Hi,
the first release tarball for Closure HTML is available, the stand-alone
version of Closure's HTML parser.
Tarballs and documentation can be found in Closure's project on
common-lisp.net. The git archive is available on repo.or.cz.
For details, please see
http://common-lisp.net/project/closure/closure-html/
d.
Hi,
there are now two new projects:
- Closure HTML:
Closure's HTML parser and some of its dependencies; previously a
part of Closure.
http://common-lisp.net/project/closure/closure-html/
- Closure Common
Everything needed by both Closure HTML and Closure XML; basically
the packages runes and runes-encodings. Previously a part of cxml,
now a separate module in cxml CVS.
http://www.cliki.net/closure-common
Picture of the dependencies:
Closure
/ \
v v
Closure HTML Closure XML (cxml)
\ /
v v
Closure Common
Ultimately, both Closure Common and Closure HTML will be released as
asdf-installable tarballs, but at this point Closure HTML is not quite
in a form that I would like to release officially.
Before a release is done I would like to see:
- documentation
- test cases, in particular to make sure that character runes work
- good serialization support (like cxml has, only for HTML instead)
- an add-on library allowing Closure HTML to be used with XML's
in-memory representations for XHTML
Some of you might already have seen `trivial-html-parser' by Ignas
Mikalaj#nas, also a stand-alone version of Closure's HTML parser.
My new Closure HTML project is based on Ignas' work (thanks to him for
getting that started), but differs from it in some respects:
- the new archive includes all history
- so far, I have preserved closure's original directory structure
- trivial-html-parser still had all of glisp, which I have reduced to
only those functions necessary (still including gstreams though).
The remaining package is called HTML-GLISP to avoid conflicts with
the real GLISP package in Closure. The latter imports
gstream-related symbols from the former for re-export.
- Mime support is in a separate package, so that closure's netlib,
from which is was taken, can now :use that package instead.
- last but not least, a different name. (The "trivial" in
trivial-html-parser would not do this code base justice at all.)
d.
