You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This plug-in appears to pass through all non-coffee files as is, while swapping out coffee files to JavaScript files. I found that surprising, and ended up chaining on another plug-in to strip out all but the JavaScript files.
Is the right idiom that a Broccoli build process should pass essentially all files through a series of plug-ins, each of which transparently passes along everything it doesn't care about? Or is the right idiom that one should use a bunch of plug-ins to process different kinds of things, then merge together the results in at the end? Maybe the former seems more suitable if most plug-ins will make a one to one file mapping? Maybe the latter seems more suitable if most plug-ins will return a tree of substantially different shape than what they accept?
I don't know the answer to the idiom question; but I do know this:
It would be easier to understand while getting started, if plug-ins would state in their documentation what they do with files they don't care about.
If you would rather have a pull request than this wall of text, just say so :-)
The text was updated successfully, but these errors were encountered:
All broccoli-filter based plugins pass files through that they don't know about. This is intentional - it can be useful e.g. when you mix .coffee, .handlebars and .js files. Thanks to the pass-through, you can chain the coffee script filter and a handlebars filter to get a tree full of .js files.
This plug-in appears to pass through all non-coffee files as is, while swapping out coffee files to JavaScript files. I found that surprising, and ended up chaining on another plug-in to strip out all but the JavaScript files.
Is the right idiom that a Broccoli build process should pass essentially all files through a series of plug-ins, each of which transparently passes along everything it doesn't care about? Or is the right idiom that one should use a bunch of plug-ins to process different kinds of things, then merge together the results in at the end? Maybe the former seems more suitable if most plug-ins will make a one to one file mapping? Maybe the latter seems more suitable if most plug-ins will return a tree of substantially different shape than what they accept?
I don't know the answer to the idiom question; but I do know this:
It would be easier to understand while getting started, if plug-ins would state in their documentation what they do with files they don't care about.
If you would rather have a pull request than this wall of text, just say so :-)
The text was updated successfully, but these errors were encountered: