A simple addon to the excellent bfj module which provides a simple parser for very large JSON files in the form of collections.
Assumptions:
- Your JSON input is huge (otherwise you would just read it in via
JSON.parse
) - Your data is in the form of a collection (i.e. an array of objects)
- You can hold each individual object in that collection in memory but not the whole input data
While BFJ is excellent at working with very large JSON files sometimes you just want to be able to read all items in a collection without having to do the parsing yourself.
The below example demonstrates how to read in a very large JSON file and emits bfjc
every time we can parse an object within the collection:
bfjc(fs.createReadStream('someBigFile.json'))
.on('bfjc', data => ... do something with the object entity ...)
.on(bfj.events.end, ()=> ... we have finished reading ...)
The main exported function functions exactly the same as bfj.walk and takes exactly the same input stream and options. See that functions documentation for more details.
Additional options are listed below:
Option | Type | Default | Description |
---|---|---|---|
pause |
boolean |
true |
Instruct the stream to pause before each bfjc emit event |
allowArrays |
boolean |
false |
Allow nested array objects e.g. [['One'], [['Two']]] |
allowScalars |
boolean |
false |
Allow scalar types (strings, numebers, booleans) as object types |
NOTES:
- The
pause
functionality assumes the events are syncronous not asyncronous. If you wish to wrap pausing in promises or other event driven functionality you will have to add your own code after disablingpause