Table of Contents
We give hints on how to upgrade the website: installing stuff from current CVS on http://logreport.org.
If you wanna upload a complete new site:
vanbaal@gelfand:~/cvs-sourceforge/docs% tar --exclude CVS -zcf \ htdocs.tar.gz htdocs vanbaal@gelfand:~/cvs-sourceforge/docs% scp htdocs.tar.gz \ hibou.logreport.org: vanbaal@hibou:~% tar zxf htdocs.tar.gz vanbaal@hibou:~% mv htdocs logreport.org vanbaal@hibou:/var/www% rm -rf logreport.org.bak vanbaal@hibou:/var/www% mv logreport.org logreport.org.bak vanbaal@hibou:/var/www% mv ~/logreport.org . vanbaal@hibou:/var/www% mv logreport.org.bak/pub logreport.org/ vanbaal@hibou:/var/www% mv logreport.org.bak/9* logreport.org/ vanbaal@hibou:/var/www% chown -R .www logreport.org vanbaal@hibou:/var/www% chmod -R g+w logreport.orgor, if you've only changed some pages:
vanbaal@gelfand:~/cvs-sourceforge/logreport/docs% scp \ htdocs/developers.phtml htdocs/toolbar.inc htdocs/news.inc \ hibou.logreport.org:/var/www/logreport.org/
Be sure the links to stuff under /pub/current are still alive. E.g. the files TODO, dev-manual.html and user-manual.html are linked to.
The DTD's are published as HTML on the website by using hibou:/usr/local/src/dtdparse/dtdparse-2.0b2-LogReportPatched.tar.gz, which is a patched version of Norman Walsh's dtdparse utility. Before the utility is run, make sure that the DocBook DTD is not included in the parsing process, because the DocBook DTD should not be published. This is done by changing the line:
<!ENTITY % load.docbookx "INCLUDE" >into:
<!ENTITY % load.docbookx "IGNORE" >
The webpages are then generated with:
perl ~/dtdparse-2.0b2-patched/dtdparse.pl --title "XML Lire Report Markup Language" --output lire.xml lire.dtd perl ~/dtdparse-2.0b2-patched/dtdformat.pl --html lire.xmlThe resulting lire directory can be tar-ed, gziped and unpacked again on hibou in the directory /var/www/logreport.org/pub/docs/dtd/.
The other two DTD's are HTML-anized similarly, but keep in mind to change the title when running dtdparse.pl.