README

Focus on your application, save your time & server resources by off-loading the heavy PDF generation tasks to us. ChromeHeadless.io will deliver the beautiful and error-free PDFs for your professional customer's invoice, data reports and more..

How it works?

Our service is provided through REST APIs so virtually it can be used with any language and system. For convenience, we are constructing the client library for each of languages such as PHP. NodeJs, .Net and Python. So if you are using those languages, you may install the client library and provide it with your secret token created from your registered account. The client library will compress your HTML together with resources such as CSS and Javascript and send over to our server farms to generate PDF version. The high definition PDF will be sent back to you.

Advantages

There are number of advantages of using our system:

1. No installation required: You do not have to install PhantomJS or Headless Chrome.
2. Off-load heavy task: Headless Chrome or PhantomJS required certain amount of CPUs and RAM which you may reserve for other crucial tasks. Beside your CPUs and RAM may not be optimized for this tasks which may affect the efficiency. Our system contains a farm of servers which is highly optimized so we can perform this tasks faster and better.
3. Avoid complicated interface: You may avoid unnecessarily complicated coding to control headless browser.
4. Everything works smoothly: We will take care to make sure that your PDFs generated nicely,all fonts are working so that you can spend more time concentrate on your application rather than be bugged with Headless Chrome issues.

Our advantages over other similar PDF cloud services

We have studied well other services before we decided to move on with our services, here are some of our advantages over them:

1. Some of them do not execute Javascript, only converting pure HTML to PDF but WE DO!
2. Many of them require your application to be online so that resources can be loaded. If your application is in localhost or intranet their solution will not work but WE DO!

Get Token Key

Token key is a secret string with 64 characters used to access our service. You need to attach your token in every request you make to our service.

Steps to generate token key:

1. If you have account with us, go to step 2 otherwise sign-up with us. An email with title "Welcome to ChromeHeadless.io" will be sent to you in few minutes after your sign up.
2. Use account credential in welcome email to log in our system.
3. Go to tokens management page
4. Hit Generate button to generate token key.

Installation

The PHP Client can be installed through composer

composer require chromeheadlessio/php-client

Example

<?php
//Use PHP Client Library
require_once "vendor/autoload.php";

//Create ChromeHeadless service with your token key specified
$service = new \chromeheadlessio\Service("my-token-key");

//Get PDF generated from html content and push it to browser
$service->export([
    "html"=>"Hello world!"
])->pdf([
    "format"=>"A4",
    "orientation"=>"portrait"
])->sendToBrowser("helloworld.pdf");

Exporting content

The export() method belongs the service class. It receives an array as parameter defining what you need to export. Below are list of properties:

Export to PDF

The pdf() method will help to generate pdf file. It takes an array as parameter defining options for your PDF. Below are available options.

Example:

$service->export(...)->pdf([
    "scale"=>1,
    "format"=>"A4",
    "landscape"=>true
])->sendToBrowser("myfile.pdf");

PDF options in view file

Some options could be set directly in the PDF view file instead of pdf() method.

header and footer

In the view file, use header and footer tags to set pdf's header and footer template:

Example:

<header>
    <div id="header-template" 
        style="font-size:10px !important; color:#808080; padding-left:10px">
        <span>Header: </span>
        {date}
        {title}
        {url}
        {pageNumber}
        {totalPages}
        <span id='pageNum' class="pageNumber"></span>
        <img src='http://www.chromium.org/_/rsrc/1438879449147/config/customLogo.gif?revision=3' />
    </div>
</header>
<footer>
...
</footer>

if either header or footer tag exists, pdf options' displayHeaderFooter will be true. PDF options' headerTemplate and footerTemplate options take priority over view file's header and footer tags. With header and footer tags, if there's no font-size style, a default style "font-size:10x" is used. Header and footer tags supports place holders like {date}, {title}, etc and img tag with link-type src. For img tag pdf options' headerTemplate and footerTemplate only support base64-type src.

margin

In the view file, use the body tag's margin style to set pdf margin:

Example:

//MyReportPDF.view.php
<body style='margin: 1in 0.5in 1in 0.5in'>
...
</body>

If either header or footer tag exists but there's no body's margin, a default margin of 1 inch will be used

Export to PNG

The png() help to generate PNG file. It take an array as parameter defining options for your PNG. Below are list of properties:

Example:

$service->export(...)->png([
    "clip"=>[
        "x"=>100,
        "y"=>100,
        "width"=>500,
        "height"=>1000,
    ]
])->sendToBrowser("myfile.png");

Export to JPG

The png() help to generate JPG file. It take an array as parameter defining options for your JPG. Below are list of properties:

Example:

$service->export(...)->jpg([
    "quality"=>80
    "clip"=>[
        "x"=>100,
        "y"=>100,
        "width"=>500,
        "height"=>1000,
    ]
])->sendToBrowser("myfile.jpg");

Getting result

In all above examples we use method sendToBrowser() to send the file to browser for user to open on browser or download as attachment. Here are all options:

Examples:

$service->export(...)->jpg([
    "quality"=>80
    "clip"=>[
        "x"=>100,
        "y"=>100,
        "width"=>500,
        "height"=>1000,
    ]
])->save("../img/myfile.jpg");

Resource cache (v1.10.0+)

Reports re-ship the same widget CSS/JS, fonts and chart libraries on every export. With the resource cache on, the client omits resources the server already holds and the server materializes them locally — a standard report export shrinks to just the HTML. It is opt-in and failure-safe: any cache-path problem falls back to a plain full-zip export, so it never breaks or blocks a render. It requires a cache-enabled service (RESOURCE_CACHE_ENABLED).

Enable it

$settings = [
    // ... your usual settings (serviceHost, token, html/url, ...) ...
    'resourceCache' => [
        'enabled' => true,
    ],
];
$report->run()->cloudExport($view)->settings($settings)->pdf($opts)->toBrowser($name);

serviceHost must point at the cache-enabled service — the client probes serviceHost/api/capabilities and only caches if it advertises support.

How the belief set works

A hash (of the asset's actual bytes — so a locally-edited asset never mismatches) is treated as "the server has this" if it is in any of:

BUNDLED   data/koolreport-hashset.json      ships warm; the KoolReport library
SYNCED    GET /api/cache/manifest (delta)   globally-shared hashes, off the hot path
SELF      X-Resource-Cached confirmations   this token's own confirmed uploads

Per export: believed assets are omitted (listed in the manifest); the server 409s with any it doesn't actually have; the client re-sends just those once. The server confirms what it now holds via X-Resource-Cached, which the client records into SELF so the next export omits them.

Staying current with the shared set (SYNCED)

BUNDLED is frozen at the client release — it only knows the library that existed when this version shipped. SYNCED keeps you current with everything the server has since made shared (custom assets that reached enough tenants to promote, newly pre-seeded library additions) without a client upgrade or re-bundle. It's a throttled delta pull:

  GET /api/cache/manifest?since=<cursor>
    ◄─ { added:[<hash>,…], cursor:<N> }      # a new client sends ?since=0 once,
                                             # then only asks for what's newer

It runs at most once per syncInterval (default daily), before the export is assembled, and is failure-safe — a down or slow manifest never blocks a render.

Round trips. The manifest GET and the export POST are deliberately separate requests (sync must not sit on the render path). So on the one export per day that a sync is due you pay two round trips; every other export that day is a single POST. It is not two round trips per export. Set sync = false to drop the GET entirely (rely on BUNDLED ∪ SELF), or syncInterval = 0 to sync on every export for a demo.

Example. You ship 1.10.0 to customer A; its bundle doesn't know a new widget asset sparkline.js. Later, three other customers each cache it, so the server promotes it to the shared pool. On A's next daily sync the client pulls that hash into SYNCED — and the next time A's report includes sparkline.js, the client omits it (0 bytes uploaded), even though A never bundled or uploaded it.

Caching your own custom resources

Library assets warm automatically from the bundled set. Your own CSS/JS are not in that set, so opt in per request with cacheCustom:

'resourceCache' => [
    'enabled'     => true,
    'cacheCustom' => ['scope' => 'tenant'], // or 'global'
],

• scope: 'tenant' — cache your custom assets privately for this token (pinned, survives eviction). The first export declares + ships them; from the second export they're omitted.
• scope: 'global' — force-share them into the pool so your other tokens/tenants dedupe immediately. Requires RESOURCE_ALLOW_CUSTOM_SHARE on the server (else it degrades to tenant); note this publishes the hashes — use only for non-secret assets.

Server warnings (global-share-disabled, pin-cap-exceeded) surface via $exporter->getWarnings().

Options

The bundled data/koolreport-hashset.json is generated by the export service's pre-seed pipeline. An empty bundle is valid — the client cold-starts and warms via SELF over successive exports instead. See the service's docs/resource-cache.md for the full protocol, deployment, and cache-cleaning.

About us

KoolPHP Inc has been in business for 10 years, we focus on building the featured rich yet easy-to-use components to help developers increase productivity and deliver highest quality applications within time and budget constraints. Our main products are KoolPHP UI and KoolReport. KoolPHP UI is a toolset for developer to construct web faster while KoolReport is an open-source reporting framework to build data reports and dashboard easier.