Merge pull request #156 from TxtDot/engines-refactor

Add jsx support to engines plugins

packages/plugins/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@txtdot/plugins",
-  "version": "1.1.1",
+  "version": "2.0.0",
   "description": "Official txtdot plugins",
   "main": "dist/lib.js",
   "types": "dist/lib.d.ts",
@@ -19,6 +19,7 @@
   "license": "MIT",
   "dependencies": {
     "@mozilla/readability": "^0.5.0",
+    "linkedom": "^0.18.0",
     "@txtdot/sdk": "workspace:*"
   },
   "devDependencies": {

packages/plugins/src/components/searchers.tsx

@@ -0,0 +1,41 @@
+import { JSX } from '@txtdot/sdk';
+
+export function PageFooter({
+  page,
+  previous,
+  next,
+}: {
+  page: number;
+  previous: string | false;
+  next: string | false;
+}) {
+  return (
+    <>
+      {page !== 1 ? (
+        <>
+          <a href={previous}>Previous </a> |
+        </>
+      ) : (
+        <></>
+      )}
+      <a href={next}> Next</a>
+    </>
+  );
+}
+
+export function ResultItem({
+  url,
+  title,
+  content,
+}: {
+  url: string;
+  title: string;
+  content: string;
+}) {
+  return (
+    <>
+      <a href={url}>{title}</a>
+      <p>{content}</p>
+    </>
+  );
+}

packages/plugins/src/engines/readability.ts

@@ -1,7 +1,7 @@
 import { Readability as OReadability } from '@mozilla/readability';
-import { EngineParseError } from '@txtdot/sdk/dist/types/errors';
 
-import { Engine } from '@txtdot/sdk';
+import { Engine, EngineParseError } from '@txtdot/sdk';
+import { parseHTML } from 'linkedom';
 
 const Readability = new Engine(
   'Readability',
@@ -10,7 +10,7 @@ const Readability = new Engine(
 );
 
 Readability.route('*path', async (input, ro) => {
-  const reader = new OReadability(input.parseDom().window.document);
+  const reader = new OReadability(input.document);
   const parsed = reader.parse();
 
   if (!parsed) {

packages/plugins/src/engines/searx.tsx

@@ -1,5 +1,7 @@
-import { Engine } from '@txtdot/sdk';
-import { HandlerInput, Route } from '@txtdot/sdk/dist/types/handler';
+import { Engine, JSX } from '@txtdot/sdk';
+import { HandlerInput, Route } from '@txtdot/sdk';
+import { parseHTML } from 'linkedom';
+import { PageFooter, ResultItem } from '../components/searchers';
 
 const SearX = new Engine('SearX', "Engine for searching with 'SearXNG'", [
   'searx.*',
@@ -9,17 +11,23 @@ async function search(
   input: HandlerInput,
   ro: Route<{ search: string; pageno?: string }>
 ) {
-  const document = input.parseDom().window.document;
+  const document = input.document;
   const search = ro.q.search;
   const page = parseInt(ro.q.pageno || '1');
 
-  const page_footer = `${
-    page !== 1
-      ? `<a href="${ro.reverse({ search, pageno: page - 1 })}">Previous </a>|`
-      : ''
-  }<a href="${ro.reverse({ search, pageno: page + 1 })}"> Next</a>`;
+  let previous: string | false;
+  let next: string | false;
+
+  if (ro.q.pageno) {
+    previous = ro.reverse({ search, pageno: page - 1 });
+    next = ro.reverse({ search, pageno: page + 1 });
+  } else {
+    previous = false;
+    next = `/search?q=${search}&pageno=${page + 1}`;
+  }
 
   const articles = Array.from(document.querySelectorAll('.result'));
+
   const articles_parsed = articles.map((a) => {
     const parsed = {
       url:
@@ -34,21 +42,24 @@ async function search(
     };
 
     return {
-      html: `<a href="${parsed.url}">${parsed.title}</a><p>${parsed.content}</p><hr>`,
+      html: <ResultItem {...parsed} />,
       text: `${parsed.title} (${parsed.url})\n${parsed.content}\n---\n\n`,
     };
   });
 
-  const content = `${articles_parsed
-    .map((a) => a.html)
-    .join('')}${page_footer}`;
+  const content = (
+    <>
+      {articles_parsed.map((a) => a.html).join('')}
+      <PageFooter page={page} previous={previous} next={next} />
+    </>
+  );
+
   const textContent = articles_parsed.map((a) => a.text).join('');
 
   return {
-    content,
+    content: content,
     textContent,
     title: `${search} - Searx - Page ${page}`,
-    lang: document.documentElement.lang,
   };
 }
 

packages/plugins/src/engines/stackoverflow/questions.tsx

@@ -1,10 +1,11 @@
-import { HandlerInput, Route } from '@txtdot/sdk/dist/types/handler';
+import { HandlerInput, Route } from '@txtdot/sdk';
+import { JSX } from '@txtdot/sdk';
 
 async function questions(
   input: HandlerInput,
   ro: Route<{ id: string; slug: string }>
 ) {
-  const document = input.parseDom().window.document;
+  const document = input.document;
 
   const questionEl = document.getElementById('question');
   const question = postParser(questionEl);
@@ -15,12 +16,15 @@ async function questions(
   const answers = allAnswers.map((a) => postParser(a));
 
   return {
-    content: `${question}<hr>${answers.length} answers <hr>${answers.join(
-      '<hr>'
-    )}`,
-    textContent: `${ro.q.id}/${ro.q.slug}\nText output not supported`, // TODO
+    content: (
+      <>
+        {question}
+        <hr />
+        {answers.length} answers <hr />
+        {answers.join(<hr />)}
+      </>
+    ),
     title,
-    lang: document.documentElement.lang,
   };
 }
 
@@ -37,12 +41,27 @@ function postParser(el: Element | null): string {
       (el.querySelector('.user-details a') as HTMLAnchorElement)?.href || '';
     const userTitle = el.querySelector('.user-action-time')?.textContent || '';
 
-    return `<h4>${userTitle}${
-      userUrl ? ` by <a href="${userUrl}">${userName}</a>` : ''
-    }</h4>`;
+    return (
+      <h4>
+        {userTitle}
+        {userUrl ? (
+          <>
+            by <a href={userUrl}>{userName}</a>
+          </>
+        ) : (
+          <></>
+        )}
+      </h4>
+    );
   });
 
-  return `<h3>${voteCount} votes</h3>${body}${footer.join('')}`;
+  return (
+    <>
+      <h3>{voteCount} votes</h3>
+      {body}
+      {footer.join('')}
+    </>
+  );
 }
 
 export default questions;

packages/plugins/src/engines/stackoverflow/users.tsx

@@ -1,10 +1,11 @@
-import { HandlerInput, Route } from '@txtdot/sdk/dist/types/handler';
+import { HandlerInput, Route } from '@txtdot/sdk';
+import { JSX } from '@txtdot/sdk';
 
 async function users(
   input: HandlerInput,
   ro: Route<{ id: string; slug: string }>
 ) {
-  const document = input.parseDom().window.document;
+  const document = input.document;
 
   const userInfo =
     document.querySelector('.md\\:ai-start > div:nth-child(2)')?.textContent ||
@@ -21,15 +22,28 @@ async function users(
       const type =
         el.querySelector('.iconAnswer, .iconQuestion')?.textContent || '';
 
-      return `<strong>${type} (${votes}) </strong><a href="${url}">${title}</a>`;
+      return (
+        <>
+          <strong>
+            {type} ({votes}){' '}
+          </strong>
+          <a href={url}>{title}</a>
+        </>
+      );
     })
-    .join('<br/>');
+    .join(<br />);
 
   return {
-    content: `${userInfo}<hr><h3>Top Posts</h3>${topPosts}`,
+    content: (
+      <>
+        {userInfo}
+        <hr />
+        <h3>Top Posts</h3>
+        {topPosts}
+      </>
+    ),
     textContent: `${ro.q.id}/${ro.q.slug}\n`, // TODO
     title: document.querySelector('title')?.textContent || '',
-    lang: document.documentElement.lang,
   };
 }
 

packages/plugins/tsconfig.json

@@ -13,13 +13,13 @@
     /* Language and Environment */
     "target": "ES2020" /* Set the JavaScript language version for emitted JavaScript and include compatible library declarations. */,
     // "lib": [],                                        /* Specify a set of bundled library declaration files that describe the target runtime environment. */
-    // "jsx": "preserve",                                /* Specify what JSX code is generated. */
+    "jsx": "react" /* Specify what JSX code is generated. */,
     // "experimentalDecorators": true,                   /* Enable experimental support for legacy experimental decorators. */
     // "emitDecoratorMetadata": true,                    /* Emit design-type metadata for decorated declarations in source files. */
     // "jsxFactory": "",                                 /* Specify the JSX factory function used when targeting React JSX emit, e.g. 'React.createElement' or 'h'. */
     // "jsxFragmentFactory": "",                         /* Specify the JSX Fragment reference used for fragments when targeting React JSX emit e.g. 'React.Fragment' or 'Fragment'. */
     // "jsxImportSource": "",                            /* Specify module specifier used to import the JSX factory functions when using 'jsx: react-jsx*'. */
-    // "reactNamespace": "",                             /* Specify the object invoked for 'createElement'. This only applies when targeting 'react' JSX emit. */
+    "reactNamespace": "JSX" /* Specify the object invoked for 'createElement'. This only applies when targeting 'react' JSX emit. */,
     // "noLib": true,                                    /* Disable including any library files, including the default lib.d.ts. */
     // "useDefineForClassFields": true,                  /* Emit ECMAScript-standard-compliant class fields. */
     // "moduleDetection": "auto",                        /* Control what method is used to detect module-format JS files. */

packages/sdk/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@txtdot/sdk",
-  "version": "1.1.1",
+  "version": "2.0.0",
   "description": "SDK for creating plugins for TxtDot",
   "main": "dist/lib.js",
   "types": "dist/lib.d.ts",

packages/sdk/src/engine.ts

@@ -2,9 +2,9 @@ import Route from 'route-parser';
 
 import {
   HandlerInput,
-  IHandlerOutput,
   EngineFunction,
   RouteValues,
+  EngineOutput,
 } from './types/handler';
 
 import { NoHandlerFoundError } from './types/errors';
@@ -33,7 +33,7 @@ export class Engine {
     this.routes.push({ route: new Route<TParams>(path), handler });
   }
 
-  async handle(input: HandlerInput): Promise<IHandlerOutput> {
+  async handle(input: HandlerInput): Promise<EngineOutput> {
     const url = new URL(input.getUrl());
     const path = url.pathname + url.search + url.hash;
     for (const route of this.routes) {

packages/sdk/src/jsx.ts

@@ -0,0 +1,35 @@
+// eslint-disable-next-line @typescript-eslint/no-namespace
+export namespace JSX {
+  export type Element = string;
+  export interface IntrinsicElements {
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    [elemName: string]: any;
+  }
+}
+
+export function createElement(
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  name: any,
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  props: { [id: string]: any },
+  ...content: string[]
+) {
+  if (typeof name === 'string') {
+    props = props || {};
+    const propsstr = Object.keys(props)
+      .map((key) => {
+        const value = props[key];
+        if (key === 'className') return `class=${value}`;
+        else return `${key}=${value}`;
+      })
+      .join(' ');
+
+    return content.length === 0
+      ? `<${name} ${propsstr}/>`
+      : `<${name} ${propsstr}>${content.join('')}</${name}>`;
+  } else if (typeof name === 'function') {
+    return name(props, ...content);
+  } else {
+    return content.join('');
+  }
+}

packages/sdk/src/lib.ts

@@ -1,3 +1,38 @@
 import { Engine } from './engine';
 
-export { Engine };
+import {
+  EngineParseError,
+  NoHandlerFoundError,
+  TxtDotError,
+} from './types/errors';
+
+import {
+  EngineFunction,
+  EngineMatch,
+  Engines,
+  RouteValues,
+  EnginesMatch,
+  HandlerInput,
+  HandlerOutput,
+  Route,
+  handlerSchema,
+} from './types/handler';
+
+import * as JSX from './jsx';
+
+export {
+  Engine,
+  EngineParseError,
+  NoHandlerFoundError,
+  TxtDotError,
+  EngineFunction,
+  EngineMatch,
+  Engines,
+  RouteValues,
+  EnginesMatch,
+  HandlerInput,
+  HandlerOutput,
+  Route,
+  handlerSchema,
+  JSX,
+};

packages/sdk/src/types/handler.ts

@@ -4,7 +4,7 @@ import { Engine } from '../engine';
 export class HandlerInput {
   private data: string;
   private url: string;
-  private dom?: Window;
+  private window?: Window;
 
   constructor(data: string, url: string) {
     this.data = data;
@@ -15,19 +15,26 @@ export class HandlerInput {
     return this.url;
   }
 
-  parseDom(): Window {
-    if (this.dom) {
-      return this.dom;
+  get document(): Document {
+    if (this.window) {
+      return this.window.document;
     }
 
-    this.dom = parseHTML(this.data);
-    return this.dom;
+    this.window = parseHTML(this.data);
+    return this.window.document;
   }
 }
 
-export interface IHandlerOutput {
+export interface HandlerOutput {
   content: string;
   textContent: string;
+  title: string;
+  lang: string;
+}
+
+export interface EngineOutput {
+  content: string;
+  textContent?: string;
   title?: string;
   lang?: string;
 }
@@ -66,7 +73,7 @@ export interface RouteValues {
 export type EngineFunction<TParams extends RouteValues> = (
   input: HandlerInput,
   ro: Route<TParams>
-) => Promise<IHandlerOutput>;
+) => Promise<EngineOutput>;
 
 export type EnginesMatch<TParams extends RouteValues> = EngineMatch<TParams>[];
 

packages/server/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@txtdot/server",
-  "version": "1.7.0",
+  "version": "1.8.0",
   "private": true,
   "description": "txtdot is an HTTP proxy that parses only text, links and pictures from pages reducing internet bandwidth usage, removing ads and heavy scripts",
   "main": "dist/app.js",

packages/server/src/distributor.ts

@@ -5,11 +5,11 @@ import { Readable } from 'stream';
 import { NotHtmlMimetypeError } from './errors/main';
 import { decodeStream, parseEncodingName } from './utils/http';
 import replaceHref from './utils/replace-href';
-import { parseHTML } from 'linkedom';
 
 import { Engine } from '@txtdot/sdk';
-import { HandlerInput, IHandlerOutput } from '@txtdot/sdk/dist/types/handler';
+import { HandlerInput, HandlerOutput } from '@txtdot/sdk';
 import config from './config';
+import { parseHTML } from 'linkedom';
 
 interface IEngineId {
   [key: string]: number;
@@ -32,7 +32,7 @@ export class Distributor {
     requestUrl: URL, // proxy URL
     engineName?: string,
     redirectPath: string = 'get'
-  ): Promise<IHandlerOutput> {
+  ): Promise<HandlerOutput> {
     const urlObj = new URL(remoteUrl);
 
     const webder_url = config.env.third_party.webder_url;
@@ -52,6 +52,7 @@ export class Distributor {
     }
 
     const engine = this.getFallbackEngine(urlObj.hostname, engineName);
+
     const output = await engine.handle(
       new HandlerInput(
         await decodeStream(data, parseEncodingName(mime)),
@@ -59,15 +60,28 @@ export class Distributor {
       )
     );
 
+    const dom = parseHTML(output.content);
+
     // post-process
     // TODO: generate dom in handler and not parse here twice
-    const dom = parseHTML(output.content);
-    replaceHref(dom, requestUrl, new URL(remoteUrl), engineName, redirectPath);
+    replaceHref(
+      dom.document,
+      requestUrl,
+      new URL(remoteUrl),
+      engineName,
+      redirectPath
+    );
 
-    const purify = DOMPurify(dom.window);
-    output.content = purify.sanitize(dom.document.toString());
+    const purify = DOMPurify(dom);
+    const content = purify.sanitize(dom.document.toString());
 
-    return output;
+    return {
+      content,
+      textContent:
+        output.textContent || dom.document.documentElement.textContent || '',
+      title: output.title || dom.document.title,
+      lang: output.lang || dom.document.documentElement.lang,
+    };
   }
 
   getFallbackEngine(host: string, specified?: string): Engine {

packages/server/src/errors/handler.ts

@@ -2,7 +2,7 @@ import { FastifyReply, FastifyRequest } from 'fastify';
 import { NotHtmlMimetypeError } from './main';
 import { getFastifyError } from './validation';
 
-import { TxtDotError } from '@txtdot/sdk/dist/types/errors';
+import { TxtDotError } from '@txtdot/sdk';
 
 import { IGetSchema } from '../types/requests/browser';
 import config from '../config';

packages/server/src/errors/main.ts

@@ -1,5 +1,5 @@
 import config from '../config';
-import { TxtDotError } from '@txtdot/sdk/dist/types/errors';
+import { TxtDotError } from '@txtdot/sdk';
 
 export class LocalResourceError extends TxtDotError {
   constructor() {

packages/server/src/types/requests/api.ts

@@ -2,7 +2,7 @@ import { FastifySchema, FastifyRequest } from 'fastify';
 import { IApiError, errorResponseSchema } from '../../errors/api';
 import { engineList } from '../../plugin_manager';
 import { FromSchema } from 'json-schema-to-ts';
-import { handlerSchema } from '@txtdot/sdk/dist/types/handler';
+import { handlerSchema } from '@txtdot/sdk';
 
 export interface IApiResponse<T> {
   data?: T;

packages/server/src/utils/replace-href.ts

@@ -2,13 +2,12 @@ import config from '../config';
 import { generateParserUrl, generateProxyUrl } from './generate';
 
 export default function replaceHref(
-  dom: Window,
+  doc: Document,
   requestUrl: URL,
   remoteUrl: URL,
   engine?: string,
   redirectPath: string = 'get'
 ) {
-  const doc: Document = dom.window.document;
   const parserUrl = (href: string) =>
     generateParserUrl(requestUrl, remoteUrl, href, engine, redirectPath);