some webpack-built files are named with their file hash while others aren't

[whatisgalen]

A common use case is that a developer has updated their project and pushes these updates to a production server (run webpack build, collectstatic, etc.) but end-users' browsers are still caching the old files. There's a cache-bursting strategy built into the logic of how webpack names these files:

resolve({
            entry: {
                ...archesCoreEntryPointConfiguration,
                ...archesApplicationsEntrypointConfiguration,
                ...projectEntryPointConfiguration,
                ...CSSFilepathLookup,
            },
            devServer: {
                port: WEBPACK_DEVELOPMENT_SERVER_PORT,
            },
            output: {
                path: Path.resolve(__dirname, APP_ROOT, 'media', 'build'),
                publicPath: STATIC_URL,
                filename: '[name].[contenthash].js', // Adds a hash to all entry points
                chunkFilename: '[id].[contenthash].js', // Adds a hash to all dynamically imported chunks
                clean: true,
                assetModuleFilename: 'img/[hash][ext][query]',
            },
            ...

The contenthash is used in creating the name of each file, capturing its updated state which then forces browsers to get the new files from the server. However when npm run build_production gets run, only so-called chunkFiles (.chunk.js) get their content hash mixed into their file names. Files like search.js or graph.js do not contain any hash values in the filenames, despite the explicit logic above in the webpack config.