unblocker

Unblocker was originally a web proxy for evading internet censorship, similar to CGIproxy / PHProxy / Glype but written in node.js. It's since morphed into a general-purpose library for proxying and rewriting remote webpages.

All data is processed and relayed to the client on the fly without unnecessary buffering, making unblocker one of the fastest web proxies available.

The magic part

The script uses "pretty" urls which, besides looking pretty, allow links with relative paths to just work without modification. (E.g. <a href="path/to/file2.html"></a>)

In addition to this, links that are relative to the root (E.g. <a href="/path/to/file2.html"></a>) can be handled without modification by checking the referrer and 307 redirecting them to the proper location in the referring site. (Although the proxy does attempt to rewrite these links to avoid the redirect.)

Cookies are proxied by adjusting their path to include the proxy's URL, and a bit of extra work is done to ensure they remain intact when switching protocols or subdomains.

Limitations

Although the proxy works well for standard login forms and even most AJAX content, OAuth login forms and anything that uses postMessage (Google, Facebook, etc.) are not likely to work out of the box. This is not an insurmountable issue, but it's not one that I expect to have fixed in the near term. Patches are welcome, including both a general-purpose fix to go into the main library, and site-specific fixes to go in the examples folder.

Running the website on your computer

See https://github.com/nfriedly/nodeunblocker.com

Using unblocker as a library in your software

npm install --save unblocker

Unblocker exports an express-compatible API, so using in an express application is trivial:

var express = require('express')
var Unblocker = require('unblocker');
var app = express();

// this must be one of the first app.use() calls and must not be on a subdirectory to work properly
app.use(new Unblocker({prefix: '/proxy/'}));

app.get('/', function(req, res) {
    //...
});

Usage without express is similarly easy, see examples/simple/server.js for an example.

Configuration

Unblocker supports the following configuration options, defaults are shown:

{
    prefix: '/proxy/',  // Path that the proxied URLs begin with. '/' is not recommended due to a few edge cases.
    host: null, // Host used in redirects (e.g `example.com` or `localhost:8080`). Default behavior is to determine this from the request headers.
    requestMiddleware: [], // Array of functions that perform extra processing on client requests before they are sent to the remote server. API is detailed below.
    responseMiddleware: [], // Array of functions that perform extra processing on remote responses before they are sent back to the client. API is detailed below.
    standardMiddleware: true, // Allows you to disable all built-in middleware if you need to perform advanced customization of requests or responses.
    processContentTypes: [ // All  built-in middleware that modifies the content of responses limits itself to these content-types.
        'text/html',
        'application/xml+xhtml',
        'application/xhtml+xml',
        'text/css'
    ],
    httpAgent: null, //override agent used to request http response from server. see https://nodejs.org/api/http.html#http_class_http_agent
    httpsAgent: null //override agent used to request https response from server. see https://nodejs.org/api/https.html#https_class_https_agent
}

Custom Middleware

Unblocker "middleware" are small functions that allow you to inspect and modify requests and responses. The majority of Unblocker's internal logic is implimented as middleware, and it's possible to write custom middleware to augment or replace the built-in middleware.

Custom middleware should be a function that accepts a single data argument and runs synchronously.

To process request and response data, create a Transform Stream to perform the processing in chunks and pipe through this stream. (Example below.)

To respond directly to a request, add a function to config.requestMiddleware that handles the clientResponse (a standard http.ServerResponse when used directly, or a Express Response when used with Express. Once a response is sent, no further middleware will be executed for that request. (Example below.)

requestMiddleware

Data example:

{
    url: 'http://example.com/',
    clientRequest: {request},
    clientResponse: {response},
    headers: {
        //...
    },
    stream: {ReadableStream of data for PUT/POST requests, empty stream for other types}
}

requestMiddleware may inspect the headers, url, etc. It can modify headers, pipe PUT/POST data through a transform stream, or respond to the request directly. If you're using express, the request and response objects will have all of the usual express goodies. For example:

function validateRequest(data) {
    if (!data.url.match(/^https?:\/\/en.wikipedia.org\//) {
        data.clientResponse.status(403).send('Wikipedia only.');
    }
}
var config = {
    requestMiddleware: [
        validateRequest
    ]
}

If any piece of middleware sends a response, no further middleware is run.

After all requestMiddleware has run, the request is forwarded to the remote server with the (potentially modified) url/headers/stream/etc.

responseMiddleware

responseMiddleware receives the same data object as the requestMiddleware, but the headers and stream fields are replaced with those of the remote server's response, and several new fields are added for the remote request and response:

Data example:

{
    url: 'http://example.com/',
    clientRequest: {request},
    clientResponse: {response},
    remoteRequest {request},
    remoteResponse: {response},
    contentType: 'text/html',
    headers: {
        //...
    },
    stream: {ReadableStream of response data}
}

For modifying content, create a new stream and then pipe data.stream to it and replace data.stream with it:

var Transform = require('stream').Transform;

function injectScript(data) {
    if (data.contentType == 'text/html') {

        // https://nodejs.org/api/stream.html#stream_transform
        var myStream = new Transform({
            decodeStrings: false,
            function(chunk, encoding, next) {
                chunk = chunk.toString.replace('</body>', '<script src="/my/script.js"></script></body>');
                this.push(chunk);
                next();
        });

        data.stream = data.stream.pipe(myStream);
    }
}

var config = {
    responseMiddleware: [
        injectScript
    ]
}

See examples/nodeunblocker.com/app.js for another example of adding a bit of middleware. Also, see any of the built-in middleware in the lib/ folder.

Built-in Middleware

Most of the internal functionality of the proxy is also implemented as middleware:

host: Corrects the host header in outgoing responses
referer: Corrects the referer header in outgoing requests
cookies: Fixes the Path attribute of set-cookie headers to limit cookies to their "path" on the proxy (e.g. Path=/proxy/http://example.com/). Also injects redirects to copy cookies from between protocols and subdomains on a given domain.
hsts: Removes Strict-Transport-Security headers because they can leak to other sites and can break the proxy.
hpkp: Removes Public-Key-Pinning headers because they can leak to other sites and can break the proxy.
csp: Removes Content-Security-Policy headers because they can leak to other sites and can break the proxy.
redirects: Rewrites urls in 3xx redirects to ensure they go through the proxy
decompress: Decompresses Content-Encoding: gzip|deflate responses and also tweaks request headers to ask for either gzip-only or no compression at all. (It will attempt to decompress deflate content, but there are some issues, so it does not advertise support for deflate.)
charsets: Converts the charset of responses to UTF-8 for safe string processing in node.js. Determines charset from headers or meta tags and rewrites all headers and meta tags in outgoing response.
urlPrefixer: Rewrites URLS of links/images/css/etc. to ensure they go through the proxy
metaRobots: Injects a ROBOTS: NOINDEX, NOFOLLOW meta tag to prevent search engines from crawling the entire web through the proxy.
contentLength: Deletes the content-length header on responses if the body was modified.

Setting the standardMiddleware configuration option to false disables all built-in middleware, allowing you to selectively enable, configure, and re-order the built-in middleware.

This configuration would mimic the defaults:

var Unblocker = require('unblocker');

var config = {
    prefix: '/proxy/',
    host: null,
    requestMiddleware: [],
    responseMiddleware: [],
    standardMiddleware: false,  // disables all built-in middleware
    processContentTypes: [
        'text/html',
        'application/xml+xhtml',
        'application/xhtml+xml'
    ]
}

var host = Unblocker.host(config);
var referer = Unblocker.referer(config);
var cookies = Unblocker.cookies(config);
var hsts = Unblocker.hsts(config);
var hpkp = Unblocker.hpkp(config);
var csp = Unblocker.csp(config);
var redirects = Unblocker.redirects(config);
var decompress = Unblocker.decompress(config);
var charsets = Unblocker.charsets(config);
var urlPrefixer = Unblocker.urlPrefixer(config);
var metaRobots = Unblocker.metaRobots(config);
var contentLength = Unblocker.contentLength(config);

config.requestMiddleware = [
    host,
    referer,
    decompress.handleRequest,
    cookies.handleRequest
    // custom requestMiddleware here
];

config.responseMiddleware = [
    hsts,
    hpkp,
    csp,
    redirects,
    decompress.handleResponse,
    charsets,
    urlPrefixer,
    cookies.handleResponse,
    metaRobots,
    // custom responseMiddleware here
    contentLength
];

app.use(new Unblocker(config));

Debugging

Unblocker is fully instrumented with debug. Enable debugging via environment variables:

DEBUG=unblocker:* node mycoolapp.js

There is also a middleware debugger that adds extra debugging middleware before and after each existing middleware function to report on changes. It's included with the default DEBUG activation and may also be selectively enabled:

DEBUG=unblocker:middleware node mycoolapp.js

... or disabled:

DEBUG=*,-unblocker:middleware node mycoolapp.js

Troubleshooting

If you're using Nginx as a reverse proxy, you probably need to disable merge_slashes to avoid endless redirects and/or other issues:

merge_slashes off;

Todo

Consider adding compress middleware to compress text-like responses
Un-prefix urls in GET / POST data
Inject js to proxy postMessage data and fix origins
More examples
Even more tests

AGPL-3.0 License

This project is released under the terms of the GNU Affero General Public License version 3.

All source code is copyright Nathan Friedly.

Commercial licensing and support are also available, contact Nathan Friedly (nathan@nfriedly.com) for details.

LCYLYM/node-unblocker