE,UAAU,EAAE,OAAO,CAAC,EAAE,SAAS,CAAC,CAAC;AACtF,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,kBAAkB,CAAC,OAAe;IAC/C,IAAI,CAAC;QACH,MAAM,cAAc,CAAC,OAAO,EAAE,CAAC,OAAO,CAAC,EAAE,IAAI,CAAC,CAAC;QAC/C,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED,KAAK,UAAU,yBAAyB,CAAC,OAAe;IACtD,IAAI,CAAC;QACH,MAAM,iBAAiB,CAAC,eAAe,OAAO,EAAE,EAAE,IAAI,CAAC,CAAC;QACxD,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,gBAAgB,CAAC,QAAgB;IAC9C,IAAI,CAAC;QACH,MAAM,EAAE,CAAC,MAAM,CAAC,QAAQ,EAAE,WAAW,CAAC,IAAI,CAAC,CAAC;QAC5C,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,sBAAsB;IAC1C,6DAA6D;IAC7D,IACE,CAAC,MAAM,kBAAkB,CAAC,aAAa,CAAC,CAAC;QACzC,CAAC,MAAM,yBAAyB,CAAC,aAAa,CAAC,CAAC,EAChD,CAAC;QACD,OAAO,aAAa,CAAC;IACvB,CAAC;IAED,8BAA8B;IAC9B,IAAI,CAAC,MAAM,kBAAkB,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,MAAM,yBAAyB,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC;QAC1F,OAAO,SAAS,CAAC;IACnB,CAAC;IAED,0CAA0C;IAC1C,MAAM,UAAU,GAAG;QACjB,sDAAsD;QACtD,0DAA0D;KAC3D,CAAC;IAEF,MAAM,YAAY,GAAG,CAAC,sDAAsD,CAAC,CAAC;IAE9E,KAAK,MAAM,OAAO,IAAI,UAAU,EAAE,CAAC;QACjC,IAAI,MAAM,gBAAgB,CAAC,OAAO,CAAC,EAAE,CAAC;YACpC,OAAO,OAAO,CAAC;QACjB,CAAC;IACH,CAAC;IAED,KAAK,MAAM,OAAO,IAAI,YAAY,EAAE,CAAC;QACnC,IAAI,MAAM,gBAAgB,CAAC,OAAO,CAAC,EAAE,CAAC;YACpC,OAAO,OAAO,CAAC;QACjB,CAAC;IACH,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,sBAAsB;IAI1C,uCAAuC;IACvC,IAAI,CAAC,MAAM,kBAAkB,CAAC,QAAQ,CAAC,CAAC,IAAI,CAAC,MAAM,yBAAyB,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC;QACxF,OAAO,EAAE,OAAO,EAAE,QAAQ,EAAE,IAAI,EAAE,EAAE,EAAE,CAAC;IACzC,CAAC;IAED,wCAAwC;IACxC,IAAI,CAAC,MAAM,kBAAkB,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,MAAM,yBAAyB,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC;QAC1F,OAAO,EAAE,OAAO,EAAE,SAAS,EAAE,IAAI,EAAE,EAAE,EAAE,CAAC;IAC1C,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,qBAAqB,CAAC,QAAgB,EAAE,SAAiB;IAC7E,MAAM,cAAc,GAAG,MAAM,sBAAsB,EAAE,CAAC;IACtD,IAAI,CAAC,cAAc,EAAE,CAAC;QACpB,MAAM,IAAI,KAAK,CACb,sNAAsN,CACvN,CAAC;IACJ,CAAC;IAED,MAAM,cAAc,CAClB,cAAc,EACd,CAAC,YAAY,EAAE,aAAa,EAAE,cAAc,EAAE,KAAK,EAAE,UAAU,EAAE,SAAS,EAAE,QAAQ,CAAC,EACrF,MAAM,CAAC,oBAAoB;KAC5B,CAAC;IAEF,+DAA+D;IAC/D,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,QAAQ,EAAE,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,CAAC;IACjE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,GAAG,QAAQ,MAAM,CAAC,CAAC;IAExD,6BAA6B;IAC7B,IAAI,CAAC;QACH,MAAM,EAAE,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;QACzB,OAAO,OAAO,CAAC;IACjB,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,2DAA2D,CAAC,CAAC;IAC/E,CAAC;AACH,CAAC;AAED,iEAAiE;AACjE,MAAM,6BAA6B,GAAG,CAAC,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,CAAC,CAAC;AAErE;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,iBAAiB,CAAC,QAAgB,EAAE,SAAiB;IACzE,MAAM,WAAW,GAAG,MAAM,sBAAsB,EAAE,CAAC;IACnD,IAAI,CAAC,WAAW,EAAE,CAAC;QACjB,MAAM,IAAI,KAAK,CACb,mMAAmM,CACpM,CAAC;IACJ,CAAC;IAED,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IACjD,MAAM,gBAAgB,GAAG,6BAA6B,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;IAErE,qDAAqD;IACrD,IAAI,gBAAgB,EAAE,CAAC;QACrB,MAAM,cAAc,GAClB,CAAC,MAAM,kBAAkB,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,yBAAyB,CAAC,IAAI,CAAC,CAAC,CAAC;QAC9E,IAAI,CAAC,cAAc,EAAE,CAAC;YACpB,MAAM,IAAI,KAAK,CACb,sCAAsC,GAAG,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,+BAA+B;gBAC7F,mHAAmH,CACtH,CAAC;QACJ,CAAC;IACH,CAAC;IAED,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,QAAQ,EAAE,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,CAAC;IACjE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,GAAG,QAAQ,MAAM,CAAC,CAAC;IAExD,IAAI,CAAC;QACH,MAAM,cAAc,CAClB,WAAW,CAAC,OAAO,EACnB,CAAC,GAAG,WAAW,CAAC,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,KAAK,EAAE,QAAQ,EAAE,eAAe,EAAE,OAAO,CAAC,EACtF,KAAK,CACN,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,MAAM,QAAQ,GAAG,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;QACxE,kDAAkD;QAClD,IAAI,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,QAAQ,CAAC,QAAQ,CAAC,mBAAmB,CAAC,EAAE,CAAC;YACtE,MAAM,IAAI,KAAK,CACb,sCAAsC,GAAG,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,+BAA+B;gBAC7F,mHAAmH,EACrH;gBACE,KAAK,EAAE,KAAK;aACb,CACF,CAAC;QACJ,CAAC;QACD,IAAI,QAAQ,CAAC,QAAQ,CAAC,wBAAwB,CAAC,IAAI,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;YAC3E,MAAM,IAAI,KAAK,CACb,6BAA6B,GAAG,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,eAAe;gBACpE,oEAAoE,EACtE;gBACE,KAAK,EAAE,KAAK;aACb,CACF,CAAC;QACJ,CAAC;QACD,MAAM,KAAK,CAAC;IACd,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,YAAY,CAChC,QAAgB;IAEhB,IAAI,CAAC;QACH,uBAAuB;QACvB,IAAI,CAAC;YACH,MAAM,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC;QAC5B,CAAC;QAAC,MAAM,CAAC;YACP,OAAO;gBACL,OAAO,EAAE,mBAAmB,QAAQ,EAAE;gBACtC,IAAI,EAAE,gBAAgB;aACvB,CAAC;QACJ,CAAC;QAED,qBAAqB;QACrB,MAAM,SAAS,GAAG,MAAM,kBAAkB,CAAC,QAAQ,CAAC,CAAC;QAErD,+BAA+B;QAC/B,IAAI,SAAS,KAAK,MAAM,EAAE,CAAC;YACzB,OAAO;gBACL,OAAO,EAAE,QAAQ;gBACjB,iBAAiB,EAAE,SAAS;aAC7B,CAAC;QACJ,CAAC;QAED,mCAAmC;QACnC,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,MAAM,EAAE,EAAE,YAAY,CAAC,CAAC,CAAC;QAEtE,6BAA6B;QAC7B,IAAI,OAAe,CAAC;QAEpB,IAAI,gBAAgB,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;YACzC,OAAO,GAAG,MAAM,qBAAqB,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;QAC1D,CAAC;aAAM,IAAI,qBAAqB,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;YACrD,OAAO,GAAG,MAAM,qBAAqB,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;QAC1D,CAAC;aAAM,IAAI,eAAe,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;YAC/C,OAAO,GAAG,MAAM,iBAAiB,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;QACtD,CAAC;aAAM,CAAC;YACN,qBAAqB;YACrB,gEAAgE;YAChE,MAAM,OAAO,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;YACrD,OAAO;gBACL,OAAO;aACR,CAAC;QACJ,CAAC;QAED,OAAO;YACL,OAAO;YACP,iBAAiB,EAAE,SAAS;SAC7B,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,OAAO;YACL,OAAO,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC;YAC/D,IAAI,EAAE,kBAAkB;SACzB,CAAC;IACJ,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,sBAAsB,CAAC,OAAe;IAC1D,IAAI,CAAC;QACH,mCAAmC;QACnC,IAAI,OAAO,CAAC,QAAQ,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,EAAE,CAAC;YAClC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;YAClC,MAAM,EAAE,CAAC,EAAE,CAAC,GAAG,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QACrD,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,wBAAwB;IAC1B,CAAC;AACH,CAAC"}

package/dist/src/conversion/convertToPdf.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=convertToPdf.test.d.ts.map

package/dist/src/conversion/convertToPdf.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"convertToPdf.test.d.ts","sourceRoot":"","sources":["../../../src/conversion/convertToPdf.test.ts"],"names":[],"mappings":""}

package/dist/src/conversion/convertToPdf.test.js

@@ -0,0 +1,208 @@
+import { vi, describe, it, expect } from "vitest";
+import { EventEmitter } from "events";
+const mockFd = {
+    read: vi.fn(),
+    close: vi.fn(),
+};
+const mockProc = {
+    stdout: new EventEmitter(),
+    stderr: new EventEmitter(),
+    kill: vi.fn(),
+    on: vi.fn(),
+};
+vi.mock("child_process", () => ({
+    spawn: vi.fn(() => mockProc),
+}));
+vi.mock("fs", async () => {
+    const actual = await vi.importActual("fs");
+    return {
+        ...actual,
+        promises: {
+            open: vi.fn(async () => {
+                return mockFd;
+            }),
+            access: vi.fn(async (path, _mode) => {
+                console.log(path);
+                const toErrorPath = [
+                    "/Applications/LibreOffice.app/Contents/MacOS/soffice",
+                    "/Applications/LibreOffice.app/Contents/MacOS/libreoffice",
+                    "C:\\Program Files\\Libreoffice\\program\\soffice.exe",
+                    "./test_fail.pdf",
+                    "test_fail.pdf",
+                    "test.docx",
+                ];
+                if (toErrorPath.includes(path)) {
+                    throw new Error("unaccessible");
+                }
+                return;
+            }),
+            mkdtemp: vi.fn(async () => {
+                return "/tmp/test";
+            }),
+            readFile: vi.fn(async () => {
+                return "hello world";
+            }),
+        },
+    };
+});
+import { guessFileExtension, findImageMagickCommand, findLibreOfficeCommand, convertOfficeDocument, convertImageToPdf, convertToPdf, } from "./convertToPdf";
+describe("test guessFileExtension", () => {
+    it("detects PDF", async () => {
+        mockFd.read.mockImplementation((buffer) => {
+            Buffer.from("%PDF").copy(buffer);
+        });
+        expect(await guessFileExtension("/some/file")).toBe(".pdf");
+    });
+    it("detects PNG", async () => {
+        mockFd.read.mockImplementation((buffer) => {
+            Buffer.from([0x89, 0x50, 0x4e, 0x47]).copy(buffer);
+        });
+        expect(await guessFileExtension("/some/file")).toBe(".png");
+    });
+    it("returns extension directly if present", async () => {
+        expect(await guessFileExtension("/some/file.pdf")).toBe(".pdf");
+    });
+});
+describe("test command availability", () => {
+    it("libreoffice available", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "/opt/bin/libreoffice");
+        const result = await findLibreOfficeCommand();
+        expect(result).toBe("libreoffice");
+    });
+    it("libreoffice not available", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(1);
+        });
+        mockProc.stderr.emit("data", "command not found");
+        // does not throw
+        const result = await findLibreOfficeCommand();
+        expect(result).toBeNull();
+    });
+    it("imagemagick available", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "/opt/bin/libreoffice");
+        const result = await findImageMagickCommand();
+        expect(result).toStrictEqual({ command: "magick", args: [] });
+    });
+    it("imagemagick not available", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(1);
+        });
+        mockProc.stderr.emit("data", "command not found");
+        // does not throw
+        const result = await findImageMagickCommand();
+        expect(result).toBeNull();
+    });
+});
+describe("test convertOfficeDocument", () => {
+    it("conversion succeeds", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertOfficeDocument("test.docx", "./");
+        expect(result).toBe("test.pdf");
+    });
+    it("conversion fails (command not found)", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(1);
+        });
+        mockProc.stdout.emit("data", "command not found");
+        await expect(convertOfficeDocument("test_command.docx", "./")).rejects.toThrow("LibreOffice is not installed. Please install LibreOffice to convert office documents. On macOS: brew install --cask libreoffice, On Ubuntu: apt-get install libreoffice");
+    });
+    it("conversion fails (output not found)", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        await expect(convertOfficeDocument("test_fail.docx", "./")).rejects.toThrow("LibreOffice conversion succeeded but output PDF not found");
+    });
+});
+describe("test convertImageToPdf", () => {
+    it("conversion succeeds", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertImageToPdf("test.png", "./");
+        expect(result).toBe("test.pdf");
+    });
+    it("conversion fails (command not found)", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(1);
+        });
+        mockProc.stdout.emit("data", "command not found");
+        await expect(convertImageToPdf("test_command.png", "./")).rejects.toThrow("ImageMagick is not installed. Please install ImageMagick to convert images. On macOS: brew install imagemagick, On Ubuntu: apt-get install imagemagick");
+    });
+});
+describe("test convertToPdf", () => {
+    it("convert PDF fails because file not found", async () => {
+        const result = await convertToPdf("test.docx");
+        expect(result).toStrictEqual({
+            message: `File not found: test.docx`,
+            code: "FILE_NOT_FOUND",
+        });
+    });
+    it("convert an office document (word)", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertToPdf("test_1.docx");
+        expect(result).toStrictEqual({
+            pdfPath: "/tmp/test/test_1.pdf",
+            originalExtension: ".docx",
+        });
+    });
+    it("convert an office document (xlsx)", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertToPdf("test.xlsx");
+        expect(result).toStrictEqual({
+            pdfPath: "/tmp/test/test.pdf",
+            originalExtension: ".xlsx",
+        });
+    });
+    it("convert an image", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertToPdf("test.png");
+        expect(result).toStrictEqual({
+            pdfPath: "/tmp/test/test.pdf",
+            originalExtension: ".png",
+        });
+    });
+    it("convert a text file", async () => {
+        mockProc.on.mockImplementation((event, cb) => {
+            if (event === "close")
+                cb(0);
+        });
+        mockProc.stdout.emit("data", "conversion successfull");
+        const result = await convertToPdf("test.txt");
+        expect(result).toStrictEqual({
+            content: "hello world",
+        });
+    });
+});
+//# sourceMappingURL=convertToPdf.test.js.map

package/dist/src/conversion/convertToPdf.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"convertToPdf.test.js","sourceRoot":"","sources":["../../../src/conversion/convertToPdf.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,EAAE,EAAE,QAAQ,EAAE,EAAE,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAClD,OAAO,EAAE,YAAY,EAAE,MAAM,QAAQ,CAAC;AAEtC,MAAM,MAAM,GAAG;IACb,IAAI,EAAE,EAAE,CAAC,EAAE,EAAE;IACb,KAAK,EAAE,EAAE,CAAC,EAAE,EAAE;CACf,CAAC;AAEF,MAAM,QAAQ,GAAG;IACf,MAAM,EAAE,IAAI,YAAY,EAAE;IAC1B,MAAM,EAAE,IAAI,YAAY,EAAE;IAC1B,IAAI,EAAE,EAAE,CAAC,EAAE,EAAE;IACb,EAAE,EAAE,EAAE,CAAC,EAAE,EAAE;CACZ,CAAC;AAEF,EAAE,CAAC,IAAI,CAAC,eAAe,EAAE,GAAG,EAAE,CAAC,CAAC;IAC9B,KAAK,EAAE,EAAE,CAAC,EAAE,CAAC,GAAG,EAAE,CAAC,QAAQ,CAAC;CAC7B,CAAC,CAAC,CAAC;AAEJ,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,KAAK,IAAI,EAAE;IACvB,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,YAAY,CAAsB,IAAI,CAAC,CAAC;IAChE,OAAO;QACL,GAAG,MAAM;QACT,QAAQ,EAAE;YACR,IAAI,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,IAAI,EAAE;gBACrB,OAAO,MAAM,CAAC;YAChB,CAAC,CAAC;YACF,MAAM,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,EAAE,IAAY,EAAE,KAAc,EAAE,EAAE;gBACnD,OAAO,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;gBAClB,MAAM,WAAW,GAAG;oBAClB,sDAAsD;oBACtD,0DAA0D;oBAC1D,sDAAsD;oBACtD,iBAAiB;oBACjB,eAAe;oBACf,WAAW;iBACZ,CAAC;gBACF,IAAI,WAAW,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;oBAC/B,MAAM,IAAI,KAAK,CAAC,cAAc,CAAC,CAAC;gBAClC,CAAC;gBACD,OAAO;YACT,CAAC,CAAC;YACF,OAAO,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,IAAI,EAAE;gBACxB,OAAO,WAAW,CAAC;YACrB,CAAC,CAAC;YACF,QAAQ,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,IAAI,EAAE;gBACzB,OAAO,aAAa,CAAC;YACvB,CAAC,CAAC;SACH;KACF,CAAC;AACJ,CAAC,CAAC,CAAC;AAEH,OAAO,EACL,kBAAkB,EAClB,sBAAsB,EACtB,sBAAsB,EACtB,qBAAqB,EACrB,iBAAiB,EACjB,YAAY,GACb,MAAM,gBAAgB,CAAC;AAExB,QAAQ,CAAC,yBAAyB,EAAE,GAAG,EAAE;IACvC,EAAE,CAAC,aAAa,EAAE,KAAK,IAAI,EAAE;QAC3B,MAAM,CAAC,IAAI,CAAC,kBAAkB,CAAC,CAAC,MAAc,EAAE,EAAE;YAChD,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACnC,CAAC,CAAC,CAAC;QAEH,MAAM,CAAC,MAAM,kBAAkB,CAAC,YAAY,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IAC9D,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,aAAa,EAAE,KAAK,IAAI,EAAE;QAC3B,MAAM,CAAC,IAAI,CAAC,kBAAkB,CAAC,CAAC,MAAc,EAAE,EAAE;YAChD,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACrD,CAAC,CAAC,CAAC;QAEH,MAAM,CAAC,MAAM,kBAAkB,CAAC,YAAY,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IAC9D,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uCAAuC,EAAE,KAAK,IAAI,EAAE;QACrD,MAAM,CAAC,MAAM,kBAAkB,CAAC,gBAAgB,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IAClE,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,2BAA2B,EAAE,GAAG,EAAE;IACzC,EAAE,CAAC,uBAAuB,EAAE,KAAK,IAAI,EAAE;QACrC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,sBAAsB,CAAC,CAAC;QACrD,MAAM,MAAM,GAAG,MAAM,sBAAsB,EAAE,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC;IACrC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,2BAA2B,EAAE,KAAK,IAAI,EAAE;QACzC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,mBAAmB,CAAC,CAAC;QAClD,iBAAiB;QACjB,MAAM,MAAM,GAAG,MAAM,sBAAsB,EAAE,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,EAAE,CAAC;IAC5B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uBAAuB,EAAE,KAAK,IAAI,EAAE;QACrC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,sBAAsB,CAAC,CAAC;QACrD,MAAM,MAAM,GAAG,MAAM,sBAAsB,EAAE,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC,EAAE,OAAO,EAAE,QAAQ,EAAE,IAAI,EAAE,EAAE,EAAE,CAAC,CAAC;IAChE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,2BAA2B,EAAE,KAAK,IAAI,EAAE;QACzC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,mBAAmB,CAAC,CAAC;QAClD,iBAAiB;QACjB,MAAM,MAAM,GAAG,MAAM,sBAAsB,EAAE,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,EAAE,CAAC;IAC5B,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,4BAA4B,EAAE,GAAG,EAAE;IAC1C,EAAE,CAAC,qBAAqB,EAAE,KAAK,IAAI,EAAE;QACnC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QAEvD,MAAM,MAAM,GAAG,MAAM,qBAAqB,CAAC,WAAW,EAAE,IAAI,CAAC,CAAC;QAC9D,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;IAClC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACpD,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,mBAAmB,CAAC,CAAC;QAElD,MAAM,MAAM,CAAC,qBAAqB,CAAC,mBAAmB,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAC5E,yKAAyK,CAC1K,CAAC;IACJ,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,qCAAqC,EAAE,KAAK,IAAI,EAAE;QACnD,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QAEvD,MAAM,MAAM,CAAC,qBAAqB,CAAC,gBAAgB,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CACzE,2DAA2D,CAC5D,CAAC;IACJ,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,wBAAwB,EAAE,GAAG,EAAE;IACtC,EAAE,CAAC,qBAAqB,EAAE,KAAK,IAAI,EAAE;QACnC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QAEvD,MAAM,MAAM,GAAG,MAAM,iBAAiB,CAAC,UAAU,EAAE,IAAI,CAAC,CAAC;QACzD,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;IAClC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACpD,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,mBAAmB,CAAC,CAAC;QAElD,MAAM,MAAM,CAAC,iBAAiB,CAAC,kBAAkB,EAAE,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CACvE,wJAAwJ,CACzJ,CAAC;IACJ,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,mBAAmB,EAAE,GAAG,EAAE;IACjC,EAAE,CAAC,0CAA0C,EAAE,KAAK,IAAI,EAAE;QACxD,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,WAAW,CAAC,CAAC;QAC/C,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC;YAC3B,OAAO,EAAE,2BAA2B;YACpC,IAAI,EAAE,gBAAgB;SACvB,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,mCAAmC,EAAE,KAAK,IAAI,EAAE;QACjD,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QACvD,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,aAAa,CAAC,CAAC;QACjD,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC;YAC3B,OAAO,EAAE,sBAAsB;YAC/B,iBAAiB,EAAE,OAAO;SAC3B,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,mCAAmC,EAAE,KAAK,IAAI,EAAE;QACjD,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QACvD,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,WAAW,CAAC,CAAC;QAC/C,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC;YAC3B,OAAO,EAAE,oBAAoB;YAC7B,iBAAiB,EAAE,OAAO;SAC3B,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,kBAAkB,EAAE,KAAK,IAAI,EAAE;QAChC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QACvD,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,UAAU,CAAC,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC;YAC3B,OAAO,EAAE,oBAAoB;YAC7B,iBAAiB,EAAE,MAAM;SAC1B,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,qBAAqB,EAAE,KAAK,IAAI,EAAE;QACnC,QAAQ,CAAC,EAAE,CAAC,kBAAkB,CAAC,CAAC,KAAK,EAAE,EAAE,EAAE,EAAE;YAC3C,IAAI,KAAK,KAAK,OAAO;gBAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;QACH,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,wBAAwB,CAAC,CAAC;QACvD,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,UAAU,CAAC,CAAC;QAC9C,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC;YAC3B,OAAO,EAAE,aAAa;SACvB,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/dist/src/core/config.d.ts

@@ -0,0 +1,4 @@
+import { LiteParseConfig } from "./types.js";
+export declare const DEFAULT_CONFIG: LiteParseConfig;
+export declare function mergeConfig(userConfig: Partial<LiteParseConfig>): LiteParseConfig;
+//# sourceMappingURL=config.d.ts.map

package/dist/src/core/config.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"config.d.ts","sourceRoot":"","sources":["../../../src/core/config.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,MAAM,YAAY,CAAC;AAE7C,eAAO,MAAM,cAAc,EAAE,eAoB5B,CAAC;AAEF,wBAAgB,WAAW,CAAC,UAAU,EAAE,OAAO,CAAC,eAAe,CAAC,GAAG,eAAe,CAKjF"}

package/dist/src/core/config.js

@@ -0,0 +1,25 @@
+export const DEFAULT_CONFIG = {
+    // OCR - defaults to in-process Tesseract for zero-setup experience
+    // If ocrServerUrl is provided, uses HTTP OCR instead
+    ocrLanguage: "en",
+    ocrEnabled: true,
+    ocrServerUrl: undefined, // If set, uses HTTP OCR; otherwise uses Tesseract
+    numWorkers: 4, // Number of pages to OCR in parallel
+    // Processing
+    maxPages: 1000,
+    targetPages: undefined,
+    dpi: 150,
+    // Output
+    outputFormat: "json",
+    // Features
+    preciseBoundingBox: true,
+    preserveVerySmallText: false,
+    preserveLayoutAlignmentAcrossPages: false,
+};
+export function mergeConfig(userConfig) {
+    return {
+        ...DEFAULT_CONFIG,
+        ...userConfig,
+    };
+}
+//# sourceMappingURL=config.js.map

package/dist/src/core/config.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"config.js","sourceRoot":"","sources":["../../../src/core/config.ts"],"names":[],"mappings":"AAEA,MAAM,CAAC,MAAM,cAAc,GAAoB;IAC7C,mEAAmE;IACnE,qDAAqD;IACrD,WAAW,EAAE,IAAI;IACjB,UAAU,EAAE,IAAI;IAChB,YAAY,EAAE,SAAS,EAAE,kDAAkD;IAC3E,UAAU,EAAE,CAAC,EAAE,qCAAqC;IAEpD,aAAa;IACb,QAAQ,EAAE,IAAI;IACd,WAAW,EAAE,SAAS;IACtB,GAAG,EAAE,GAAG;IAER,SAAS;IACT,YAAY,EAAE,MAAM;IAEpB,WAAW;IACX,kBAAkB,EAAE,IAAI;IACxB,qBAAqB,EAAE,KAAK;IAC5B,kCAAkC,EAAE,KAAK;CAC1C,CAAC;AAEF,MAAM,UAAU,WAAW,CAAC,UAAoC;IAC9D,OAAO;QACL,GAAG,cAAc;QACjB,GAAG,UAAU;KACd,CAAC;AACJ,CAAC"}

package/dist/src/core/config.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=config.test.d.ts.map

package/dist/src/core/config.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"config.test.d.ts","sourceRoot":"","sources":["../../../src/core/config.test.ts"],"names":[],"mappings":""}

package/dist/src/core/config.test.js

@@ -0,0 +1,21 @@
+import { describe, it, expect } from "vitest";
+import { DEFAULT_CONFIG, mergeConfig } from "./config";
+describe("test config", () => {
+    it("test merge config with no partial config", () => {
+        const user_config = {};
+        const config = mergeConfig(user_config);
+        expect(config).toStrictEqual(DEFAULT_CONFIG);
+    });
+    it("test merge config overrides defaults", () => {
+        const user_config = {
+            ocrLanguage: "fr",
+            ocrServerUrl: "http://localhost:8000",
+            dpi: 200,
+        };
+        const config = mergeConfig(user_config);
+        expect(config.ocrLanguage).toBe(user_config.ocrLanguage);
+        expect(config.ocrServerUrl).toBe(user_config.ocrServerUrl);
+        expect(config.dpi).toBe(user_config.dpi);
+    });
+});
+//# sourceMappingURL=config.test.js.map

package/dist/src/core/config.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"config.test.js","sourceRoot":"","sources":["../../../src/core/config.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAE9C,OAAO,EAAE,cAAc,EAAE,WAAW,EAAE,MAAM,UAAU,CAAC;AAEvD,QAAQ,CAAC,aAAa,EAAE,GAAG,EAAE;IAC3B,EAAE,CAAC,0CAA0C,EAAE,GAAG,EAAE;QAClD,MAAM,WAAW,GAA6B,EAAE,CAAC;QACjD,MAAM,MAAM,GAAG,WAAW,CAAC,WAAW,CAAC,CAAC;QACxC,MAAM,CAAC,MAAM,CAAC,CAAC,aAAa,CAAC,cAAc,CAAC,CAAC;IAC/C,CAAC,CAAC,CAAC;IACH,EAAE,CAAC,sCAAsC,EAAE,GAAG,EAAE;QAC9C,MAAM,WAAW,GAA6B;YAC5C,WAAW,EAAE,IAAI;YACjB,YAAY,EAAE,uBAAuB;YACrC,GAAG,EAAE,GAAG;SACT,CAAC;QACF,MAAM,MAAM,GAAG,WAAW,CAAC,WAAW,CAAC,CAAC;QACxC,MAAM,CAAC,MAAM,CAAC,WAAW,CAAC,CAAC,IAAI,CAAC,WAAW,CAAC,WAAW,CAAC,CAAC;QACzD,MAAM,CAAC,MAAM,CAAC,YAAY,CAAC,CAAC,IAAI,CAAC,WAAW,CAAC,YAAY,CAAC,CAAC;QAC3D,MAAM,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IAC3C,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}