From:  Andy Sterland <>
Date:  16 Jan 2014 08:40:10 Hong Kong Time

Question on escaping for sourcesContent


I've got a couple questions/comments on sourcesContent. Hopefully this is the right place to mail them too :).

Presumably the intent for sourcesContent is that it's just a regular JSON string and the compiler should encode the source file as Unicode and escape certain characters like string terminators, line breaks, tabs, control characters, etc? Which sounds reasonable. Though I was wondering if in some circumstances it might be easier for the compiler and debugger to treat the source file as binary. Essentially encode the source files as base64 and then let the debugger decode it as they would for a file. Granted you might need additional metadata such as content-encoding (if it's not captured in a BOM or similar).  Thoughts?

Apologies if this has been discussed at length and I missed it.