key-file-storage
Simple key-value storage (a persistent data structure) directly on file system, maps each key to a separate file.
- Simple key-value storage model
- Very easy to learn and use
- Both Synchronous and Asynchronous APIs
- One JSON containing file per each key
- Built-in configurable cache
- Both Promise and Callback support
const kfs = require("key-file-storage")('my/storage/path')
// Write something to file 'my/storage/path/myfile'
kfs.myfile = { x: 123 }
// Read contents of file 'my/storage/path/myfile'
const x = kfs.myfile.x
// Delete file 'my/storage/path/myfile'
delete kfs.myfileA very nice alternative for any of these node modules: node-persist, configstore, flat-cache, conf, simple-store and more...
Installation
Installing package on Node.js:
$ npm install key-file-storageInitialization
Initializing a key-file storage:
const keyFileStorage = require("key-file-storage")
const kfs = keyFileStorage('/storage/directory/path', caching)The value of caching can be
-
true(By default, if not specified) : Unlimited cache, anything will be cached on memory, good for small data volumes. -
false: No cache, read the files from disk every time, good when other applications can modify the files' contents arbitrarily. -
n(An integer number) : Limited cache, only thenlatest referred key-values will be cached, good for large data volumes where only a fraction of data is being used frequently .
Usage
Synchronous API
As simple as native javascript objects:
kfs['key'] = value // Write filekfs['key'] // Read filedelete kfs['key'] // Delete filedelete kfs['*'] // Delete all storage files'key' in kfs // Check for file existence
//=> true or false-
You can use
kfs.keyNameinstead ofkfs['keyName']anywhere if the key name allows. -
undefinedis not supported as a savable value, butnullis. Saving a key with valueundefinedis equivalent to remove it. So, you can usekfs['key'] = undefinedor evenkfs['*'] = undefinedto delete files. -
Synchronous API will throw an exception if any errors happen, so you shall handle it your way.
Asynchronous API with Promises
Every one of the following calls returns a promise:
kfs('key', value) // Write filekfs('key') // Read filenew kfs('key') // Delete filenew kfs('*') /* or */
new kfs() /* or */
new kfs // Delete all storage files('key' in kfs(), kfs()) // Check for file existence
// Resolves to true or false- Once again,
undefinedis not supported as a savable value, butnullis. Saving a key with valueundefinedis equivalent to remove it. So, you can usekfs('key', undefined)or evenkfs('*', undefined)to delete files.
Asynchronous API with Callbacks
The same as asynchronous with promises, but with callback function as the last input parameter of kfs() :
kfs('key', value, cb) // Write filekfs('key', cb) // Read filenew kfs('key', cb) // Delete filenew kfs('*', cb) /* or */
new kfs(cb) // Delete all storage files'key' in kfs(cb) // Check for file existence
// without promise output
/* or */
('key' in kfs(), kfs(cb))
// Check for file existence
// with promise output-
These calls still return a promise on their output (except for
'key' in kfs(callback)form of existence check). -
The first input parameter of all callback functions is
err, so you shall handle it within the callback. Reading and Existence checking callbacks provide the return values as their second input parameter.
Folders as Collections
Every folder in the storage can be treated as a collection of key-values.
You can query the list of all containing keys (filenames) within a collection (folder) like this (Note that a collection path must end with a forward slash '/'):
Synchronous API
try {
const keys = kfs['col/path/']
// keys = ['col/path/key1', 'col/path/sub/key2', ... ]
} catch (error) {
// handle error...
}Asynchronous API with Promises
kfs('col/path/')
.then(keys => {
// keys = ['col/path/key1', 'col/path/sub/key2', ... ]
})
.catch(error => {
// handle error...
})Asynchronous API with Callbacks
kfs('col/path/', (error, keys) => {
if (error) {
// handle error...
}
// keys = ['col/path/key1', 'col/path/sub/key2', ... ]
})Notes
-
NOTE 1 : Each key will map to a separate file (using the key itself as its relative path). Therefore, keys may be relative paths, e.g:
'data.json','/my/key/01'or'any/other/relative/path/to/a/file'. The only exception is strings including'..'(double dot) which will not be accepted for security reasons. -
NOTE 2 : You may have hidden key files by simply add a
'.'before the filename in the key path. -
NOTE 3 : If a key's relative path ends with a forward slash
'/', it will be considered to be a collection (folder) name. So,'data/set/'is a collection and'data/set/key'is a key in that collection. -
NOTE 4 : This module has a built-in implemented cache, so, when activated, accessing a certain key more than once won't require file-system level operations again for that file.
-
NOTE 5 : When activated, caching will include queries on collections too.
Example
const keyFileStorage = require("key-file-storage")
// Locate 'db' folder in the current directory as the storage path,
// Require 100 latest accessed key-values to be cached:
const kfs = keyFileStorage('./db', 100)
// Create file './db/users/hessam' containing this user data, synchronously:
kfs['users/hessam'] = ({
name: "Hessam",
skills: {
java: 10,
csharp: 15
}
})
// Read file './db/users/hessam' as a JSON object, asynchronously:
kfs('users/hessam').then(hessam => {
console.log(`Hessam's java skill is ${hessam.skills.java}.`)
})
// Check whether file './db/users/mahdiar' exists or not, asynchronously:
'users/mahdiar' in kfs((error, exists) => {
if (exists) {
console.log("User Mahdiar exists!")
}
})
// List all the keys in './db/users/', synchronously:
const allUsers = kfs['users/']
//=> ['users/hessam', 'users/mahdiar', ... ]Contribute
It would be very appreciated if you had any suggestions or contribution on this repository or submitted any issue.
- See the code on GitHub
- Contact me by my gmail address (Hessamoddin A Shokravi)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent