[cl-markdown-devel] #. compile problem
He Gary, Sbcl is not totally happy with cl-markdown. In line-is-link-label-p it won't fasl-ize #.(ppcre:create-scanner '(:sequence link-label)) eg: ; file: /home/zeno/lisp/lib/cl-markdown/dev/markdown.lisp ; in: DEFUN LINE-IS-LINK-LABEL-P ; (CL-PPCRE:SCAN #<CLOSURE # {D461E9D}> CL-MARKDOWN::LINE) ; --> CL-PPCRE:SCAN LOAD-TIME-VALUE SB-KERNEL:MAKE-VALUE-CELL ; --> CL-PPCRE:CREATE-SCANNER ; ==> ; #<CLOSURE (LAMBDA #) {D461E9D}> ; ; caught ERROR: ; Objects of type FUNCTION can't be dumped into fasl files. the form should work fine without reading it in advance, or ..? btw, i just got cl-markdown hooked up in the web-app i'm working on. It functions quite nicely. A whole lot better than the nl2br methodology, thanks! Greets, Ties
Hi, Thanks for the notice. I got into the habit of #. my create-scanner forms as an optimization but should probably just trust the compiler.
btw, i just got cl-markdown hooked up in the web-app i'm working on. It functions quite nicely. A whole lot better than the nl2br methodology, thanks!
Hey, great. Please keep me up to date with any problems you encounter or ideas that would make CL-Markdown easier to use. -- Gary Warren King metabang.com http://www.metabang.com/ (413) 210 7511 gwking on #lisp (occasionally)
It functions quite nicely.
ehh... having said that. I didn't actually try more than one sentence. when i render (without the indentation): # bla www.frankrijk.com http://www.spanje.com ties@sleep.se i get (again without the indentation): <h1>bla www.frankrijk.com http://www.spanje.com ties@sleep.se </h1> e.g. a long line of big text. Shouldn't markdown have wrapped them all in <p>s? Bulletting frankrijk and spanje for example bullets them correctly, so i gather it's in the paragraph detection code. I tried to figure it out, but i'm not that fast. greets, Ties
Hi Ties, I think that this is a bug that is fixed. Where I try your sample, I get:
<H1>bla </H1> <P>www.frankrijk.com </P> <P>http://www.spanje.com </P> <P>ties@sleep.se </P>
which is correct as far as the structure goes but is missing the automatic URL conversation -- I'll try to look into that later today or tomorrow. On Jun 7, 2006, at 1:44 PM, Ties Stuij wrote:
It functions quite nicely.
ehh... having said that. I didn't actually try more than one sentence.
when i render (without the indentation):
# bla
www.frankrijk.com
ties@sleep.se
i get (again without the indentation):
<h1>bla
www.frankrijk.com
ties@sleep.se </h1>
e.g. a long line of big text. Shouldn't markdown have wrapped them all in <p>s? Bulletting frankrijk and spanje for example bullets them correctly, so i gather it's in the paragraph detection code. I tried to figure it out, but i'm not that fast.
greets, Ties
-- Gary Warren King metabang.com http://www.metabang.com/ (413) 210 7511 gwking on #lisp (occasionally)
ehh... having said that
well i pulled like hell on all markdowns dependencies, removed every potential stale fasl (in hindsight probably only fresh as grass fasls) but no dice. what helped a bit was removing (some-element-p (chunk-enders (current-chunk-parser)) (lambda (p) (funcall p line))) from chunk-source, right under ;; end current chunk this lets my program regognize the end of a <h> and wraps a maximum of two text blocks seperated by an empty line in <p>s. If there are more than two, all but the last text blocks are wrapped in the one and the last is wrapped in the other. It's voodoo to me. I will now leave the world of programming and start a farm on the countryside greets, Ties
I will make sure that the latest stuff is out that in darcs and ASDF- Install land and get back to you. On Jun 7, 2006, at 5:02 PM, Ties Stuij wrote:
ehh... having said that
well i pulled like hell on all markdowns dependencies, removed every potential stale fasl (in hindsight probably only fresh as grass fasls) but no dice.
what helped a bit was removing
(some-element-p (chunk-enders (current-chunk-parser)) (lambda (p) (funcall p line)))
from chunk-source, right under ;; end current chunk
this lets my program regognize the end of a <h> and wraps a maximum of two text blocks seperated by an empty line in <p>s. If there are more than two, all but the last text blocks are wrapped in the one and the last is wrapped in the other.
It's voodoo to me. I will now leave the world of programming and start a farm on the countryside
greets, Ties
-- Gary Warren King metabang.com http://www.metabang.com/ (413) 210 7511 gwking on #lisp (occasionally)
Hi again, FWIW, I did a fresh install with the ASDF-Installable versions this morning and CL-Markdown worked as far as the paragraph structures go. I haven't gotten to testing a fresh darcs pull yet though and I'm having some hard disk conniptions that are slowing up progress a bit (current drive starting to fail, new drive on the way...). On Jun 7, 2006, at 5:02 PM, Ties Stuij wrote:
ehh... having said that
well i pulled like hell on all markdowns dependencies, removed every potential stale fasl (in hindsight probably only fresh as grass fasls) but no dice.
what helped a bit was removing
(some-element-p (chunk-enders (current-chunk-parser)) (lambda (p) (funcall p line)))
from chunk-source, right under ;; end current chunk
this lets my program regognize the end of a <h> and wraps a maximum of two text blocks seperated by an empty line in <p>s. If there are more than two, all but the last text blocks are wrapped in the one and the last is wrapped in the other.
It's voodoo to me. I will now leave the world of programming and start a farm on the countryside
greets, Ties
-- Gary Warren King metabang.com http://www.metabang.com/ (413) 210 7511 gwking on #lisp (occasionally)
participants (2)
-
Gary King -
Ties Stuij