Have a logo and favicon for the project.

Maybe in the root webpage repo/submodule.

As @WybrenKoelmans suggested in his PR #59 , it should be possible to extract part of an image by passing the following parameters:

extract-top-x
extract-top-y
extract-bottom-x
extract-bottom-y

My experience with GraphicsMagic has been far better than with ImageMagik, at least on a windows machine. But I haven't given it too complex tasks.

http://blog.trackduck.com/2014/11/12/imagemagick-vs-graphicsmagick-one-better-process-images-website/

many say in the web, that GraphicsMagis is faster. I'm not sure if the API is the same in every aspect. Again my limited experience says that they use have the same API.

Add more badge to Readme file

• MIT license
• the latest stable version number

Add PHPCS to travis build file

After merging the last PR #105 the unit test is failing because of the new parameters.
to fix that it need to add new param in OutputImageTest.php

Currently only returning 500
Even if I try to exit with error code the framework seems to be catching me first.

public function saveNewFile($sourceFile, $newFileName, $options)
    {
        $newFilePath = TMP_DIR . $newFileName;
        try {
            $tmpFile = $this->saveToTemporaryFile($sourceFile);
        } catch (Exception $e) {
            http_response_code($e->getCode());
            exit($e->getMessage());
        }
        $commandStr = $this->generateCmdString($newFilePath, $tmpFile, $options);

        exec($commandStr, $output, $code);
        if (count($output) === 0) {
            $output = $code;
        } else {
            $output = implode(PHP_EOL, $output);
        }

        if ($code !== 0) {
            throw new \Exception($output . ' Command line: ' . $commandStr);
        }
        $this->filesystem->write($newFileName, stream_get_contents(fopen($newFilePath, 'r')));
        unlink($tmpFile);
        unlink($newFilePath);
    }

and then

public function saveToTemporaryFile($fileUrl)
    {
        //check restricted_domains is enabled
        if ($this->params['restricted_domains'] &&
            is_array($this->params['whitelist_domains']) &&
            !in_array(parse_url($fileUrl, PHP_URL_HOST), $this->params['whitelist_domains'])
        ) {
            throw  new \Exception('Restricted domains enabled, the domain your fetching from is not allowed: ' . parse_url($fileUrl, PHP_URL_HOST), 403);

        }

        if (!$resource = @fopen($fileUrl, "r")) {
            throw  new \Exception('Error occured while trying to read the file Url : ' . $fileUrl, 400);
        }
        $content = "";
        while ($line = fread($resource, 1024)) {
            $content .= $line;
        }
        $tmpFile = TMP_DIR . uniqid("", true);
        file_put_contents($tmpFile, $content);
        return $tmpFile;
    }

but it doesn't work

Enabling xDebug profiler shows too many called to parser function in the Yaml Symfony component.
We should replace it with Yaml pecl extension and see results.

Add more Unit test and increase Code Coverage

Fix CodeClimate Warning

In order to Enable Xdebug, we should add a documentation how to enable it and use it.

Clarify the mime-type output logic, currently a couple of cases are mixed

Using ab Apache (Apache HTTP server benchmarking tool) reveal concurrency issue with the app.
After checking, it appears like it's related to the hash generated to each image, in the case of refresh option it should add another random hash to image generated.

Update docs about the new feature added recently image extraction

Move array params to separate classes:

• defaultParams
• ParsedOptions

Fix the latest SensioLabsInsight warning.

Improve Documentation

Add Community links (Github and Twitter)

Current architecture needs to be refactored and improved.

• Split logic Image processing between resize and face detection.
• Separating the Image Entity class into 2 classes: InputImage, OutputImage
• Improve folders and classes naming.

Hey,

Didn't find this in the doc, but maybe it's already implemented ?

Goal of this RFC is to be able to use flyimg as a public hosting media (no domain restriction) but still provides security to avoid "bad" users to generate their owns image (like doing a loop with +1 pixel width) and doing a DDOS attack.

A way to achieve that is to have on the page creating the url to this service, and on flyimage a shared key whichs allows to encode all the parameters with the url into a signature, and check this on the server (if signature invalid throw a 403 / 404 ?). It's very similar on how thumbor works.

WDYT ?

More UnitTest should be added to the application.

Very good compression
https://jpeg.org/jpegxr/

Because for any production use there should be a caching layer in front of this service, like a CDN.

It could be useful later on when we add smart header handling, like eTags based on files checksum. But for now it's overkill.

In order to add more Security logic and checks, It'll be good to move the check restricted domains logic to a specific Hander: SecurityHandler.

Add Travis ci Test coverage

See this:
https://hacks.mozilla.org/2017/01/using-immutable-caching-to-speed-up-the-web/

Consider the cases where that should be added.

Consider using phantomJS to load a URL, render it, and return a part or all of the rendered page.

The part that would be selected should be passed as a css selector in the first parameter.

Something like this:

http://flyimg.io/render/.header__title--big/https://some .webpage.com

https://developers.google.com/speed/webp/
WebP is natively supported in Google Chrome and the Opera browser.
Todo: based on header on the request

If you request a resource to the server's URL it should match automatically to the stored file, where ever it may be.

clarification needed

It was left from the beginning , it's already private but it's still on the list of URL params when it should be a config level thing.

I would like to be able to resize mode with servo skuchevaya not it.

for example
https://dimains.com/upload/w_300,h_200/imagefolder/12541.jpg

in the current documentation is not seen as an opportunity to implement resize or tell me how to do it?

Improve Logging output:

• Remove access log from the container output
• Add ExceptionHandler in app.php
• Add Handler to output catched exception to Stderr

Consider guetzli JPEG Encoder
https://opensource.googleblog.com/2017/03/guetzli-new-open-source-jpeg-encoder.html
https://github.com/google/guetzli/

After merging #104 we need to add the proper documentation how to use signature generation feature.

Some of the important security headers are missing from the Response:
To check: https://schd.io/4kDA

Remove ParserTraitusage

Add UnitTest to cover the extract feature.

Currently we convert everything to jpg, if the input resource is PNG, the default output should be PNG.

Currently we only have a text that says "hello from docker"
Maybe make this a submodule project?

maybe with Zopfli for compression
more: https://ariya.io/2016/06/using-zopfli-to-optimize-png-images

The idea is to return a URL to the resource instead of an image.
Consider the case where the resource has been uploaded to s3 maybe give the amazon URL.
This may be configurable according to the storage selected.

Test performance under several conditions.

• Transformation performance
• Serving performance

how a bout this?

A Dockerized PHP Microservice application to resize and crop images on the fly. Get optimized images with MozJPEG, WebP or PNG using ImageMagick to do Face detection, cropping, Face Blurring, Rotation and many other options. Runs inside a Docker container for easy to deployment. Abstract storage based on FlySystem in order to store images on any provider (local, S3,...).

No tests written for it

When not requesting a refresh response should be cacheable, but we are sending: Cache-Control:no-cache with all responses.

As a bonus cache expiry times should be configurable. (In the params file maybe? )

In order to generate Changelog file we'll use this tool:
https://github.com/lalitkapoor/github-changes

GIF should be straightforward
for adding animated gif resizing check the following:

http://askubuntu.com/questions/257831/how-can-i-resize-an-animated-gif-file-using-imagemagick

• add tests for gif
• add tests for animated gif

In order to pptimize composer install when running the benchmark,
it should add --no-dev --optimize-autoloader to composer install

Hi,
i'm trying to make flyimg running without Docker (replace with Apache, and htaccess rewrite) but no luck. Any suggestion ?

Update composer