I'm working on an angular application which requires html files to be extracted as plain HTML files and at same time should check for any <img src="..."> to require those images (as assets). In addition, images are based upon root path (so /images/something.png), they need to be resolved relatively to webpack context setting (the base path).
How can I achieve this? Can't get html-loader to play nicely with file-loader. Since the former outputs a JS file (with the require statements) and the latter expects a plain HTML file.
I found the solution by myself by digging a lot into source codes of the loaders in question.
Basically, it doesn't work by default because file-loader expects a "raw" file, so if you want to output an HTML file, you need to have html source, not a JS one. However, html-loader takes an HTML file and outputs a JS file (with require assets and the content).
The solution was this deeply hidden and fantastic extract-loader which parses the JS coming out of html-loader, converts it back to plain html, the assets are still required and replaced even with hash for cachebusting.
That's perfect, you pass the output to file-loader and you finally have your html files!
Example configuration
In my case, my loader configuration looks like:
'file-loader!extract-loader!html-loader' + '?root=' + encodeURIComponent(sourcePath.toString())