diff --git a/index.js b/index.js index fab32e2..3aa0dbc 100644 --- a/index.js +++ b/index.js @@ -36,8 +36,15 @@ app.use( app.get('/', (req, res) => { res.send('Hello There!'); }) +function addProtocolToUrl(url) { + if (!url.startsWith('http://') && !url.startsWith('https://')) { + url = 'https://' + url; + } + return url; +} function parseUrlWithoutHashAndQuery(fullUrl) { + fullUrl = addProtocolToUrl(fullUrl); const parsedUrl = new URL(fullUrl); // Set the hash and search/query to empty strings @@ -93,29 +100,27 @@ app.post('/hash', async (req, res) => { try { let { urls } = req.body; if (!urls) { - return res.status(400).json({ error: 'Missing URL in the request parameters' }); + return res.status(400).json({ error: 'Missing in the request parameters' }); } if (!Array.isArray(urls)) urls = [urls]; const promises = urls.map(async (url) => { const urlWithoutHashAndQuery = parseUrlWithoutHashAndQuery(url); - console.log(url, `Fetching and hashing ${urlWithoutHashAndQuery}`); const hashedContent = await fetchAndHashContent(urlWithoutHashAndQuery); const fileHash = await hashContent(Buffer.from(hashedContent, 'utf-8')); - return { urls, fileHash }; + return { url, fileHash }; }); let results = await Promise.all(promises); - results = results.reduce((acc, { urls, fileHash }) => { - acc[urls] = fileHash; + results = results.reduce((acc, { url, fileHash }) => { + acc[url] = fileHash; return acc; }, {}); res.json(results); } catch (error) { - console.error('Error:', error.message); - res.status(500).json({ error: 'Internal Server Error' }); + res.status(500).json({ error: error.message }); } }); @@ -135,7 +140,7 @@ app.post('/download-repos', async (req, res) => { let { urls } = req.body; if (!urls) { - return res.status(400).json({ error: 'Missing urls in the request parameters' }); + return res.status(400).json({ error: 'Missing in the request parameters' }); } if (!Array.isArray(urls)) { urls = [urls]; @@ -167,8 +172,7 @@ app.post('/download-repos', async (req, res) => { // Pipe the zip file to the response archive.pipe(res); } catch (error) { - console.error('Error:', error.message); - res.status(500).json({ error: 'Internal Server Error' }); + res.status(500).json({ error: error.message }); } }); diff --git a/index.min.js b/index.min.js index 0aeea67..ae7e2e0 100644 --- a/index.min.js +++ b/index.min.js @@ -1 +1 @@ -require("dotenv").config();const express=require("express"),cors=require("cors"),axios=require("axios"),{createHash:createHash}=require("crypto"),archiver=require("archiver"),rateLimit=require("express-rate-limit"),{parse:parseUrl,URL:URL}=require("url"),{parse:parseHtml}=require("node-html-parser"),allowedDomains=process.env.ALLOWED_DOMAINS.split(","),app=express();app.use(cors());const port=process.env.PORT||3e3,host=process.env.HOST||"0.0.0.0";function parseUrlWithoutHashAndQuery(fullUrl){const parsedUrl=new URL(fullUrl);parsedUrl.hash="",parsedUrl.search="";return parsedUrl.toString()}async function hashContent(content){const hash=createHash("sha256");return hash.update(content),hash.digest("hex")}async function fetchAndHashContent(url,visitedUrls=new Set){if(visitedUrls.has(url))return"";visitedUrls.add(url);const content=(await axios.get(url,{responseType:"arraybuffer",timeout:1e4})).data.toString("utf-8"),linkedResources=parseHtml(content).querySelectorAll('link[rel="stylesheet"], script[src]');return`${content}_${(await Promise.all(linkedResources.map((async resource=>{const resourceUrl=parseUrl(resource.getAttribute("href")||resource.getAttribute("src"),!0);let absoluteResourceUrl=resourceUrl.href;resourceUrl.hostname||(resourceUrl.path.startsWith("/")||url.endsWith("/")||(url+="/"),absoluteResourceUrl=`${url}${resourceUrl.path}`);const resourceContent=await fetchAndHashContent(absoluteResourceUrl,visitedUrls);return`${resourceUrl.path}_${resourceContent}`})))).join("_")}`}async function downloadGitHubRepo(owner,repo){if(!owner||!repo)throw new Error("Missing owner or repo");const zipUrl=`https://github.com/${owner}/${repo}/archive/refs/heads/master.zip`;return(await axios.get(zipUrl,{responseType:"arraybuffer"})).data}app.use(express.json()),app.use(rateLimit({windowMs:6e4,max:10})),app.get("/",((req,res)=>{res.send("Hello There!")})),app.post("/hash",(async(req,res)=>{try{let{urls:urls}=req.body;if(!urls)return res.status(400).json({error:"Missing URL in the request parameters"});Array.isArray(urls)||(urls=[urls]);const promises=urls.map((async url=>{const urlWithoutHashAndQuery=parseUrlWithoutHashAndQuery(url);console.log(url,`Fetching and hashing ${urlWithoutHashAndQuery}`);const hashedContent=await fetchAndHashContent(urlWithoutHashAndQuery),fileHash=await hashContent(Buffer.from(hashedContent,"utf-8"));return{urls:urls,fileHash:fileHash}}));let results=await Promise.all(promises);results=results.reduce(((acc,{urls:urls,fileHash:fileHash})=>(acc[urls]=fileHash,acc)),{}),res.json(results)}catch(error){console.error("Error:",error.message),res.status(500).json({error:"Internal Server Error"})}})),app.post("/download-repos",(async(req,res)=>{try{let{urls:urls}=req.body;if(!urls)return res.status(400).json({error:"Missing urls in the request parameters"});Array.isArray(urls)||(urls=[urls]);const archive=archiver("zip");res.attachment("repos.zip");const downloadPromises=urls.map((async url=>{const[owner,name]=url.split("/").slice(-2);if(!owner||!name)return void console.error(`Invalid url format: ${url}`);const zipBuffer=await downloadGitHubRepo(owner,name);archive.append(zipBuffer,{name:`${owner}-${name}.zip`})}));await Promise.all(downloadPromises),archive.finalize(),archive.pipe(res)}catch(error){console.error("Error:",error.message),res.status(500).json({error:"Internal Server Error"})}})),app.listen(port,host,(()=>{console.log(`Server is running at http://${host}:${port}`)})),module.exports=app; \ No newline at end of file +require("dotenv").config();const express=require("express"),cors=require("cors"),axios=require("axios"),{createHash:createHash}=require("crypto"),archiver=require("archiver"),rateLimit=require("express-rate-limit"),{parse:parseUrl,URL:URL}=require("url"),{parse:parseHtml}=require("node-html-parser"),allowedDomains=process.env.ALLOWED_DOMAINS.split(","),app=express();app.use(cors());const port=process.env.PORT||3e3,host=process.env.HOST||"0.0.0.0";function addProtocolToUrl(url){return url.startsWith("http://")||url.startsWith("https://")||(url="https://"+url),url}function parseUrlWithoutHashAndQuery(fullUrl){fullUrl=addProtocolToUrl(fullUrl);const parsedUrl=new URL(fullUrl);parsedUrl.hash="",parsedUrl.search="";return parsedUrl.toString()}async function hashContent(content){const hash=createHash("sha256");return hash.update(content),hash.digest("hex")}async function fetchAndHashContent(url,visitedUrls=new Set){if(visitedUrls.has(url))return"";visitedUrls.add(url);const content=(await axios.get(url,{responseType:"arraybuffer",timeout:1e4})).data.toString("utf-8"),linkedResources=parseHtml(content).querySelectorAll('link[rel="stylesheet"], script[src]');return`${content}_${(await Promise.all(linkedResources.map((async resource=>{const resourceUrl=parseUrl(resource.getAttribute("href")||resource.getAttribute("src"),!0);let absoluteResourceUrl=resourceUrl.href;resourceUrl.hostname||(resourceUrl.path.startsWith("/")||url.endsWith("/")||(url+="/"),absoluteResourceUrl=`${url}${resourceUrl.path}`);const resourceContent=await fetchAndHashContent(absoluteResourceUrl,visitedUrls);return`${resourceUrl.path}_${resourceContent}`})))).join("_")}`}async function downloadGitHubRepo(owner,repo){if(!owner||!repo)throw new Error("Missing owner or repo");const zipUrl=`https://github.com/${owner}/${repo}/archive/refs/heads/master.zip`;return(await axios.get(zipUrl,{responseType:"arraybuffer"})).data}app.use(express.json()),app.use(rateLimit({windowMs:6e4,max:10})),app.get("/",((req,res)=>{res.send("Hello There!")})),app.post("/hash",(async(req,res)=>{try{let{urls:urls}=req.body;if(!urls)return res.status(400).json({error:"Missing in the request parameters"});Array.isArray(urls)||(urls=[urls]);const promises=urls.map((async url=>{const urlWithoutHashAndQuery=parseUrlWithoutHashAndQuery(url),hashedContent=await fetchAndHashContent(urlWithoutHashAndQuery);return{url:url,fileHash:await hashContent(Buffer.from(hashedContent,"utf-8"))}}));let results=await Promise.all(promises);results=results.reduce(((acc,{url:url,fileHash:fileHash})=>(acc[url]=fileHash,acc)),{}),res.json(results)}catch(error){res.status(500).json({error:error.message})}})),app.post("/download-repos",(async(req,res)=>{try{let{urls:urls}=req.body;if(!urls)return res.status(400).json({error:"Missing in the request parameters"});Array.isArray(urls)||(urls=[urls]);const archive=archiver("zip");res.attachment("repos.zip");const downloadPromises=urls.map((async url=>{const[owner,name]=url.split("/").slice(-2);if(!owner||!name)return void console.error(`Invalid url format: ${url}`);const zipBuffer=await downloadGitHubRepo(owner,name);archive.append(zipBuffer,{name:`${owner}-${name}.zip`})}));await Promise.all(downloadPromises),archive.finalize(),archive.pipe(res)}catch(error){res.status(500).json({error:error.message})}})),app.listen(port,host,(()=>{console.log(`Server is running at http://${host}:${port}`)})),module.exports=app; \ No newline at end of file