It also ensures that imported code doesn't need to be loaded again for every template in the application. This makes it relatively easy to support all of JavaScript and JSON and anything else you might want to import in one go. maybe an option to choose between them.

decide whether this just transpiles to a require call, or whether it inlines the imported code.

add an 'ImportStatement' node in the parser.

add a lexer token for 'import statement' - possibly just use 'babylon' to parse the entire statement?.

This is not a small amount of work, but if you are willing to do that, I will definitely be happy to review/merge pull requests, and offer help where I can. At the moment the data fetching looks like: My main discomfort is the long path to the json. I can’t just include the json, because I need the json in the JS part to select and modify the strings. I readFile them because require is cached for the server runtime and I modify the Files on runtime. just give as a local and you have the might of everything in the template.Īt the beginning of a mixin I readFile the json. Now the only object we make available for the pug pages is require. I don’t know the available names of the variables and the jsons are located far away from the rest of the component. Often texts are coming from jsons and in the first iterations die jsons got required in the express part and given as locale variables to the pug.īut that meens you have a lot of ghost like global variables in every page.

Some gulp tasks watch over all the files, generate the css, js and browsersyncs. When developing we have a express server generating the html. We are building a design system and styleguide with pug.

Get array length in pug template