Remove getSiteFiles function to simplify and speed up the export process

[sejas]

Related issues

• Fixes STU-536
• Related to Follow symlink directories for Jetpack backups #1448

Proposed Changes

• Simplify and speed up the export logic by removing getBackupContents and getSiteFiles functions.

Testing Instructions

• Tests should pass
• Run npm start
• Go to import/export tab
• Click on export entire site
• Observe the produced file contains all the correct files

Pre-merge Checklist

• Have you checked for TypeScript, React or other console errors?

[epeicher]

Thanks for simplifying this and reducing the work required in the export process. It works as advertised. I have tested different exports, and I have not found any issues. I have also exported broken symlinks and no errors are returned. LGTM!

[epeicher]

I love it!

😄

[katinthehatsite]

What do you think about cleaning up these comments as they seem self-explanatory?

[katinthehatsite]

The functionality looks good and works as expected 👍

From what I understand, the logic that you are removing is now handled by:

	private addWpContent(): void {
		const categories = (
			[ 'uploads', 'plugins', 'themes', 'muPlugins', 'fonts' ] as BackupContentsCategory[]
		 ).filter( ( category ) => this.options.includes[ category ] );
		this.emit( ExportEvents.WP_CONTENT_EXPORT_START );
		for ( const category of categories ) {
			const folderName = category === 'muPlugins' ? 'mu-plugins' : category;
			const absolutePath = path.join( this.options.site.path, 'wp-content', folderName );
			const archivePath = path.relative( this.options.site.path, absolutePath );
			this.archive.directory( absolutePath, archivePath, ( entry ) => {
				const fullArchivePath = path.join( archivePath, entry.name );
				const isExcluded = this.pathsToExclude.some( ( pathToExclude ) =>
					fullArchivePath.startsWith( path.normalize( pathToExclude ) )
				);
				if (
					isExcluded ||
					entry.name.includes( '.git' ) ||
					entry.name.includes( 'node_modules' )
				) {
					return false;
				}
				return entry;
			} );
			this.emit( ExportEvents.WP_CONTENT_EXPORT_PROGRESS, { directory: absolutePath } );
		}
		this.emit( ExportEvents.WP_CONTENT_EXPORT_COMPLETE );
	}

Is that right?

[epeicher]

The functionality looks good and works as expected 👍

From what I understand, the logic that you are removing is now handled by:

	private addWpContent(): void {
(…)

Is that right?

Great question @katinthehatsite! Let me chime in with an answer. That is absolutely right, before the related changes, this function depended on backupContent.wpcomSites, but after those changes, the backupContent.wpcomSites was not required anymore (and the functions that populated it). We decided to leave the refactor as a follow-up to avoid mixing a refactor with new functionality.

[sejas]

The functionality looks good and works as expected 👍
From what I understand, the logic that you are removing is now handled by:

	private addWpContent(): void {
(…)

Is that right?

Great question @katinthehatsite! Let me chime in with an answer. That is absolutely right, before the related changes, this function depended on backupContent.wpcomSites, but after those changes, the backupContent.wpcomSites was not required anymore (and the functions that populated it). We decided to leave the refactor as a follow-up to avoid mixing a refactor with new functionality.

Yes, reading all the files and saving that list structure was slowly replaced over time. And the last push was when we replaced archive.file with archive.directory on #1448

My assumption is that Antony added the list of files in memory, thinking about future use for it in the progress bar, but that never happened.