@zoom7/cachebust
v0.0.1
Published
A simple cache buster
Downloads
5
Readme
cachebust
Straight forward asset fingerprinting "cachebuster" for production applications.
Requires Node 8.5.0 or higher.
Using cachebust
Installing
npm install @sparkbox/cachebust
The recommended way to use cachebust
is to either add your
input data in package.json
or in a .cachebust.config
file (see config).
Running
cachebust
: will create your fingerprinted assests & update fingerprinted "source file" names inside the
target "template file(s)" as specified in the package.json file or the .cachebust.config file.
If you need to restore your template files, you can run: cachebust --restore
Config
Globbing is allowed for target
files. So "target": ["path/to/target/**/*.html"]
is allowed.
Package.json
Create a key in your package.json
file at the root level, with source/target keys.
For example:
{
...
"cachebust": {
"source": [
"path/to/source/css/style.css",
"path/to/source/js/script.js"
],
"target": [
"path/to/target/template.html"
]
}
...
}
Config File
Also you may create a config file in the root of the project directory called .cachebust.config
This will be a JSON object like the following:
{
"cachebust": {
"source": [
"path/to/source/css/style.css",
"path/to/source/js/script.js"
],
"target": [
"path/to/target/**/*.html"
]
}
}
Direct Input
cachebust allows you to directly specify the source and the target you'll need to pass in a comma seperated list for example:
cachebust -s path/to/source/css/style.css,path/to/source/js/script.js -t path/to/target/template.html
Note Direct input will override the config and package json cacehebust paramaters.
This will fingerprint the following files:
- path/to/source/css/style.css
- path/to/source/js/script.js
This will then update your template "target" file with the updated filenames:
- path/to/target/template.html
To restore these you'll have to again pass the target and source files:
cachebust -r -s path/to/source/css/style.css,path/to/source/js/script.js -t path/to/target/template.html
Running on production
** WARNING ** If run this without a backup, it is descructive. It will not backup your files.
Added a -n
or --no-backup
option to cachebust, this will no generate a manifest file. See https://github.com/sparkbox/cachebust/issues/9
In some cases this is desired, say on a build server, or in a git versioned repo, there is no reason to keep a manifest since each deploys a copy of the original is retained from the source control.
CLI Options
You can run cachebust --help
for more options.
-V, --version output the version number
-s, --source [files] source file(s) to be fingerprinted; comma seperated file list
-t, --target [files] target file(s), template files that need the fingerprinted asset file names; comma seperated file list
-r, --restore copies the backup file(s) back to the original; backup file(s) are removed.
-q, --quiet', All normal output is surpressed. Some errors will still be generated.
-n, --no-backup Warning: this will not generate the manifest file. This is only to be used on PRODUCTION servers or under version control. This is descructive!
-h, --help output usage information
Ensure that the file lists for -s
or --source
and -t
and --target
are comma seperated.
-s dir/path/file.ext,dir/otherpath/otherfile.ext
etc.
You can pass in one file (no comma, or mutliple files for both source and target).
You may also specify a single file, without a comma.
License
The MIT License (MIT)