Tommy: Web Assets converter and optimizer

Tommy will process and optimize all your static assets ready for the web, no matter what the file extension happens to be; and, to avoid wasting time, it will also save processed items so that unmodified assets will not be processed further.

Tommy also gives you the option of syncing all your processed assets to a S3 bucket.

Pass runtime options

• --src specify the source directory (where your assets are located)
• --dst specify the destination directory (where your assets will be generated)
• --force regenerate all assets ignoring cache
• --config specify a JSON file containing an extension to the configuration

Configuration

If you pass to Tommy an additional JSON file, it will be merged with initial config.

tommy --config config.json

Example:

{
  "version": "v1",
  "processor": {
    "resize": true,
    "poster": false
  },
  "resize": {
    "dimensions": [300, 600, 1200]
  },
  "ignore": [".dockerignore"]
}

First-level keys

Key	Type	Applicable to	Description	Default
version	string	-	Version of the configuration	v1
ignore	string[]	-	Pattern to ignore	see config.json
remoteSync	bool	-	On/Off sync to remote bucket	false
s3Bucket	bool	-	S3 Bucket name	null
processor.resize	bool	jpg/png images	On/Off generation of various resized images	true
processor.image	bool	jpg/png images	On/Off imagemagick processor	true
processor.poster	bool	videos	On/Off generation of poster image	true
processor.jpg	bool	jpg images	On/Off jpegoptim optimizer	true
processor.png	bool	png images	On/Off pngquant optimizer	true
processor.gif	bool	gif images	On/Off gifsicle optimizer	true
processor.svg	bool	svg images	On/Off svgo optimizer	true
processor.lazyLoadBlurried	bool	jpg/png images	On/Off generation of a blurry image	true
processor.videoThumbs	bool	videos	On/Off generation of thumbnails extracted from video	true
converter.mp4	bool	videos	On/Off conversion to MP4	true
converter.webm	bool	videos	On/Off conversion to WEBM	true
converter.webp	bool	jpg/png images	On/Off conversion to WEBP	true

resize

Key	Type	Description	Default
resize.dimensions	number[]	Dimensions of resized images in PX (longest side)	[200,400,800,1200]
resize.quality	number	Quality of images	80

image

Key	Type	Description	Default
image.quality	number	Quality of image	80

videoThumbs

Key	Type	Description	Default
videoThumbs.fps	string	Specify FPS of shoots (1/60 = every 1 minute)	1/30

lazyLoadBlurried

Key	Type	Description	Default
lazyLoadBlurried.size	number	Length of longest side	10

How to: run with Docker

I suggest you to use with Docker to avoid installing all dependencies in your host.

docker run \
-v "$(pwd)/volumes/src":/src \
-v "$(pwd)/volumes/dst":/dst \
-v "$(pwd)/config.json:/root/config.json" \
kopiro/tommy \
--src /src \
--dst /dst \
--config /root/config.json

Explanation of mounts:

• -v "$(pwd)/volumes/src":/src mount the source directory (where your assets are located) into container /src
• -v "$(pwd)/volumes/dst":/dst mount the destination directory (where your assets will be generated) into container /dst
• -v "$(pwd)/config.json:/root/config.json" mount your (optional) configuration JSON file into container /root/config.json

How to: run in MacOS

Upfront, install dependencies by running macos.sh, then install with NPM:

npm -g i tommy

Then run as a binary:

tommy --src ./volumes/src --dst ./volumes/dst

How: build locally Docker image

Download the repository, then run ./test-docker.sh.

It will build the Docker image locally and test with volumes present in current repository.

Configure AWS credentials

To use the remote sync feature, you need to configure the binary aws to locale credentials.

When using with docker, you can just share your .awscredentials file through a volume in the path /root/.aws/credentials.

Otherwise, you can configure your aws via aws configure.

Example file:

[default]
aws_access_key_id=...
aws_secret_access_key=...

To mount:

docker run
...
-v "$(pwd)/.awscredentials":"/root/.aws/credentials"
...

License

MIT