chunk-csv

chunk-csv

Splits a CSV read stream into multiple write streams or strings.

This library was forked from csv-split-stream, an extra method was added and the previous code was updated to support async functions within the callback functions. The original library hasn't been updated since 2017, so I decided to revive it. Feel free to submit a PR or issue containing any bug fixes or feature requests.

Install

npm install chunk-csv

Usage

Using write streams

• Split a local CSV file into multiple CSV files (10000 lines each, excluding the header row):

const chunkCsv = require('chunk-csv');

return chunkCsv.splitStream(
  fs.createReadStream('input.csv'),
  {
    lineLimit: 10000
  },
  (index) => fs.createWriteStream(`output-${index}.csv`)
)
.then(csvSplitResponse => {
  console.log('csv split succeeded!', csvSplitResponse);
  // outputs: {
  //  "totalChunks": 350,
  //  "options": {
  //    "delimiter": "\n",
  //    "lineLimit": "10000"
  //  }
  // }
}).catch(csvSplitError => {
  console.log('csv split failed!, csvSplitError);
});

• Download a large CSV file via HTTP, split it into chunks of 10000 lines and upload each of them to s3:

const http           = require('http'),
const chunkCsv = require('csv-split-stream');
const AWS            = require('aws-sdk'),
const s3Stream       = require('s3-upload-stream')(new AWS.S3());

function downloadAndSplit(callback) {
  http.get({...}, downloadStream => {
    chunkCsv.splitStream(
      downloadStream,
      {
        lineLimit: 10000
      },
      (index) => s3Stream.upload({
        Bucket: 'testBucket',
        Key: `output-${index}.csv`
      })
    )
    .then(csvSplitResponse => {
      console.log('csv split succeeded!', csvSplitResponse);
      callback(...);
    }).catch(csvSplitError => {
      console.log('csv split failed!', csvSplitError);
      callback(...);
    });
  });    
}

Using strings

• Split a local CSV file into multiple strings (10000 lines each, excluding the header row):

const chunkCsv = require('chunk-csv');

chunkCsv.split(
  fs.createReadStream('input.csv'),
  {
    lineLimit: 10000
  },
  async (chunk, index) => {
      const data = await neatCsv(chunk);
      console.log("Processed Chunk", index);
      console.log(data);
  }
)
.then(csvSplitResponse => {
  console.log('csv split succeeded!', csvSplitResponse);
  // outputs: {
  //  "totalChunks": 350,
  //  "options": {
  //    "delimiter": "\n",
  //    "lineLimit": "10000"
  //  }
  // }
}).catch(csvSplitError => {
  console.log('csv split failed!', csvSplitError);
});

Methods

splitStream(readable, options, callback(index))
The splitStream method splits a CSV readable stream into multiple write streams and takes 3 arguments.

• readable is a readable stream from your csv file.
• options object: delimiter (defaults to "\n"), lineLimit number of lines per chunk.
• callback(index) - for every chunk of csv this callback function will be executed. A stream that is writable must be returned from this callback (can be a write or read/write stream). The chunks of csv data will be piped to this write stream. The index argument is given which specifies the chunk number being processed.
  

split(readable, options, callback(chunk, index))
The split method splits a CSV readable stream into multiple, smaller strings.

• readable is a readable stream from your csv file.
• options object: delimiter (defaults to "\n"), lineLimit number of lines per chunk.
• callback(chunk, index) - for every chunk of csv this callback function will be executed. The chunk argument will be the raw csv data with the specified number of lines and index specifies the chunk number being processed.

Notes

This module will use the first row as a header, so make sure your CSV has a header row. Currently working on a solution to support a "no headers" option.