<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> </head> <body bgcolor="#ffffff" text="#000000"> Matthew Flatt wrote at 08/28/2011 08:40 AM: <blockquote cite="mid:20110828124000.DCE8A6500BC@mail-svr1.cs.utah.edu" type="cite"> <pre wrap="">At Sun, 28 Aug 2011 02:46:18 -0400, Neil Van Dyke wrote: </pre> <blockquote type="cite"> <pre wrap="">Oh, another option to get rid of the two modes: write out a file that specifies a custom reader (coming from a PLaneT package), and that reader can read syntax objects and return those rather than making syntax objects from reading text. </pre> </blockquote> <pre wrap=""> That sounds right to me, except I'm not sure about the "write out" part. I image your document source should be #lang planet neil/toscribble/doc "sourcefile" where the reader specified by `neil/toscribble/doc' reads the named file and produces a syntax object for a module that is a Scribble document. </pre> </blockquote> <br> I should have mentioned that my need is for API documentation embedded in a normal "#lang racket/base" file.  So, although it's the source file for the documentation, I can't simply hand that file directly to Scribble.  I need to translate the file first, and then either call Scribble directly with syntax objects or (sounds like) write out a file that Scribble then processes.<br> <br> If I write out a file, but I want Scribble error messages to point to locations in the original "#lang racket/base" source file, then I think I'll need the written-out file to have the "#lang planet neil/toscribble/doc" line that will do a reader trick to make syntax objects with locations that refer to the original "#lang racket/base" file.<br> <br> <div class="moz-signature">-- <br> <a class="moz-txt-link-freetext" href="http://www.neilvandyke.org/">http://www.neilvandyke.org/</a> </div> </body> </html>