@aiquants/html-to-markdown 0.2.5 → 0.3.0

package/README.md

@@ -136,12 +136,15 @@ npx --package=@aiquants/html-to-markdown aiq-html2md-mcp-stream
   - `url` (required*): The URL of the web page to convert
   - `html_content` (required*): HTML content as a string to convert (alternative to URL)
   - `locale` (optional): Browser locale (`en-US` or `ja-JP`, defaults to `en-US`)
-  - `output_format` (optional): Output format (`markdown`, `html`, or `both`, defaults to `markdown`)
-  - `save_path` (optional): Full path (including filename) where to save the converted content
-  - `save_directory` (optional): Directory path where to save the converted content with auto-generated filename
 
   *Either `url` or `html_content` is required
 
+- **save_content_to_file**: Save text content to a file with specified path or directory
+  - `content` (required): Text content to save to file (will be saved as Markdown format)
+  - `save_path` (optional): Complete file path including filename to save content
+  - `save_directory` (optional): Directory path to save content with auto-generated filename
+  - `filename` (optional): Base filename to use when save_directory is specified (extension .md will be added automatically)
+
 **File Saving Options:**
 
 - **`save_path`**: Specify the complete file path including filename and extension where you want to save the content.
@@ -157,11 +160,10 @@ npx --package=@aiquants/html-to-markdown aiq-html2md-mcp-stream
 
 **File Saving Behavior:**
 
-- When `output_format` is `markdown`: Only a `.md` file is saved
-- When `output_format` is `html`: Only a `.html` file is saved
-- When `output_format` is `both`: Both `.md` and `.html` files are saved
-  - With `save_path`: Files are saved as `path.md` and `path.html`
-  - With `save_directory`: Files are saved as `filename.md` and `filename.html`
+- All content is saved as Markdown format (`.md` files)
+- The directory will be created automatically if it doesn't exist
+- With `save_path`: File is saved to the exact specified path
+- With `save_directory`: File is saved with auto-generated filename based on the URL or custom filename if provided
 
 - **html_to_markdown_streamable**: Same as above but with streamable MCP support for progress updates
 

package/dist/{core-CFnkfU5N.js → core-CouDTni-.js}

@@ -61,12 +61,12 @@ const getMessage = (locale, key2, replacements) => {
 const getHtmlWithPlaywright = async (url, locale) => {
   let browser;
   try {
-    console.error(getMessage(locale, "playwright_launch"));
+    console.info(getMessage(locale, "playwright_launch"));
     browser = await chromium.launch();
     const page = await browser.newPage({ locale });
-    console.error(getMessage(locale, "playwright_goto", { url, locale }));
+    console.info(getMessage(locale, "playwright_goto", { url, locale }));
     await page.goto(url, { waitUntil: "networkidle" });
-    console.error(getMessage(locale, "playwright_get_content"));
+    console.info(getMessage(locale, "playwright_get_content"));
     const html2 = await page.content();
     return html2;
   } catch (error) {
@@ -75,7 +75,7 @@ const getHtmlWithPlaywright = async (url, locale) => {
   } finally {
     if (browser) {
       await browser.close();
-      console.error(getMessage(locale, "playwright_close"));
+      console.info(getMessage(locale, "playwright_close"));
     }
   }
 };
@@ -20681,7 +20681,7 @@ const handleTableCell = (_state, node2, _parent) => {
   return { type: "tableCell", children: [{ type: "html", value }] };
 };
 const convertHtmlToMarkdown = async (html2, baseUrl, locale) => {
-  console.error(getMessage(locale, "convert_start"));
+  console.info(getMessage(locale, "convert_start"));
   const slugger2 = new BananaSlug();
   const file = await unified().use(rehypeParse, { fragment: true }).use(rehypeRaw).use(rehypeSanitizeHtml).use(rehypeWikipediaFootnotes).use(rehypeNamedAnchors).use(rehypeSlug).use(rehypeParagraphWrapper).use(rehypeAbsoluteLinks, { baseUrl }).use(rehypeRemark, {
     handlers: {
@@ -20783,7 +20783,7 @@ const convertHtmlToMarkdown = async (html2, baseUrl, locale) => {
   }).process(html2);
   const unescapedMarkdown = String(file).replace(/\\\[\^(.+?)\\?\](?!\s*:)/g, "[^$1]").replace(/\]\(\\#/g, "](#");
   const correctedMarkdown = unescapedMarkdown.replace(/(\s*)- - /g, "$1- ").replace(/(\s*)\* \* /g, "$1* ");
-  console.error(getMessage(locale, "convert_success"));
+  console.info(getMessage(locale, "convert_success"));
   return correctedMarkdown;
 };
 const htmlToMarkdown = async (urlOrHtml, options = {}) => {

package/dist/index.cjs

@@ -1 +1 @@
-const e=new(require("happy-dom").Window);Object.assign(global,{document:e.document,window:e,self:e}),Object.defineProperty(exports,Symbol.toStringTag,{value:"Module"});const r=require("./core-D2dlEt1l.cjs"),t=require("./mcp-server-Bo_mCvzz.cjs"),c=require("./mcp-server-streamable-Cc4QsZQe.cjs");exports.getMessage=r.getMessage,exports.htmlToMarkdown=r.htmlToMarkdown,exports.createMcpServer=t.createMcpServer,exports.runMcpServer=t.runMcpServer,exports.createStreamableMcpServer=c.createStreamableMcpServer,exports.runStreamableMcpServer=c.runStreamableMcpServer;
+const e=new(require("happy-dom").Window);Object.assign(global,{document:e.document,window:e,self:e}),Object.defineProperty(exports,Symbol.toStringTag,{value:"Module"});const r=require("./core-D2dlEt1l.cjs"),t=require("./mcp-server-CCu0Pcm3.cjs"),c=require("./mcp-server-streamable-mF26pg3E.cjs");exports.getMessage=r.getMessage,exports.htmlToMarkdown=r.htmlToMarkdown,exports.createMcpServer=t.createMcpServer,exports.runMcpServer=t.runMcpServer,exports.createStreamableMcpServer=c.createStreamableMcpServer,exports.runStreamableMcpServer=c.runStreamableMcpServer;

package/dist/index.js

@@ -5,9 +5,9 @@ Object.assign(global, {
   window,
   self: window
 });
-import { g, h } from "./core-CFnkfU5N.js";
-import { c, r } from "./mcp-server-CueTvF4M.js";
-import { c as c2, r as r2 } from "./mcp-server-streamable-BC-Qlr-9.js";
+import { g, h } from "./core-CouDTni-.js";
+import { c, r } from "./mcp-server-BAmy7d8r.js";
+import { c as c2, r as r2 } from "./mcp-server-streamable-CVDJHcWu.js";
 export {
   c as createMcpServer,
   c2 as createStreamableMcpServer,

package/dist/main.js

@@ -6,7 +6,7 @@ Object.assign(global, {
   window,
   self: window
 });
-import { g as getMessage, h as htmlToMarkdown } from "./core-CFnkfU5N.js";
+import { g as getMessage, h as htmlToMarkdown } from "./core-CouDTni-.js";
 const runCli = async () => {
   const yargsParser = (await import("yargs-parser")).default;
   const fs = await import("node:fs");

package/dist/mcp-server-BAmy7d8r.js

@@ -0,0 +1,210 @@
+import { Window } from "happy-dom";
+const window = new Window();
+Object.assign(global, {
+  document: window.document,
+  window,
+  self: window
+});
+import { J as JSONRPCMessageSchema, g as getPackageVersion, S as Server, L as ListToolsRequestSchema, M as MCP_TOOL_SCHEMAS, C as CallToolRequestSchema, p as parseHtmlToMarkdownArgs, c as createSuccessResult, a as createErrorResult, b as parseSaveContentArgs, s as saveContentToFile } from "./version-BF9-U8Ef.js";
+import process from "node:process";
+import { h as htmlToMarkdown } from "./core-CouDTni-.js";
+class ReadBuffer {
+  append(chunk) {
+    this._buffer = this._buffer ? Buffer.concat([this._buffer, chunk]) : chunk;
+  }
+  readMessage() {
+    if (!this._buffer) {
+      return null;
+    }
+    const index = this._buffer.indexOf("\n");
+    if (index === -1) {
+      return null;
+    }
+    const line = this._buffer.toString("utf8", 0, index).replace(/\r$/, "");
+    this._buffer = this._buffer.subarray(index + 1);
+    return deserializeMessage(line);
+  }
+  clear() {
+    this._buffer = void 0;
+  }
+}
+function deserializeMessage(line) {
+  return JSONRPCMessageSchema.parse(JSON.parse(line));
+}
+function serializeMessage(message) {
+  return JSON.stringify(message) + "\n";
+}
+class StdioServerTransport {
+  constructor(_stdin = process.stdin, _stdout = process.stdout) {
+    this._stdin = _stdin;
+    this._stdout = _stdout;
+    this._readBuffer = new ReadBuffer();
+    this._started = false;
+    this._ondata = (chunk) => {
+      this._readBuffer.append(chunk);
+      this.processReadBuffer();
+    };
+    this._onerror = (error) => {
+      var _a;
+      (_a = this.onerror) === null || _a === void 0 ? void 0 : _a.call(this, error);
+    };
+  }
+  /**
+   * Starts listening for messages on stdin.
+   */
+  async start() {
+    if (this._started) {
+      throw new Error("StdioServerTransport already started! If using Server class, note that connect() calls start() automatically.");
+    }
+    this._started = true;
+    this._stdin.on("data", this._ondata);
+    this._stdin.on("error", this._onerror);
+  }
+  processReadBuffer() {
+    var _a, _b;
+    while (true) {
+      try {
+        const message = this._readBuffer.readMessage();
+        if (message === null) {
+          break;
+        }
+        (_a = this.onmessage) === null || _a === void 0 ? void 0 : _a.call(this, message);
+      } catch (error) {
+        (_b = this.onerror) === null || _b === void 0 ? void 0 : _b.call(this, error);
+      }
+    }
+  }
+  async close() {
+    var _a;
+    this._stdin.off("data", this._ondata);
+    this._stdin.off("error", this._onerror);
+    const remainingDataListeners = this._stdin.listenerCount("data");
+    if (remainingDataListeners === 0) {
+      this._stdin.pause();
+    }
+    this._readBuffer.clear();
+    (_a = this.onclose) === null || _a === void 0 ? void 0 : _a.call(this);
+  }
+  send(message) {
+    return new Promise((resolve) => {
+      const json = serializeMessage(message);
+      if (this._stdout.write(json)) {
+        resolve();
+      } else {
+        this._stdout.once("drain", resolve);
+      }
+    });
+  }
+}
+class HtmlToMarkdownMcpServer {
+  constructor() {
+    this.version = getPackageVersion();
+    this.server = new Server(
+      {
+        name: "@aiquants/html-to-markdown",
+        version: this.version,
+        description: "Convert web pages and HTML to Markdown with JavaScript dynamic content support and table structure preservation. Includes separate file saving tool."
+      },
+      {
+        capabilities: {
+          tools: {}
+        }
+      }
+    );
+    this.setupToolHandlers();
+  }
+  /**
+   * Setup tool handlers for the MCP server.
+   * MCP サーバーのツールハンドラーを設定
+   */
+  setupToolHandlers() {
+    this.server.setRequestHandler(ListToolsRequestSchema, async () => {
+      return {
+        tools: [
+          {
+            name: "html_to_markdown",
+            description: "Convert web pages or HTML strings to Markdown format. Handles JavaScript-rendered dynamic content and preserves table structures.",
+            inputSchema: MCP_TOOL_SCHEMAS.htmlToMarkdown
+          },
+          {
+            name: "save_content_to_file",
+            description: "Save text content to a file with specified path or directory. Supports automatic filename generation and format detection.",
+            inputSchema: MCP_TOOL_SCHEMAS.saveContentToFile
+          }
+        ]
+      };
+    });
+    this.server.setRequestHandler(CallToolRequestSchema, async (request) => {
+      switch (request.params.name) {
+        case "html_to_markdown": {
+          return await this.handleHtmlToMarkdown(request.params.arguments);
+        }
+        case "save_content_to_file": {
+          return await this.handleSaveContentToFile(request.params.arguments);
+        }
+        default:
+          throw new Error(`Unknown tool: ${request.params.name}`);
+      }
+    });
+  }
+  /**
+   * Handle HTML to Markdown conversion tool call.
+   * HTML から Markdown への変換ツール呼び出しを処理
+   */
+  async handleHtmlToMarkdown(args) {
+    try {
+      const validatedArgs = parseHtmlToMarkdownArgs(args);
+      const { url, html_content, locale = "en-US" } = validatedArgs;
+      if (url) {
+        try {
+          new URL(url);
+        } catch {
+          throw new Error(`Invalid URL: ${url}`);
+        }
+      }
+      const result = await htmlToMarkdown(url || "direct-html-input", {
+        locale,
+        htmlContent: html_content
+      });
+      const responseText = result.markdown;
+      return createSuccessResult(responseText);
+    } catch (error) {
+      return createErrorResult(error);
+    }
+  }
+  /**
+   * Handle save content to file tool call.
+   * ファイル保存ツール呼び出しを処理
+   */
+  async handleSaveContentToFile(args) {
+    try {
+      const validatedArgs = parseSaveContentArgs(args);
+      const { content, save_path, save_directory, filename } = validatedArgs;
+      const savedPath = await saveContentToFile(content, save_path, save_directory, filename);
+      const responseText = `File saved successfully to: ${savedPath}`;
+      return createSuccessResult(responseText);
+    } catch (error) {
+      return createErrorResult(error);
+    }
+  }
+  /**
+   * Start the MCP server.
+   * MCP サーバーを開始
+   */
+  async start() {
+    const transport = new StdioServerTransport();
+    await this.server.connect(transport);
+    console.info("HTML to Markdown MCP server started");
+  }
+}
+const createMcpServer = () => {
+  return new HtmlToMarkdownMcpServer();
+};
+const runMcpServer = async () => {
+  const server = createMcpServer();
+  await server.start();
+};
+export {
+  createMcpServer as c,
+  runMcpServer as r
+};

package/dist/mcp-server-CCu0Pcm3.cjs

@@ -0,0 +1 @@
+const t=new(require("happy-dom").Window);Object.assign(global,{document:t.document,window:t,self:t});const e=require("./version-Cg6D6H9g.cjs"),r=require("node:process"),s=require("./core-D2dlEt1l.cjs");class a{append(t){this._buffer=this._buffer?Buffer.concat([this._buffer,t]):t}readMessage(){if(!this._buffer)return null;const t=this._buffer.indexOf("\n");if(-1===t)return null;const r=this._buffer.toString("utf8",0,t).replace(/\r$/,"");return this._buffer=this._buffer.subarray(t+1),function(t){return e.JSONRPCMessageSchema.parse(JSON.parse(t))}(r)}clear(){this._buffer=void 0}}class n{constructor(t=r.stdin,e=r.stdout){this._stdin=t,this._stdout=e,this._readBuffer=new a,this._started=!1,this._ondata=t=>{this._readBuffer.append(t),this.processReadBuffer()},this._onerror=t=>{var e;null===(e=this.onerror)||void 0===e||e.call(this,t)}}async start(){if(this._started)throw new Error("StdioServerTransport already started! If using Server class, note that connect() calls start() automatically.");this._started=!0,this._stdin.on("data",this._ondata),this._stdin.on("error",this._onerror)}processReadBuffer(){for(var t,e;;)try{const e=this._readBuffer.readMessage();if(null===e)break;null===(t=this.onmessage)||void 0===t||t.call(this,e)}catch(r){null===(e=this.onerror)||void 0===e||e.call(this,r)}}async close(){var t;this._stdin.off("data",this._ondata),this._stdin.off("error",this._onerror);0===this._stdin.listenerCount("data")&&this._stdin.pause(),this._readBuffer.clear(),null===(t=this.onclose)||void 0===t||t.call(this)}send(t){return new Promise(e=>{const r=function(t){return JSON.stringify(t)+"\n"}(t);this._stdout.write(r)?e():this._stdout.once("drain",e)})}}class o{constructor(){this.version=e.getPackageVersion(),this.server=new e.Server({name:"@aiquants/html-to-markdown",version:this.version,description:"Convert web pages and HTML to Markdown with JavaScript dynamic content support and table structure preservation. Includes separate file saving tool."},{capabilities:{tools:{}}}),this.setupToolHandlers()}setupToolHandlers(){this.server.setRequestHandler(e.ListToolsRequestSchema,async()=>({tools:[{name:"html_to_markdown",description:"Convert web pages or HTML strings to Markdown format. Handles JavaScript-rendered dynamic content and preserves table structures.",inputSchema:e.MCP_TOOL_SCHEMAS.htmlToMarkdown},{name:"save_content_to_file",description:"Save text content to a file with specified path or directory. Supports automatic filename generation and format detection.",inputSchema:e.MCP_TOOL_SCHEMAS.saveContentToFile}]})),this.server.setRequestHandler(e.CallToolRequestSchema,async t=>{switch(t.params.name){case"html_to_markdown":return await this.handleHtmlToMarkdown(t.params.arguments);case"save_content_to_file":return await this.handleSaveContentToFile(t.params.arguments);default:throw new Error(`Unknown tool: ${t.params.name}`)}})}async handleHtmlToMarkdown(t){try{const r=e.parseHtmlToMarkdownArgs(t),{url:a,html_content:n,locale:o="en-US"}=r;if(a)try{new URL(a)}catch{throw new Error(`Invalid URL: ${a}`)}const i=(await s.htmlToMarkdown(a||"direct-html-input",{locale:o,htmlContent:n})).markdown;return e.createSuccessResult(i)}catch(r){return e.createErrorResult(r)}}async handleSaveContentToFile(t){try{const r=e.parseSaveContentArgs(t),{content:s,save_path:a,save_directory:n,filename:o}=r,i=`File saved successfully to: ${await e.saveContentToFile(s,a,n,o)}`;return e.createSuccessResult(i)}catch(r){return e.createErrorResult(r)}}async start(){const t=new n;await this.server.connect(t)}}const i=()=>new o;exports.createMcpServer=i,exports.runMcpServer=async()=>{const t=i();await t.start()};