Round-trip tests

README.md

@@ -6,7 +6,7 @@
 [![minzipped](https://img.shields.io/bundlephobia/minzip/hyparquet-writer)](https://www.npmjs.com/package/hyparquet-writer)
 [![workflow status](https://github.com/hyparam/hyparquet-writer/actions/workflows/ci.yml/badge.svg)](https://github.com/hyparam/hyparquet-writer/actions)
 [![mit license](https://img.shields.io/badge/License-MIT-orange.svg)](https://opensource.org/licenses/MIT)
-![coverage](https://img.shields.io/badge/Coverage-95-darkred)
+![coverage](https://img.shields.io/badge/Coverage-96-darkred)
 [![dependencies](https://img.shields.io/badge/Dependencies-1-blueviolet)](https://www.npmjs.com/package/hyparquet-writer?activeTab=dependencies)
 
 Hyparquet Writer is a JavaScript library for writing [Apache Parquet](https://parquet.apache.org) files. It is designed to be lightweight, fast and store data very efficiently. It is a companion to the [hyparquet](https://github.com/hyparam/hyparquet) library, which is a JavaScript library for reading parquet files.

eslint.config.js

@@ -43,7 +43,7 @@ export default [
       'jsdoc/require-returns-type': 'error',
       'jsdoc/sort-tags': 'error',
       'no-constant-condition': 'off',
-      'no-extra-parens': 'error',
+      'no-extra-parens': 'warn',
       'no-multi-spaces': 'error',
       'no-trailing-spaces': 'error',
       'no-undef': 'error',

test/write.roundtrip.test.js

@@ -0,0 +1,34 @@
+import fs from 'fs'
+import { asyncBufferFromFile, parquetMetadataAsync, parquetReadObjects, parquetSchema } from 'hyparquet'
+import { describe, expect, it } from 'vitest'
+import { parquetWriteBuffer } from '../src/index.js'
+
+describe('parquetWrite round-trip', () => {
+  const files = fs.readdirSync('test/files').filter(f => f.endsWith('.parquet'))
+
+  files.forEach(filename => {
+    it(`round-trips data from ${filename}`, async () => {
+      const file = await asyncBufferFromFile(`test/files/${filename}`)
+      const metadata = await parquetMetadataAsync(file)
+      const rows = await parquetReadObjects({ file })
+
+      // transpose the row data
+      const schema = parquetSchema(metadata)
+      const columnData = schema.children.map(({ element }) => ({
+        ...element,
+        data: /** @type {any[]} */ ([]),
+      }))
+      for (const row of rows) {
+        for (const { name, data } of columnData) {
+          data.push(row[name])
+        }
+      }
+
+      const buffer = parquetWriteBuffer({ columnData })
+      const output = await parquetReadObjects({ file: buffer })
+
+      expect(output.length).toBe(rows.length)
+      expect(output).toEqual(rows)
+    })
+  })
+})