Completely customizable file monitoring using states, allowing file changes to be detected between script runs.
A slightly different breed of file monitoring for NodeJS. The state of all monitored files will be stored and compared to the current state on the next run to detect changes. It is heavily inspired by the file change detection formerly used by the Android Gradle Plugin for incremental task runs.
Install with npm
npm i file-state-monitor --save
require and use it in your code:
const FileMonitor = FileMonitor;const LastModifiedState = LastModifiedStatelet monitor = LastModifiedState;let stateFile = '/path/to/states.json';let loaded = monitor;monitor;let changedFiles = monitor;monitor;
On the first run this will give you a
Map of all files under the given path with a state of
created. If you run the script again you will get a list of
deleted files. When no files changed an empty
Map will be returned.
The API is pretty much self explanatory. You first create a new
FileMonitor and pass one of the available change detection strategies. After that
load() the previous state from disk and add directories or files you want to monitor with
monitorPath(). Calling this will compare the files with their previous state and you can get a list of changed files with
getChangedFiles(). To persist the state back to disk simply call
write() on the file monitor instance.
Visit https://janvennemann.github.io/file-state-monitor/?api for a complete API documentation.
Change detection strategies
This library uses special file state classes to let you choose which method you want to use to detect file changes. By subclassing the
BaseFileState you can also define your own change detection strategy, giving you complete control about what files you consider as changed. File state implementations that come bundled with the library:
LastModifiedState: Uses the last modified timestamp to detect if a file changed
SizeState: Uses the file size to detect if a file changed
ContentHashState: Computes a SHA-1 hash of the file's content and uses that hash to detect if a file changed
CombinedState: Uses all of the above checks in series, marking a file changed as soon as the first check returns true. Starts with the inexpensive checks for modification time and file size, and only then does the expensive content hash check.
Write your own state
Need a more sophisticated check other than only checking file modification time, size or content hash? No problem, just create you own state class that extends from BaseFileState. For a proper implementation you are required to at least define your own
toJson methods. Everything else is up to you. Take a look at the bundles states to get an idea of how it works.