From fd05c4d162c4e076ae00d66febb7b0169b8550d4 Mon Sep 17 00:00:00 2001 From: Peter Johnson Date: Wed, 1 Dec 2021 11:40:57 +0100 Subject: [PATCH] feat(bom): strip byte-order-mark from CSV files (#495) --- lib/streams/recordStream.js | 14 +++++++------- package.json | 2 +- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/lib/streams/recordStream.js b/lib/streams/recordStream.js index 8d3fafbed..8c0db064b 100644 --- a/lib/streams/recordStream.js +++ b/lib/streams/recordStream.js @@ -1,13 +1,12 @@ -const fs = require( 'fs' ); -const path = require( 'path' ); - -const csvParse = require( 'csv-parse' ); -const combinedStream = require( 'combined-stream' ); -const _ = require( 'lodash' ); +const _ = require('lodash'); +const fs = require('fs'); +const path = require('path'); +const csvParse = require('csv-parse').parse; +const combinedStream = require('combined-stream'); const through = require('through2'); const split = require('split2'); -const logger = require( 'pelias-logger' ).get( 'openaddresses' ); +const logger = require('pelias-logger').get('openaddresses'); const config = require('pelias-config').generate(); const CleanupStream = require('./cleanupStream'); @@ -100,6 +99,7 @@ function fileStreamDispatcher(stream, filePath) { } return stream.pipe(csvParse({ + bom: true, trim: true, skip_empty_lines: true, relax_column_count: true, diff --git a/package.json b/package.json index 4e88040b1..e28a0054f 100644 --- a/package.json +++ b/package.json @@ -11,7 +11,7 @@ "async": "^3.1.0", "bottleneck": "^2.19.5", "combined-stream": "^1.0.7", - "csv-parse": "^4.0.0", + "csv-parse": "^5.0.3", "fs-extra": "^8.1.0", "glob": "^7.0.0", "lodash": "^4.16.0",