Merge pull request #59 from observerly/feature/fits/FITS

Author: michealroberts

samples/RosettaNebula/Rosetta_Nebula_[Ha]_Monochrome_M_300s_2024-11-26T17_20_00Z.fits

@@ -0,0 +1,226 @@
+/*****************************************************************************************************************/
+
+// @author         Michael Roberts <michael@observerly.com>
+// @package        @observerly/fits
+// @license        Copyright © 2021-2025 observerly
+
+/*****************************************************************************************************************/
+
+import { afterAll, beforeAll, describe, expect, it } from 'vitest'
+
+import { ImageData, createCanvas } from 'canvas'
+import * as fs from 'node:fs'
+import * as path from 'node:path'
+
+import { FITS, ZScaleInterval } from '../..'
+
+/*****************************************************************************************************************/
+
+const filename = 'Rosetta_Nebula_[Ha]_Monochrome_M_300s_2024-11-26T17_20_00Z.fits'
+
+/*****************************************************************************************************************/
+
+const input = path.join(__dirname, `../../../samples/RosettaNebula/${filename}`)
+
+/*****************************************************************************************************************/
+
+const output = path.join(__dirname, '../../../output')
+
+/*****************************************************************************************************************/
+
+let file: File
+
+/*****************************************************************************************************************/
+
+beforeAll(() => {
+  // Read the FITS file and create a File object
+  const fileBuffer = fs.readFileSync(input)
+  file = new File([fileBuffer], 'sample.fit', { type: 'application/octet-stream' })
+})
+
+/*****************************************************************************************************************/
+
+afterAll(() => {
+  // Clean up
+  // fs.rmdirSync(output, { recursive: true })
+})
+
+/*****************************************************************************************************************/
+
+describe('FITS', () => {
+  it('should be defined', () => {
+    expect(FITS).toBeDefined()
+  })
+
+  it('should create an instance of FITS', () => {
+    const fits = new FITS()
+    expect(fits).toBeInstanceOf(FITS)
+  })
+
+  it('should read a FITS file from a File object', async () => {
+    const fits = new FITS()
+    await fits.readFromFile(file)
+    expect(fits.getHeaders()).toBeInstanceOf(Map)
+  })
+
+  it('should read a FITS file from a Blob object', async () => {
+    const blob = new Blob([file], { type: 'application/octet-stream' })
+    const fits = new FITS()
+    await fits.readFromBlob(blob, filename)
+    expect(fits.getHeaders()).toBeInstanceOf(Map)
+  })
+
+  it('should read a FITS file from an ArrayBuffer', async () => {
+    const buffer = await file.arrayBuffer()
+    const fits = new FITS()
+    await fits.readFromBuffer(buffer, filename)
+    expect(fits.getHeaders()).toBeInstanceOf(Map)
+  })
+
+  it('should correctly parse headers from a FITS file', async () => {
+    const fits = new FITS()
+    await fits.readFromFile(file)
+    const headers = fits.getHeaders()
+
+    expect(headers).toBeInstanceOf(Map)
+    expect(headers.size).toBeGreaterThan(0)
+
+    expect(headers.get('SIMPLE')).toEqual({
+      key: 'SIMPLE',
+      value: true,
+      comment: 'FITS Standard 4.0'
+    })
+
+    expect(headers.get('BITPIX')).toEqual({
+      key: 'BITPIX',
+      value: -32,
+      comment: 'Number of bits per data pixel'
+    })
+
+    expect(headers.get('NAXIS')).toEqual({
+      key: 'NAXIS',
+      value: 2,
+      comment: '[1] Number of array dimensions'
+    })
+
+    expect(headers.get('NAXIS1')).toEqual({
+      key: 'NAXIS1',
+      value: 1463,
+      comment: '[1] Length of data axis 1'
+    })
+
+    expect(headers.get('NAXIS2')).toEqual({
+      key: 'NAXIS2',
+      value: 1168,
+      comment: '[1] Length of data axis 2'
+    })
+
+    expect(headers.get('BSCALE')).toEqual({
+      key: 'BSCALE',
+      value: 1,
+      comment: ''
+    })
+
+    expect(headers.get('BZERO')).toEqual({
+      key: 'BZERO',
+      value: 0,
+      comment: ''
+    })
+  })
+
+  it('should correctly destroy the FITS instance', async () => {
+    const fits = new FITS()
+    await fits.readFromFile(file)
+    fits.destroy()
+    expect(fits.getHeaders()).toBeInstanceOf(Map)
+    expect(fits.getHeaders().size).toBe(0)
+    expect(fits.width).toBe(0)
+    expect(fits.height).toBe(0)
+    expect(fits.getImageHDU()).toBeInstanceOf(Float32Array)
+    expect(fits.getImageHDU().length).toBe(0)
+    expect(fits.bitpix).toBe(0)
+    expect(fits.bzero).toBe(0)
+    expect(fits.bscale).toBe(1)
+  })
+})
+
+/*****************************************************************************************************************/
+
+describe('FITS', () => {
+  it('should parse FITS from file, retrieve normalized data, and write a JPEG', async () => {
+    // Read the file from disk into a Buffer:
+    const fitsBuffer = fs.readFileSync(input)
+    const file = new File([fitsBuffer], 'test.fits')
+    const fits = await new FITS().readFromFile(file)
+    expect(fits.width).toBeGreaterThan(0)
+    expect(fits.height).toBeGreaterThan(0)
+
+    // Now, getImageHDU() returns a Float32Array of normalized [0..255] data:
+    let data = fits.getImageHDU()
+
+    // Compute the zscale interval for normalization giving us vmin and vmax:
+    const { vmin, vmax } = ZScaleInterval(data)
+
+    const resolution = fits.width * fits.height
+
+    // Normalize the data to the [0..255] range
+    const normalizedData = new Float32Array(resolution)
+
+    for (let i = 0; i < resolution; i++) {
+      normalizedData[i] = ((data[i] - vmin) / (vmax - vmin)) * 255
+    }
+
+    data = normalizedData
+
+    // For consistency, confirm we have width*height pixels:
+    const width = fits.width
+    const height = fits.height
+    const pixels = width * height
+    expect(data.length).toBe(pixels)
+
+    // Create a canvas and 2D context to paint the ImageData:
+    const canvas = createCanvas(width, height)
+    const ctx = canvas.getContext('2d')
+    if (!ctx) {
+      throw new Error('Could not get 2D context from node-canvas.')
+    }
+
+    // Build an RGBA buffer from the normalized grayscale data:
+    const rgba = new Uint8ClampedArray(pixels * 4)
+    for (let i = 0; i < pixels; i++) {
+      // Already scaled to [0..255] by getImageData():
+      const gray = Math.floor(data[i])
+
+      const idx = i * 4
+      rgba[idx + 0] = gray // R
+      rgba[idx + 1] = gray // G
+      rgba[idx + 2] = gray // B
+      rgba[idx + 3] = 255 // A
+    }
+
+    // Paint the ImageData on the canvas at (0, 0):
+    const imageData = new ImageData(rgba, width, height)
+    ctx.putImageData(imageData, 0, 0)
+
+    // Convert to JPEG buffer with 90% quality:
+    const image = canvas.toBuffer('image/jpeg', { quality: 0.9 })
+
+    // Ensure the output directory exists before writing the JPEG file:
+    if (!fs.existsSync(output)) {
+      fs.mkdirSync(output, { recursive: true })
+    }
+
+    // Write the final JPEG file to disk:
+    const outputPath = path.join(output, 'output.jpg')
+
+    // Convert Buffer to Uint8Array for fs.writeFile():
+    const uint8Array = new Uint8Array(image.buffer, image.byteOffset, image.byteLength)
+    fs.writeFileSync(outputPath, uint8Array)
+
+    // Confirm the output file is not empty:
+    const stats = fs.statSync(outputPath)
+    expect(stats.size).toBeGreaterThan(0)
+  })
+})
+
+/*****************************************************************************************************************/