openjpa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kevin Sutter <kwsut...@gmail.com>
Subject Re: Exporting website from CWIKI
Date Fri, 16 Nov 2012 01:04:30 GMT
Tables are ugly in this markdown syntax.  The initial conversion to mdtext
did a partial job, but every table definition needs massaging.  For an
example, look at the documentation.mdtext page that I just committed.  I
did this one by hand (and some find/replace macros).  But, we need some
type of sed macro or something to do this to all of the pages with tables.
Here are some examples:

<table>
needs to be...
<table border="1">

[manual](
http://openjpa.apache.org/builds/apache-openjpa-1.3.0-SNAPSHOT/docs/manual/main.html
)
needs to be...
<a href="
http://openjpa.apache.org/builds/apache-openjpa-1.3.0-SNAPSHOT/docs/manual/main.html
">manual</a>

[manual</td><td>
http://openjpa.apache.org/builds/apache-openjpa-1.3.0-SNAPSHOT/docs/manual/manual.pdf
]
needs to be...
<a href="
http://openjpa.apache.org/builds/apache-openjpa-1.3.0-SNAPSHOT/docs/manual/manual.pdf
">manual</a>

The basic issue is that the simple table support in markdown is way too
simple.  It doesn't provide borders for one thing.  And, lining up columns
is a pain.  So, the recommendation is to use html <table>.  Hopefully,
somebody is more verse with editing tools or scripts or templates or
something to make this job less painful...

Thanks,
Kevin

On Thu, Nov 15, 2012 at 2:33 PM, Kevin Sutter <kwsutter@gmail.com> wrote:

> Okay, now that I am hearing a couple more success stories, it's time to
> divvy up the work.  Since some of the pages take more time than others, I
> don't want people to be wasting time on a page that is already "checked
> out".  So, I created a simple text file:
> https://svn.apache.org/repos/asf/openjpa/site/trunk/WhoIsWorkingOnWhat.txt
>
> User           Pages
> kwsutter       documentation.mdtext, documentation-top-link.mdtext
> rgcurtis
> leealber
> helenxu
> etc, etc
>
>
>
>
> Done pages
> quick-start.mdtext
>
>
> Synchronize this with your workspace and edit it when you pick a new file
> to update.  When it's done to your liking, check it back in and then move
> the name of the file to the "done" section.  About as simple as you can
> get.  Hope it helps...
>
> Thanks for your help!
> Kevin
>
>
> On Thu, Nov 15, 2012 at 1:45 PM, Kevin Sutter <kwsutter@gmail.com> wrote:
>
>> YooHoo!  finally had success!
>>
>> Here's what I did...  I went with an Ubuntu vm image (
>> http://www.trendsigma.net/vmware/ubuntu1010t.html, there are others, but
>> this is the one I worked with).
>>
>> I installed the Python packages.  First, the setuptools and then the
>> three packages:
>> easy_install Pygments
>> easy_install ElementTree
>> easy_install Markdown
>>
>> I installed the Perl packages.  I used the force option (-f) since some
>> of them failed the first time:
>>
>> cpan XML::Atom::Feed
>> cpan XML::RSS::Parser::Lite
>> cpan XML::Parser::Lite
>> cpan YAML
>> cpan SVN::Client
>>
>> I installed the SVN client:
>> apt-get update            // updates the list of packages available (may
>> not be necessary in your env)
>> apt-get install subversion
>>
>> I then following the cms local build instructions:
>> http://www.apache.org/dev/cmsref.html#local-build
>>
>> Mike's scripts should also work, but I was having so many issues, I
>> decided to go the old fashioned way...
>>
>> The final build is then done via this command:
>> perl build_site.pl --source-base /content/root/folder --target-base
>> /desired/output/folder
>>
>> Where you have to fill in your site root folder and target folders.  The
>> build went rather fast.  I then brought up Firefox and looked at the site:
>> file:///home/user/openjpa/site/trunk/target/content/quick-start.html
>>
>> Some pages are in better shape than others...  I am going to pick a page
>> and try to clean it up.  Once we get another person or two up-to-speed,
>> then we can try to divvy up the remaining work.  Good luck!
>>
>> Kevin
>>
>>
>> On Thu, Nov 15, 2012 at 9:08 AM, Albert Lee <allee8285@gmail.com> wrote:
>>
>>> Mike,
>>>
>>> Thanks for all the help. Will give it another try.
>>>
>>> Albert.
>>>
>>> On Wed, Nov 14, 2012 at 8:41 PM, Michael Dick <michael.d.dick@gmail.com
>>> >wrote:
>>>
>>> > Sorry I can't reply much during the work day I'll try answer
>>> everything.
>>> >
>>> >
>>> > On Wed, Nov 14, 2012 at 4:45 PM, Kevin Sutter <kwsutter@gmail.com>
>>> wrote:
>>> >
>>> > > Okay, I think I'm giving up on the Windows approach...  The build
>>> scripts
>>> > > all seem to require .sh scripts and other linux-specific items.  I
>>> might
>>> > > follow Rick's lead and try a VMWare linux image...
>>> > >
>>> >
>>> > Believe it or don't, I tried Windows first too. And gave up.
>>> >
>>> >
>>> > > Maybe Mike is the only one that can get this build running and we
>>> just
>>> > need
>>> > > to pitch to in $20 per to buy him a nice dinner...  :-)
>>> > >
>>> >
>>> > Preferred payment plan is 25 year old Highland Park (will also accept
>>> Old
>>> > Pulteney)
>>> >
>>> > <snip>
>>> >
>>> >
>>> > > > cpan SVN::Client
>>> > > >
>>> > > > These all seemed to work except for the SVN::Client module.  This
>>> one
>>> > > > seemed to require a different build process and it eventually
blew
>>> > > chunks.
>>> > > > I don't know if this module is really needed or not.  I'll try
a
>>> couple
>>> > > of
>>> > > > other things, but otherwise, I'll move on and see if I really
need
>>> it.
>>> > >
>>> >
>>> > I think I had a similar problem, but it just worked for me. The build
>>> > scripts don't commit anything so I think it's not needed.
>>> >
>>> > <snip>
>>> >
>>> > > >> setuptools, pygments and ElementTree installed with no problem.
>>> > However
>>> > > >> when tried to install Markdown, I got:
>>> > > >>
>>> > > >> [root@atlanta ~]# sudo easy_install Markdown
>>> > > >> Searching for Markdown
>>> > > >> Reading http://pypi.python.org/simple/Markdown/
>>> > > >> Reading http://www.freewisdom.org/projects/python-markdown
>>> > > >> Reading
>>> https://sourceforge.net/project/showfiles.php?group_id=153041
>>> > > >> Reading http://packages.python.org/Markdown/
>>> > > >> Reading http://www.freewisdom.org/projects/python-markdown/
>>> > > >> Best match: Markdown 2.2.1
>>> > > >> Downloading
>>> > >
>>> >
>>> > You might have better luck if you use apt (debian based distros) or yum
>>> > (RedHat based) to directly install the python module for markdown. I'm
>>> > guessing you're on RHEL, so I can't help much, other than I think it's
>>> > sudo yum --search markdown (then install the one with python)
>>> >
>>> > That said:
>>> >
>>> >  >>
>>> > > >>
>>> > >
>>> >
>>> "/usr/lib/python2.4/site-packages/setuptools-0.6c11-py2.4.egg/setuptools/sandbox.py",
>>> > > >> line 64, in <lambda>
>>> > > >>     {'__file__':setup_script, '__name__':'__main__'}
>>> > > >>   File "setup.py", line 130
>>> > > >>     with codecs.open('docs/_template.html', encoding='utf-8')
as
>>> f:
>>> > > >>               ^
>>> > > >> SyntaxError: invalid syntax
>>> > > >> [root@atlanta ~]#
>>> > > >>
>>> > > >>
>>> > > >> Any idea why it has SyntaxErro on the codecs token in setup.py?
>>> > >
>>> >
>>> > You need to update python. The 'with' syntax was introduced in python
>>> 2.6
>>> > or 2.7 I think. Yum / apt should help here.
>>> >
>>> > Regarding the files to edit : look for the mdtext ones, those are
>>> markdown
>>> > format. More advanced web developers than me may build templates and
>>> other
>>> > cool things! But I just wanted to get the site to look more or less
>>> > correct. The cwiki extension should be the original confluence wiki
>>> text,
>>> > and I'm not sure what the files without extensions are.
>>> >
>>> > Regarding locking the files, well, we've never had to lock files
>>> before,
>>> > and I don't know a good way to do so right now. If folks want to sign
>>> up
>>> > for a page we could add a table to the wiki (no irony there),
>>> resurrect the
>>> > irc room, or just email when we're starting to work on a page.
>>> >
>>> > Perhaps my brute force approach really isn't the best way to go about
>>> this.
>>> > I suspect that there are lot of pages which aren't needed or aren't
>>> > reachable anymore. Status pages, placeholders (tables of contents,
>>> etc.),
>>> > and legacy pages could be saved off and coverted later. I didn't go
>>> through
>>> > an identify those pages before I committed them - partly because I ran
>>> out
>>> > of time, and partly because I didn't want to be the only one to make
>>> the
>>> > call.
>>> >
>>> > -mike
>>> >
>>>
>>>
>>>
>>> --
>>> Albert Lee.
>>>
>>
>>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message