Often we save settings or configurations in json
files.
If you've spent some time using NodeJS, you'd know that there are quite a few
ways to read json files in NodeJS
Two of the most common ones are : https://gist.github.com/c9536d535c31c4c07ae0d4192086f97b
The other way is using the fs
module.
We can either do it synchronously https://gist.github.com/243832e0a1246cba12f7d6b7c1ae843f
Or we can do it asynchronously -
https://gist.github.com/8d40b176603cfde343f9a47a68229303
Now obviously the question that comes to mind is which method to use, and if there are any obvious benefits to any one method
Let us discuss the differences -
require()
will cache the file in the require graph.
So during the lifetime of the node app, if the file.json
is changed,
you will not get the new data, even if you re-run require('./file.json')
On the other hand fs.readFile
or fs.readFileSync
will always
re-read the file, and pick up changes.
In require
you cannot define the file encoding. 99% of the times, that is
not a problem.
Nevertheless if your json is not encoded in Unicode/UTF-8, you'd have to use
fs.readFile
as it supports encodings such as ascii or latin1 or even
base64 (yes, yes, I hear you, no one saves JSON as base64)
require()
is synchronous, and hence blocking in nature.
fs
provides both sync and async methods as shown above. If you want to
read your JSON file without blocking, then fs.readFile
is your only option.
NOTE: From NodeJS 10.x, import file from './file.json'
would be possible
because of support for modules
, which would allow async reading of json files
without fs
I hope you'll be able to take an informed decision in your future projects based on the points discussed here.