README

Command line tool to Analyze and store your Media file's metadata.

Do you or your organization have a bunch of images/videos lumped into some hard drive? If those are stored in a non-structured way, like there is no real scheme on how it is organized (by date, by album etc). Performing analysis and retrieving a certain set of files will be really tough to do.

Scalp allows you analyze and extract a lot of information from those bunch of media files and build "something" out of that data. It is purely a backend tool that is designed to support any kind of application you might be developing on that top of the extracted info.

Few situations it could be useful:

Build a private Search engine, that allows user to input certain keywords and return a list of files matching that keyword
Use as a backend store for your CMS
Generate thumbnails (small/medium/large) and host the processed files to CDN

Below is the overview of how Scalp works

Your media files can be stored in the same server or a dedicated NAS server (needs to be mounted to a location where scalp can access it )
Scalp reads the files and extract Metadata from it (represented as JSON object)
Scalp generates thumbnails (small/medium/large)
The data processed can then be forwarded to any of the supported backend storage (File, ElasticSearch, S3, BackBlaze etc)
Your application accesses the processed data via those backend storage

Advantages

You can continously re-organize how the files is presented to your users without having to physically move around the actual files. Since the representation can be abstracted by the Application itself
Your raw images/videos is left untouched (Scalp never move, resize or modify its properties)
Saves a lot of bandwidth, since instead of serving the raw images while browsing items. You can just use the thumbnails created by Scalp. and only access the raw (usually big file) image when user requested.

Requirements

Linux-based Server (Debian/RHEL etc)
PHP 5.4 or greater

Install

It is very easy to install Scalp as a CLI utility:
via wget

wget https://downloads.buonzz.com/scalp.phar
sudo mv scalp.phar  /usr/local/bin/scalp
chmod +x /usr/local/bin/scalp

via curl

curl -o scalp.phar 'https://downloads.buonzz.com/scalp.phar'
sudo mv scalp.phar  /usr/local/bin/scalp
chmod +x /usr/local/bin/scalp

After this, scalp command is available anywhere in your computer. To check if the scalp is installed properly, just execute

scalp -V

This should output

Scalp Metadata Extraction Tool by Darwin Biler version v2

via Composer - Globally

You can install scalp globally in your machine:

composer global require 'buonzz/scalp=dev-master'

Simply add this directory to your PATH in your ~/.bash_profile (or ~/.bashrc) like this:

export PATH=~/.composer/vendor/bin:$PATH

via Composer - per-project

just require the buonzz/scalp in your composer project

{
"require":{
    "buonzz/scalp": "1.*"
  }
}

Usage

Scalp can extract the metadata information from your media files and export it to following:

Static JSON files
Thumbnails

First, you need to create a configuration file called .env This will be used by scalp to retrieve certain information:

Metadata Extraction

To generate static JSON files

scalp metadata:extract

Sample extracted metadata

{
   "last_modified":"2017-03-25T12:10:39+00:00",
   "last_accessed":"2017-04-09T12:07:11+00:00",
   "file_permissions":"0644",
   "date_indexed":"2017-04-10T03:04:23+00:00",
   "human_filesize":"974.83 kB",
   "filepath":"IMG_1123.JPG",
   "path_tags":[

   ],
   "filesize":"998225",
   "fileformat":"jpg",
   "filename":"IMG_1123.JPG",
   "mime_type":"image\/jpeg",
   "exif":{
      "DateTimeDigitized":"2016-05-17T16:19:26+00:00",
      "ExposureTime":0.016666666666667,
      "FNumber":4,
      "ISOSpeedRatings":200,
      "ShutterSpeedValue":6,
      "ApertureValue":4,
      "FocalLength":20
   },
   "date_tags":[
      "Tue",
      "17th",
      "Tuesday",
      "May",
      "May",
      "2016",
      "4pm",
      "UTC",
      "17"
   ],
   "file_contents_hash":"4c7e796bc250b14fe7964694c4db5852eca34ddee24991371f848c3e8097436d",
   "width":"1920",
   "height":"1280"
}

Thumbnail Creation

scalp thumbnail:create

Running in the background

If you got a relatively large collection of media files. It might take hours for the process to complete. You can place the process in the background, so that it will continue to execute even after you had logged out in the terminal.

nohup scalp metadata:extract > scalp.log &
nohup scalp thumbnail:create > scalp.log &

Backend Storage

The processed data can be stored in multitude of ways. Allowing you to freely use its data in whatever purpose it may serve for your use-case.

File

The simpliest backend is the file storage. It simply dumps everything to whatever you had defined in the OUTPUT_FOLDER configuration in your .env file. You can either use the generated files and host it to your web server/CDN or pass it to your application for further processing.

ElasticSearch

When using ElasticSearch as a backend. Make sure to provide the following entries in your .env file

Load it to ElasticSearch

scalp es:index

Save thumbnails to ElasticSearch

scalp thumb:save

Viewing the thumbnails

You can get a preview of those thumbnails by using the PHP's built in web server

php -S localhost:8080 /usr/local/bin/scalp

then in your localhost, just append the url given by search command.

BackBlaze's B2 Cloud Storage

B2 is a great way to store both your metadata and thumbnails if you plan to serve those data in a web site with high traffic. It is way cheaper than the other alternative, plus it offers you a free 10GB initial storage so you don't really need to pay anything unless you decide to upload more contents.

Provide the following entries in your .env file prior to using the BackBlaze

Then execute the following to extract, create thumbnail and upload the data to B2:

scalp b2:upload scalp-test dist

the first parameter (scalp-test) is the bucket name you created in your backblaze account, and the second parameter is the folder where the files to be uploaded is located.

AWS S3

TODO

buonzz / scalp

Maintainers

Details