@aiquants/html-to-markdown 0.1.4 → 0.1.6

package/dist/{core-DOzMLPij.js → core-DU_246Q7.js}

@@ -19494,9 +19494,39 @@ function bail(error) {
     throw error;
   }
 }
+var commonjsGlobal = typeof globalThis !== "undefined" ? globalThis : typeof window !== "undefined" ? window : typeof global !== "undefined" ? global : typeof self !== "undefined" ? self : {};
 function getDefaultExportFromCjs(x) {
   return x && x.__esModule && Object.prototype.hasOwnProperty.call(x, "default") ? x["default"] : x;
 }
+function getAugmentedNamespace(n) {
+  if (Object.prototype.hasOwnProperty.call(n, "__esModule")) return n;
+  var f = n.default;
+  if (typeof f == "function") {
+    var a2 = function a3() {
+      var isInstance = false;
+      try {
+        isInstance = this instanceof a3;
+      } catch {
+      }
+      if (isInstance) {
+        return Reflect.construct(f, arguments, this.constructor);
+      }
+      return f.apply(this, arguments);
+    };
+    a2.prototype = f.prototype;
+  } else a2 = {};
+  Object.defineProperty(a2, "__esModule", { value: true });
+  Object.keys(n).forEach(function(k) {
+    var d = Object.getOwnPropertyDescriptor(n, k);
+    Object.defineProperty(a2, k, d.get ? d : {
+      enumerable: true,
+      get: function() {
+        return n[k];
+      }
+    });
+  });
+  return a2;
+}
 var extend$1;
 var hasRequiredExtend;
 function requireExtend() {
@@ -20662,11 +20692,7 @@ const convertHtmlToMarkdown = async (html2, baseUrl, locale) => {
           if (children.length === 0) {
             return { type: "html", value: `<span id="${String(id)}"></span>` };
           }
-          return [
-            { type: "html", value: `<span id="${String(id)}">` },
-            ...children,
-            { type: "html", value: "</span>" }
-          ];
+          return [{ type: "html", value: `<span id="${String(id)}">` }, ...children, { type: "html", value: "</span>" }];
         }
         return state.all(node2);
       },
@@ -20760,23 +20786,35 @@ const convertHtmlToMarkdown = async (html2, baseUrl, locale) => {
   console.log(getMessage(locale, "convert_success"));
   return correctedMarkdown;
 };
-const htmlToMarkdown = async (url, options = {}) => {
-  const { locale = "en-US" } = options;
+const htmlToMarkdown = async (urlOrHtml, options = {}) => {
+  const { locale = "en-US", htmlContent } = options;
   try {
-    new URL(url);
-    const htmlContent = await getHtmlWithPlaywright(url, locale);
-    if (!htmlContent) {
-      console.error(`No HTML content found for URL: ${url}`);
+    let htmlContentToProcess;
+    let sourceUrl;
+    if (htmlContent) {
+      htmlContentToProcess = htmlContent;
+      sourceUrl = urlOrHtml || "direct-html-input";
+    } else {
+      new URL(urlOrHtml);
+      htmlContentToProcess = await getHtmlWithPlaywright(urlOrHtml, locale);
+      sourceUrl = urlOrHtml;
+    }
+    if (!htmlContentToProcess) {
+      console.error(`No HTML content found for source: ${sourceUrl}`);
       return { html: "", markdown: "" };
     }
-    const markdown = await convertHtmlToMarkdown(htmlContent, url, locale);
-    return { html: htmlContent, markdown };
+    const markdown = await convertHtmlToMarkdown(htmlContentToProcess, sourceUrl, locale);
+    return { html: htmlContentToProcess, markdown };
   } catch (error) {
-    console.error(`An error occurred during the conversion process for URL: ${url}`, error);
+    const source = htmlContent ? "HTML string" : `URL: ${urlOrHtml}`;
+    console.error(`An error occurred during the conversion process for ${source}`, error);
     return { html: "", markdown: "" };
   }
 };
 export {
+  getAugmentedNamespace as a,
+  getDefaultExportFromCjs as b,
+  commonjsGlobal as c,
   getMessage as g,
   htmlToMarkdown as h
 };

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 o=require("./core-CVMYjtlI.cjs");exports.getMessage=o.getMessage,exports.htmlToMarkdown=o.htmlToMarkdown;
+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-D4Mj1IQ9.cjs"),c=require("./mcp-server-streamable-Bu7QAr3W.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

@@ -5,8 +5,14 @@ Object.assign(global, {
   window,
   self: window
 });
-import { g, h } from "./core-DOzMLPij.js";
+import { g, h } from "./core-DU_246Q7.js";
+import { c, r } from "./mcp-server-CpmnW1p8.js";
+import { c as c2, r as r2 } from "./mcp-server-streamable-CQD5zK30.js";
 export {
+  c as createMcpServer,
+  c2 as createStreamableMcpServer,
   g as getMessage,
-  h as htmlToMarkdown
+  h as htmlToMarkdown,
+  r as runMcpServer,
+  r2 as runStreamableMcpServer
 };

package/dist/main.cjs

@@ -1,2 +1,2 @@
 #!/usr/bin/env node
-Object.create,Object.defineProperty,Object.getOwnPropertyDescriptor,Object.getOwnPropertyNames,Object.getPrototypeOf,Object.prototype.hasOwnProperty;const e=new(require("happy-dom").Window);Object.assign(global,{document:e.document,window:e,self:e});const t=require("./core-CVMYjtlI.cjs");var r="undefined"!=typeof document?document.currentScript:null;(async()=>{(("undefined"==typeof document?require("url").pathToFileURL(__filename).href:r&&"SCRIPT"===r.tagName.toUpperCase()&&r.src||new URL("main.cjs",document.baseURI).href).endsWith(process.argv[1])||process.argv[1]&&process.argv[1].includes("vite-node"))&&(async()=>{const e=(await import("yargs-parser")).default,r=await import("node:fs"),o=await import("node:path"),c=e(process.argv.slice(2),{string:["locale","output"],alias:{output:["o"]},default:{locale:"en-US"}}),s=c._[0],n=c.locale,a=c.output;s||process.exit(1);try{new URL(s)}catch(i){process.exit(1)}try{const{markdown:e,html:c}=await t.htmlToMarkdown(s,{locale:n});let i;if(a){i=o.resolve(process.cwd(),a);const e=o.dirname(i);r.existsSync(e)||r.mkdirSync(e,{recursive:!0})}else{const e=`${s.replace(/https?:\/\//,"").replace(/[^a-zA-Z0-9]/g,"_")}_${Date.now()}.md`,t=o.join(process.cwd(),".outputs","raw");r.existsSync(t)||r.mkdirSync(t,{recursive:!0}),i=o.join(t,e)}r.writeFileSync(i,e);const p=o.parse(i),l=o.join(p.dir,`${p.name}.html`);r.writeFileSync(l,c)}catch(i){process.exit(1)}})()})();
+Object.create,Object.defineProperty,Object.getOwnPropertyDescriptor,Object.getOwnPropertyNames,Object.getPrototypeOf,Object.prototype.hasOwnProperty;const e=new(require("happy-dom").Window);Object.assign(global,{document:e.document,window:e,self:e});const t=require("./core-D2dlEt1l.cjs");var r="undefined"!=typeof document?document.currentScript:null;(async()=>{(("undefined"==typeof document?require("url").pathToFileURL(__filename).href:r&&"SCRIPT"===r.tagName.toUpperCase()&&r.src||new URL("main.cjs",document.baseURI).href).endsWith(process.argv[1])||process.argv[1]&&process.argv[1].includes("vite-node"))&&(async()=>{const e=(await import("yargs-parser")).default,r=await import("node:fs"),o=await import("node:path"),c=e(process.argv.slice(2),{string:["locale","output"],alias:{output:["o"]},default:{locale:"en-US"}}),s=c._[0],n=c.locale,a=c.output;s||process.exit(1);try{new URL(s)}catch(i){process.exit(1)}try{const{markdown:e,html:c}=await t.htmlToMarkdown(s,{locale:n});let i;if(a){i=o.resolve(process.cwd(),a);const e=o.dirname(i);r.existsSync(e)||r.mkdirSync(e,{recursive:!0})}else{const e=`${s.replace(/https?:\/\//,"").replace(/[^a-zA-Z0-9]/g,"_")}_${Date.now()}.md`,t=o.join(process.cwd(),".outputs","raw");r.existsSync(t)||r.mkdirSync(t,{recursive:!0}),i=o.join(t,e)}r.writeFileSync(i,e);const p=o.parse(i),l=o.join(p.dir,`${p.name}.html`);r.writeFileSync(l,c)}catch(i){process.exit(1)}})()})();

package/dist/main.js

@@ -6,7 +6,7 @@ Object.assign(global, {
   window,
   self: window
 });
-import { g as getMessage, h as htmlToMarkdown } from "./core-DOzMLPij.js";
+import { g as getMessage, h as htmlToMarkdown } from "./core-DU_246Q7.js";
 const runCli = async () => {
   const yargsParser = (await import("yargs-parser")).default;
   const fs = await import("node:fs");

package/dist/mcp-server-CpmnW1p8.js

@@ -0,0 +1,283 @@
+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, C as CallToolRequestSchema } from "./version-BRVBmpor.js";
+import process from "node:process";
+import { h as htmlToMarkdown } from "./core-DU_246Q7.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);
+      }
+    });
+  }
+}
+function parseHtmlToMarkdownArgs(args) {
+  if (!args || typeof args !== "object") {
+    throw new Error("Arguments are required");
+  }
+  const { url, html_content, locale = "en-US", output_format = "markdown" } = args;
+  if (!url && !html_content) {
+    throw new Error("Either 'url' or 'html_content' is required");
+  }
+  if (url && typeof url !== "string") {
+    throw new Error("'url' must be a string");
+  }
+  if (html_content && typeof html_content !== "string") {
+    throw new Error("'html_content' must be a string");
+  }
+  if (locale && !["en-US", "ja-JP"].includes(locale)) {
+    throw new Error("'locale' must be 'en-US' or 'ja-JP'");
+  }
+  if (output_format && !["markdown", "html", "both"].includes(output_format)) {
+    throw new Error("'output_format' must be 'markdown', 'html', or 'both'");
+  }
+  return {
+    url,
+    html_content,
+    locale,
+    output_format
+  };
+}
+class HtmlToMarkdownMcpServer {
+  constructor() {
+    this.version = getPackageVersion();
+    this.server = new Server(
+      {
+        name: "@aiquants/html-to-markdown",
+        version: this.version,
+        description: "A tool to dynamically fetch a web page from a given URL and convert it to Markdown."
+      },
+      {
+        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 HTML content from a URL or HTML string to Markdown format. Supports dynamic content loading and preserves table structures.",
+            inputSchema: {
+              type: "object",
+              properties: {
+                url: {
+                  type: "string",
+                  description: "The URL of the web page to convert to Markdown"
+                },
+                html_content: {
+                  type: "string",
+                  description: "HTML content as a string to convert to Markdown (alternative to URL)"
+                },
+                locale: {
+                  type: "string",
+                  description: "The locale to use for browser context (en-US or ja-JP)",
+                  enum: ["en-US", "ja-JP"],
+                  default: "en-US"
+                },
+                output_format: {
+                  type: "string",
+                  description: "Output format: 'markdown' (default), 'html', or 'both'",
+                  enum: ["markdown", "html", "both"],
+                  default: "markdown"
+                }
+              },
+              anyOf: [{ required: ["url"] }, { required: ["html_content"] }]
+            }
+          }
+        ]
+      };
+    });
+    this.server.setRequestHandler(CallToolRequestSchema, async (request) => {
+      switch (request.params.name) {
+        case "html_to_markdown": {
+          return await this.handleHtmlToMarkdown(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", output_format = "markdown" } = 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
+      });
+      let content;
+      switch (output_format) {
+        case "markdown":
+          content = [
+            {
+              type: "text",
+              text: result.markdown
+            }
+          ];
+          break;
+        case "html":
+          content = [
+            {
+              type: "text",
+              text: result.html
+            }
+          ];
+          break;
+        case "both":
+          content = [
+            {
+              type: "text",
+              text: result.markdown
+            },
+            {
+              type: "text",
+              text: result.html
+            }
+          ];
+          break;
+        default:
+          throw new Error(`Invalid output format: ${output_format}`);
+      }
+      return {
+        content,
+        isError: false
+      };
+    } catch (error) {
+      const errorMessage = error instanceof Error ? error.message : String(error);
+      return {
+        content: [
+          {
+            type: "text",
+            text: `Error: ${errorMessage}`
+          }
+        ],
+        isError: true
+      };
+    }
+  }
+  /**
+   * Start the MCP server.
+   * MCP サーバーを開始
+   */
+  async start() {
+    const transport = new StdioServerTransport();
+    await this.server.connect(transport);
+    console.error("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-D4Mj1IQ9.cjs

@@ -0,0 +1 @@
+const t=new(require("happy-dom").Window);Object.assign(global,{document:t.document,window:t,self:t});const r=require("./version-DTaUKNBS.cjs"),e=require("node:process"),o=require("./core-D2dlEt1l.cjs");class n{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 e=this._buffer.toString("utf8",0,t).replace(/\r$/,"");return this._buffer=this._buffer.subarray(t+1),function(t){return r.JSONRPCMessageSchema.parse(JSON.parse(t))}(e)}clear(){this._buffer=void 0}}class s{constructor(t=e.stdin,r=e.stdout){this._stdin=t,this._stdout=r,this._readBuffer=new n,this._started=!1,this._ondata=t=>{this._readBuffer.append(t),this.processReadBuffer()},this._onerror=t=>{var r;null===(r=this.onerror)||void 0===r||r.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,r;;)try{const r=this._readBuffer.readMessage();if(null===r)break;null===(t=this.onmessage)||void 0===t||t.call(this,r)}catch(e){null===(r=this.onerror)||void 0===r||r.call(this,e)}}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(r=>{const e=function(t){return JSON.stringify(t)+"\n"}(t);this._stdout.write(e)?r():this._stdout.once("drain",r)})}}class a{constructor(){this.version=r.getPackageVersion(),this.server=new r.Server({name:"@aiquants/html-to-markdown",version:this.version,description:"A tool to dynamically fetch a web page from a given URL and convert it to Markdown."},{capabilities:{tools:{}}}),this.setupToolHandlers()}setupToolHandlers(){this.server.setRequestHandler(r.ListToolsRequestSchema,async()=>({tools:[{name:"html_to_markdown",description:"Convert HTML content from a URL or HTML string to Markdown format. Supports dynamic content loading and preserves table structures.",inputSchema:{type:"object",properties:{url:{type:"string",description:"The URL of the web page to convert to Markdown"},html_content:{type:"string",description:"HTML content as a string to convert to Markdown (alternative to URL)"},locale:{type:"string",description:"The locale to use for browser context (en-US or ja-JP)",enum:["en-US","ja-JP"],default:"en-US"},output_format:{type:"string",description:"Output format: 'markdown' (default), 'html', or 'both'",enum:["markdown","html","both"],default:"markdown"}},anyOf:[{required:["url"]},{required:["html_content"]}]}}]})),this.server.setRequestHandler(r.CallToolRequestSchema,async t=>{if("html_to_markdown"===t.params.name)return await this.handleHtmlToMarkdown(t.params.arguments);throw new Error(`Unknown tool: ${t.params.name}`)})}async handleHtmlToMarkdown(t){try{const r=function(t){if(!t||"object"!=typeof t)throw new Error("Arguments are required");const{url:r,html_content:e,locale:o="en-US",output_format:n="markdown"}=t;if(!r&&!e)throw new Error("Either 'url' or 'html_content' is required");if(r&&"string"!=typeof r)throw new Error("'url' must be a string");if(e&&"string"!=typeof e)throw new Error("'html_content' must be a string");if(o&&!["en-US","ja-JP"].includes(o))throw new Error("'locale' must be 'en-US' or 'ja-JP'");if(n&&!["markdown","html","both"].includes(n))throw new Error("'output_format' must be 'markdown', 'html', or 'both'");return{url:r,html_content:e,locale:o,output_format:n}}(t),{url:e,html_content:n,locale:s="en-US",output_format:a="markdown"}=r;if(e)try{new URL(e)}catch{throw new Error(`Invalid URL: ${e}`)}const i=await o.htmlToMarkdown(e||"direct-html-input",{locale:s,htmlContent:n});let c;switch(a){case"markdown":c=[{type:"text",text:i.markdown}];break;case"html":c=[{type:"text",text:i.html}];break;case"both":c=[{type:"text",text:i.markdown},{type:"text",text:i.html}];break;default:throw new Error(`Invalid output format: ${a}`)}return{content:c,isError:!1}}catch(r){return{content:[{type:"text",text:`Error: ${r instanceof Error?r.message:String(r)}`}],isError:!0}}}async start(){const t=new s;await this.server.connect(t)}}const i=()=>new a;exports.createMcpServer=i,exports.runMcpServer=async()=>{const t=i();await t.start()};