libera/#commonlisp - IRC Chatlog
Search
9:59:26
scymtym
does anyone know what ldb meant by dpans-parser? i can only find mentions of https://github.com/robert-strandh/dpANS-parser but that repository seems to not exist anymore
10:14:38
beach
scymtym: At least, ldb was aware of your work, so I assumed that's what was being referred to.
10:18:25
beach
One could check how they all deal with that page that scymtym identified as problematic, because it is incorrectly translated in the Common Lisp HyperSpec.
10:21:25
scymtym
i think each of the existing parsers (including mine), as well as the dpANS tex sources themselves, have certain issues. gilberth's approach is probably flawless in terms of accurate translation but as i understand it, the result is not a document object tree suitable for further processing
10:25:43
scymtym
i'm not sure how easy that would be since he (again, in my understanding) executes a tex implementation by compiling the pascal source code to CL and capturing certain aspects of the output. he shared a video of the process but i don't know whether it was intended for general consumption
10:26:43
beach
Now that I think about it, that sounds just like something he (and only he) would do. :)
10:44:25
hayley
That description seems accurate - I saw some S-expression Pascal code of his somewhere.
11:42:43
utis
does anyone happen to know why hunchentoot serves another 404 that that in www/errors/404.html?
11:46:51
splittist
utis: have a look at https://github.com/edicl/hunchentoot/blob/ef175853ce4e2f6b3c7f5927450d8e8c44e70a4c/acceptor.lisp#L763 . You can specialize ACCEPTOR-STATUS-MESSAGE on the HTTP-STATUS-CODE argument (e.g. (http-status-code (eql ...))) and serve whatever you want.
11:55:53
splittist
(see also the :error-template-directory initarg to the acceptor object and the initialize-instance :after method on it to see what happens to that argument)
15:33:43
ldb
scymtym: I finally figured it out how to use the architecture.builder-protocol to produce valid XML from your dpans parser, so I can at least something to continue work on.
15:36:16
ldb
I presume the transformation does not drop anything important from concept-*.tex and dict-*.tex files?
15:37:46
ldb
I have only enabled "make-drop-[1,2]","cleanup-math", "cleanup-components", "cleanup-issues", "simplify".
15:40:03
ldb
The TeX macroexpansion is not something I need since I consider the orginal markup is already mostly semantic based.
16:05:15
ldb
Well, I just found scymtym 's parser treats \SymbolTokens. as a command with name "SymbolTokens." rather than "\SymbolTokens" + "."