fetch

Fetch URL contents

Usage no npm install needed!

<script type="module">
  import fetch from 'https://cdn.skypack.dev/fetch';
</script>

README

fetch

Fetch url contents. Supports gzipped content for quicker download, redirects (with automatic cookie handling, so no eternal redirect loops), streaming and piping etc.

Build Status

Install

npm install fetch

Usage

See examples folder for a complete example

Fetch from URL

fetch.fetchUrl(url [, options], callback)

Where

  • url is the url to fetch
  • options is an optional options object
  • callback is the callback to run - callback(error, meta, body)

Example

var fetchUrl = require("fetch").fetchUrl;

// source file is iso-8859-15 but it is converted to utf-8 automatically
fetchUrl("http://kreata.ee/iso-8859-15.php", function(error, meta, body){
    console.log(body.toString());
});

NB If the file has been marked with charset other than utf-8, it is converted automatically.

By default iconv-lite is used for charset conversion. If you want to use node-iconv module instead, add "iconv": "*" to your package.json file, it will be picked up by fetch automatically.

Streaming

fetch.FetchStream(url [, options]) -> Stream

Where

  • url is the url to fetch
  • options is an optional options object

With events:

  • data with a data chunk - function(chunk){}
  • meta with some information about the response function(meta){}
  • end when the receiving is ready
  • error

Example

var FetchStream = require("fetch").FetchStream;

var fetch = new FetchStream("http://google.com");

fetch.on("data", function(chunk){
    console.log(chunk);
});

Options

Possible option values

  • maxRedirects how many redirects allowed, defaults to 10
  • disableRedirects set to true if redirects are not allowed, defaults to false
  • headers optional header fields, in the form of {'Header-Field':'value'}
  • maxResponseLength maximum allowd length for the file, the remainder is cut off. Defaults to Infinity
  • method defaults to GET
  • payload request body
  • disableGzip set to false, to disable content gzipping, needed for Node v0.5.9 which has buggy zlib
  • cookies an array of cookie definitions in the form of ['name=val']
  • cookieJar for sharing cookies between requests, see below
  • outputEncoding valid for fetchUrl
  • disableDecoding valid for fetchUrl, set to true to disable automatic charset decoding to utf-8
  • overrideCharset valid for fetchUrl, set input encoding
  • asyncDnsLoookup use high performance asyncronous DNS resolution based on c-ares instead of a thread pool calling getaddrinfo(3)
  • timeout set a timeout in ms
  • agentHttps pass-through http.request agent parameter for https
  • agentHttp pass-through http.request agent parameter for http
  • agent pass-through http.request agent parameter as fallback, if agentHttps or agentHttp are not specified
  • rejectUnauthorized whether to reject self-signed certificates (true, default behavior), or ignore and allow them (false)

Meta object

Meta object contains following fields:

  • status HTTP status code
  • responseHeaders response headers
  • finalUrl last url value, useful with redirects
  • redirectCount how many redirects happened
  • cookieJar CookieJar object for sharing/retrieving cookies

Headers

Request headers can be set with options.headers

options = {
    headers:{
        "X-My-Header": "This is a custom header field"
    }
}

User-Agent

User-Agent value can be set with options.headers['User-Agent'] value. Defaults to "FetchStream"

options = {
    headers: {
        "User-Agent": "MyUseragent/1.0"
    }
}

Cookies

Cookies can be set with options.cookies which takes an array with cookie definitions

options = {
    cookie: ["name=value", "key=value; path=/; secure"]
}

Paths, domain, expire and other cookie settings are honored, so try not to set cookies with expire dates in the past. If domain is not set, any domain will pass, same for paths.

NB Do not set cookie field directly in request header as it will be overwritten.

Cookie sharing

Cookies can be shared between different requests, this can be achieved with CookieJar

var fetch = require("fetch");

var cookies = new fetch.CookieJar();

// add one cookie for testing
cookies.setCookie('alfa=beta; path=/;');

// create a FetchStream with custom CookieJar
var f = fetch.FetchStream("http://www.example.com/page1",{cookieJar: cookies});

f.on("end", function(){
    // if cookies were set with the previos request, the data is
    // saved in 'cookieJar' and passed to the next request
    fetch.FetchStream("http://www.example.com/page1",{cookieJar: cookies});
});

Redirects

Redirects are on by default, use options.disableRedirects to disable. Maximum redirect count can be set with options.maxRedirects (defaults to 10)

options = {
    disableRedirects: true
}

options = {
    maxRedirects: 100
}

Disable Gzip support

Gzip and Deflate support is automatically on. This is problematic in Node v0.5.9 and below since Zlib support on these versions is buggy with unpacking and tends to yield in error.

options = {
    disableGzip: true
}

Piping to file

FetchStream is a readable Stream object and thus can be piped. For example stream URL contents directly to a file:

var FetchStream = require("fetch").FetchStream,
    fs = require("fs"),
    out;

out = fs.createWriteStream('file.html');
new FetchStream("http://www.example.com/index.php").pipe(out);

License

BSD