@aiquants/html-to-markdown 0.1.7 → 0.2.0

package/README.md

@@ -133,9 +133,35 @@ npx --package=@aiquants/html-to-markdown aiq-html2md-mcp-stream
 #### MCP Tools Available
 
 - **html_to_markdown**: Convert HTML content from a URL to Markdown format
-  - `url` (required): The URL of the web page to convert
+  - `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
+
+**File Saving Options:**
+
+- **`save_path`**: Specify the complete file path including filename and extension where you want to save the content.
+  - Example: `/path/to/output/my-page.md`
+  - The directory will be created automatically if it doesn't exist.
+  - You cannot use both `save_path` and `save_directory` at the same time.
+
+- **`save_directory`**: Specify only the directory where you want to save the file. The filename will be auto-generated based on the URL.
+  - Example: `/path/to/output/` (filename will be auto-generated like `page.md`)
+  - For URLs like `https://example.com/articles/my-article`, the filename becomes `my-article.md`
+  - The directory will be created automatically if it doesn't exist.
+  - You cannot use both `save_path` and `save_directory` at the same time.
+
+**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`
 
 - **html_to_markdown_streamable**: Same as above but with streamable MCP support for progress updates
 

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-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;
+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-B5lrAMe3.cjs"),c=require("./mcp-server-streamable-B3R3qi85.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

@@ -6,8 +6,8 @@ Object.assign(global, {
   self: window
 });
 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";
+import { c, r } from "./mcp-server-Ct2TYHcU.js";
+import { c as c2, r as r2 } from "./mcp-server-streamable-C6-qOLdK.js";
 export {
   c as createMcpServer,
   c2 as createStreamableMcpServer,

package/dist/mcp-server-B5lrAMe3.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-DTaUKNBS.cjs"),r=require("node:process"),n=require("fs"),o=require("path"),a=require("./core-D2dlEt1l.cjs");class s{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$/,"");return this._buffer=this._buffer.subarray(t+1),function(t){return e.JSONRPCMessageSchema.parse(JSON.parse(t))}(r)}clear(){this._buffer=void 0}}class i{constructor(t=r.stdin,e=r.stdout){this._stdin=t,this._stdout=e,this._readBuffer=new s,this._started=!1,this._ondata=t=>{this._readBuffer.append(t),this.processReadBuffer()},this._onerror=t=>{var e;null===(e=this.onerror)||void 0===e||e.call(this,t)}}async start(){if(this._started)throw new Error("StdioServerTransport already started! If using Server class, note that connect() calls start() automatically.");this._started=!0,this._stdin.on("data",this._ondata),this._stdin.on("error",this._onerror)}processReadBuffer(){for(var t,e;;)try{const e=this._readBuffer.readMessage();if(null===e)break;null===(t=this.onmessage)||void 0===t||t.call(this,e)}catch(r){null===(e=this.onerror)||void 0===e||e.call(this,r)}}async close(){var t;this._stdin.off("data",this._ondata),this._stdin.off("error",this._onerror);0===this._stdin.listenerCount("data")&&this._stdin.pause(),this._readBuffer.clear(),null===(t=this.onclose)||void 0===t||t.call(this)}send(t){return new Promise(e=>{const r=function(t){return JSON.stringify(t)+"\n"}(t);this._stdout.write(r)?e():this._stdout.once("drain",e)})}}function c(t,e){if(t&&"direct-html-input"!==t)try{let e=new URL(t).pathname.split("/").pop()||"page";return e=e.replace(/\.[^.]*$/,""),e=e.replace(/[<>:"/\\|?*]/g,"_"),e&&!e.match(/^\.+$/)||(e="page"),e}catch{return"page"}return"page"}async function h(t,e,r,a,s){if(!e&&!r)return null;let i;if(e)i=o.resolve(e);else{if(!r)return null;{const t=c(a),e="html"===s?".html":".md";i=o.resolve(r,`${t}${e}`)}}return await n.promises.mkdir(o.dirname(i),{recursive:!0}),await n.promises.writeFile(i,t,"utf-8"),i}class l{constructor(){this.version=e.getPackageVersion(),this.server=new e.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(e.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"},save_path:{type:"string",description:"Full path (including filename) where to save the converted content. Cannot be used with save_directory."},save_directory:{type:"string",description:"Directory path where to save the converted content with auto-generated filename. Cannot be used with save_path."}},anyOf:[{required:["url"]},{required:["html_content"]}]}}]})),this.server.setRequestHandler(e.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 e=function(t){if(!t||"object"!=typeof t)throw new Error("Arguments are required");const{url:e,html_content:r,locale:n="en-US",output_format:o="markdown",save_path:a,save_directory:s}=t;if(!e&&!r)throw new Error("Either 'url' or 'html_content' is required");if(e&&"string"!=typeof e)throw new Error("'url' must be a string");if(r&&"string"!=typeof r)throw new Error("'html_content' must be a string");if(n&&!["en-US","ja-JP"].includes(n))throw new Error("'locale' must be 'en-US' or 'ja-JP'");if(o&&!["markdown","html","both"].includes(o))throw new Error("'output_format' must be 'markdown', 'html', or 'both'");if(a&&"string"!=typeof a)throw new Error("'save_path' must be a string");if(s&&"string"!=typeof s)throw new Error("'save_directory' must be a string");if(a&&s)throw new Error("Cannot specify both 'save_path' and 'save_directory'");return{url:e,html_content:r,locale:n,output_format:o,save_path:a,save_directory:s}}(t),{url:r,html_content:n,locale:s="en-US",output_format:i="markdown",save_path:l,save_directory:d}=e;if(r)try{new URL(r)}catch{throw new Error(`Invalid URL: ${r}`)}const u=await a.htmlToMarkdown(r||"direct-html-input",{locale:s,htmlContent:n}),m=[];if(l||d)switch(i){case"markdown":{const t=await h(u.markdown,l,d,r,"markdown");t&&m.push(t);break}case"html":{const t=await h(u.html,l,d,r,"html");t&&m.push(t);break}case"both":{const t=l?l.endsWith(".md")?l:`${l}.md`:d?o.join(d,`${c(r)}.md`):void 0;if(t){const e=await h(u.markdown,t,void 0,r,"markdown");e&&m.push(e)}const e=l?l.endsWith(".html")?l:`${l}.html`:d?o.join(d,`${c(r)}.html`):void 0;if(e){const t=await h(u.html,e,void 0,r,"html");t&&m.push(t)}break}}let f="";switch(i){case"markdown":f=u.markdown;break;case"html":f=u.html;break;case"both":f=`=== Markdown ===\n${u.markdown}\n\n=== HTML ===\n${u.html}`;break;default:throw new Error(`Invalid output format: ${i}`)}m.length>0&&(f+=`\n\n=== Files Saved ===\n${m.map(t=>`- ${t}`).join("\n")}`);return{content:[{type:"text",text:f}],isError:!1}}catch(e){return{content:[{type:"text",text:`Error: ${e instanceof Error?e.message:String(e)}`}],isError:!0}}}async start(){const t=new i;await this.server.connect(t)}}const d=()=>new l;exports.createMcpServer=d,exports.runMcpServer=async()=>{const t=d();await t.start()};

package/dist/{mcp-server-CpmnW1p8.js → mcp-server-Ct2TYHcU.js}

@@ -7,6 +7,8 @@ Object.assign(global, {
 });
 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 { promises } from "fs";
+import { join, resolve, dirname } from "path";
 import { h as htmlToMarkdown } from "./core-DU_246Q7.js";
 class ReadBuffer {
   append(chunk) {
@@ -86,12 +88,12 @@ class StdioServerTransport {
     (_a = this.onclose) === null || _a === void 0 ? void 0 : _a.call(this);
   }
   send(message) {
-    return new Promise((resolve) => {
+    return new Promise((resolve2) => {
       const json = serializeMessage(message);
       if (this._stdout.write(json)) {
-        resolve();
+        resolve2();
       } else {
-        this._stdout.once("drain", resolve);
+        this._stdout.once("drain", resolve2);
       }
     });
   }
@@ -100,7 +102,7 @@ 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;
+  const { url, html_content, locale = "en-US", output_format = "markdown", save_path, save_directory } = args;
   if (!url && !html_content) {
     throw new Error("Either 'url' or 'html_content' is required");
   }
@@ -116,13 +118,59 @@ function parseHtmlToMarkdownArgs(args) {
   if (output_format && !["markdown", "html", "both"].includes(output_format)) {
     throw new Error("'output_format' must be 'markdown', 'html', or 'both'");
   }
+  if (save_path && typeof save_path !== "string") {
+    throw new Error("'save_path' must be a string");
+  }
+  if (save_directory && typeof save_directory !== "string") {
+    throw new Error("'save_directory' must be a string");
+  }
+  if (save_path && save_directory) {
+    throw new Error("Cannot specify both 'save_path' and 'save_directory'");
+  }
   return {
     url,
     html_content,
     locale,
-    output_format
+    output_format,
+    save_path,
+    save_directory
   };
 }
+function generateFilename(url, outputFormat) {
+  if (url && url !== "direct-html-input") {
+    try {
+      const urlObj = new URL(url);
+      let filename = urlObj.pathname.split("/").pop() || "page";
+      filename = filename.replace(/\.[^.]*$/, "");
+      filename = filename.replace(/[<>:"/\\|?*]/g, "_");
+      if (!filename || filename.match(/^\.+$/)) {
+        filename = "page";
+      }
+      return filename;
+    } catch {
+      return "page";
+    }
+  }
+  return "page";
+}
+async function saveContentToFile(content, savePath, saveDirectory, url, outputFormat) {
+  if (!savePath && !saveDirectory) {
+    return null;
+  }
+  let finalPath;
+  if (savePath) {
+    finalPath = resolve(savePath);
+  } else if (saveDirectory) {
+    const filename = generateFilename(url);
+    const extension = outputFormat === "html" ? ".html" : ".md";
+    finalPath = resolve(saveDirectory, `${filename}${extension}`);
+  } else {
+    return null;
+  }
+  await promises.mkdir(dirname(finalPath), { recursive: true });
+  await promises.writeFile(finalPath, content, "utf-8");
+  return finalPath;
+}
 class HtmlToMarkdownMcpServer {
   constructor() {
     this.version = getPackageVersion();
@@ -173,6 +221,14 @@ class HtmlToMarkdownMcpServer {
                   description: "Output format: 'markdown' (default), 'html', or 'both'",
                   enum: ["markdown", "html", "both"],
                   default: "markdown"
+                },
+                save_path: {
+                  type: "string",
+                  description: "Full path (including filename) where to save the converted content. Cannot be used with save_directory."
+                },
+                save_directory: {
+                  type: "string",
+                  description: "Directory path where to save the converted content with auto-generated filename. Cannot be used with save_path."
                 }
               },
               anyOf: [{ required: ["url"] }, { required: ["html_content"] }]
@@ -198,7 +254,7 @@ class HtmlToMarkdownMcpServer {
   async handleHtmlToMarkdown(args) {
     try {
       const validatedArgs = parseHtmlToMarkdownArgs(args);
-      const { url, html_content, locale = "en-US", output_format = "markdown" } = validatedArgs;
+      const { url, html_content, locale = "en-US", output_format = "markdown", save_path, save_directory } = validatedArgs;
       if (url) {
         try {
           new URL(url);
@@ -210,39 +266,64 @@ class HtmlToMarkdownMcpServer {
         locale,
         htmlContent: html_content
       });
-      let content;
+      const savedPaths = [];
+      if (save_path || save_directory) {
+        switch (output_format) {
+          case "markdown": {
+            const savedPath = await saveContentToFile(result.markdown, save_path, save_directory, url, "markdown");
+            if (savedPath) savedPaths.push(savedPath);
+            break;
+          }
+          case "html": {
+            const savedPath = await saveContentToFile(result.html, save_path, save_directory, url, "html");
+            if (savedPath) savedPaths.push(savedPath);
+            break;
+          }
+          case "both": {
+            const markdownPath = save_path ? save_path.endsWith(".md") ? save_path : `${save_path}.md` : save_directory ? join(save_directory, `${generateFilename(url)}.md`) : void 0;
+            if (markdownPath) {
+              const savedMarkdownPath = await saveContentToFile(result.markdown, markdownPath, void 0, url, "markdown");
+              if (savedMarkdownPath) savedPaths.push(savedMarkdownPath);
+            }
+            const htmlPath = save_path ? save_path.endsWith(".html") ? save_path : `${save_path}.html` : save_directory ? join(save_directory, `${generateFilename(url)}.html`) : void 0;
+            if (htmlPath) {
+              const savedHtmlPath = await saveContentToFile(result.html, htmlPath, void 0, url, "html");
+              if (savedHtmlPath) savedPaths.push(savedHtmlPath);
+            }
+            break;
+          }
+        }
+      }
+      let responseText = "";
       switch (output_format) {
         case "markdown":
-          content = [
-            {
-              type: "text",
-              text: result.markdown
-            }
-          ];
+          responseText = result.markdown;
           break;
         case "html":
-          content = [
-            {
-              type: "text",
-              text: result.html
-            }
-          ];
+          responseText = result.html;
           break;
         case "both":
-          content = [
-            {
-              type: "text",
-              text: result.markdown
-            },
-            {
-              type: "text",
-              text: result.html
-            }
-          ];
+          responseText = `=== Markdown ===
+${result.markdown}
+
+=== HTML ===
+${result.html}`;
           break;
         default:
           throw new Error(`Invalid output format: ${output_format}`);
       }
+      if (savedPaths.length > 0) {
+        responseText += `
+
+=== Files Saved ===
+${savedPaths.map((path) => `- ${path}`).join("\n")}`;
+      }
+      const content = [
+        {
+          type: "text",
+          text: responseText
+        }
+      ];
       return {
         content,
         isError: false