pyblosxom-users — Mailing list for pyblosxom users and hackers.

[pyblosxom-users] small fix to make py['depth'] work

[jens p. <je...@pe...>]

Hello, I'm deploying a site based on pyblosxom and found a bug in the
handling of recursive directories. If I have som directoris as follows:

A-B1-C1
 |  -C2
 |  -C3
 |  -C4
 -B2
 -B3
 -B4

then the settings:
py['datadir'] = 'A'
py['depth'] = 2

will only list the contents of A, B1, C1 and B2.

The inclosed patch against CVS fixes this behaviour

Best of everything and thanks for a good tool
/jp

$ diff -Naur pyblosxom/libs/tools.py pyblosxom.cx/libs/tools.py
--- pyblosxom/libs/tools.py     2002-11-24 06:26:39.000000000 +0100
+++ pyblosxom.cx/libs/tools.py  2003-01-05 13:10:34.000000000 +0100
@@ -93,8 +93,7 @@
         # recursively scan other folders, appending results
         if (recurse == 0) or (recurse > 1):
             if os.path.isdir(fullname) and not os.path.islink(fullname):
-                recurse = ( recurse > 1 and recurse -  1 or 0)
-                result = result + Walk( fullname, recurse, pattern, return_folders )
+                result = result + Walk( fullname, ( recurse > 1 and recurse -  1 or 0), pattern, return_folders )
              
     return result
  


-- 
 jens persson         #
 <je...@pe...>    #       This space intentionally left blank
 Mäster Olofsväg 24   #
 S-224 66 LUND;SWEDEN #

[pyblosxom-users] num_entries in subblogs?

[Mark P. <mar...@ma...>]

PyBlosxom seems to not honor the num_entries setting when viewing the main 
page of a nonroot weblog, whereas Blosxom does. My two subblogs have 
entirely different templates and purposes, so showing only num_entries on 
the main page of those subblogs is much more useful. Is this right? Line 
278 of libs/pyblosxom.py reads:

if (self.py['pi_bl'] == '' and self.py['pi_yr'] == '') and count >= 
int(self.py['num_entries']):

when it should (to be like Blosxom) read:

if self.py['pi_yr'] == '' and count >= int(self.py['num_entries']):

Sorry if this has been covered already.

--
Mark Paschal
mar...@ma...
http://markpasc.org/weblog/

[pyblosxom-users] drop-in plugins and item 6

[will <wi...@bl...>]

I've been using pyblosxom for a month or two (or three--I forget) now and 
overall it's been ok.  I made some changes locally to handle drop-in 
plugins that don't require me to change the pyblosxom.py file (like 
with pyarchives.py).  My driving motivation was:

  a) i wanted to reference archives in templates other than foot
  b) i wanted to add other plugins that create variable 
     content like pyarchives (like a calendar, for instance), 
     by dropping the file into the directory without having to
     edit pyblosxom.py

So I fixed up libs/plugins/__init__.py to load plugin modules and let them 
throw variables into the py dict.  It works for my specific needs, but it 
could be generalized for modules that do things other than add variables 
to the py dict.  I don't know what folks have in mind though.

I started tossing around item 6.  I also re-wrote pyarchives to figure
things out only if it's asked to.  It uses __str__ to do it, though, which
doesn't allow you to pass data in for the calculations.  

One way to take that a step further is to add to the variable syntax 
something like this:

   $varname.somefunc(a, b)

where the py variable would be varname with some method somefunc that 
takes in arguments a, b.  Then the regexp would be something like 
this:

   r'\$([\w_\-]+(?:\.\w+\([^\)]+\))?)'

that matches:

   $varname
   $varname.func()
   $varname.func(a)
   $varname.func(a, b, "blah")

It creates the issue of wanting to embed other variables in the variables 
like:

   $varname.func("$file_path", "$tb_id")

which has some cheesy but doable solutions.  Though I think it'd be better
if we added some kind of eval mechanism instead.  Then you could access
python things as well as pyblosxom things.  Out of curiosity, how does 
blosxom do it?  (Or do they?)

It's 3:00am and I'm rambling.  I have a replica of the codebase on my
web-site with the original files, my versions, and diff files (diff -u).  
I didn't want to spam the list with diffs and such.

   http://www.bluesock.org/~willg/pyblosxom/

/will

-- 
whatever it is, you can find it at http://www.bluesock.org/~willg/
except Will--you can only see him in real life.

[pyblosxom-users] Things to do in PyBlosxom (Any takers?)

[Wari W. <wa...@ce...>]

There are several things that I have planned for pyblosxom. All toward
the goal of extendability. Here's a list of things to do, may not be
in any order though.

 1. Break up the configuration to a seperate file.
    Right now we have the configuration of pyblosxom in the main CGI.
    This does pose a problem when I want to put in features in
    pyblosxom, like comments, etc. Also, the fact that XMLRPC is not
    compatible with GET and POST in the same CGI is a problem when it
    comes to put in trackback or pingbacks.
    
 2. Break up the XMLRPC and the main pyblosxom.
    Before I can do this, I will have to break up the configuration (See
    1.) I want pyblosxom to just do the blosxom things, and I want the
    bloggerAPI support to just do the bloggerAPI things (You can delete
    it if you don't want it). This is a good thing. You'll have less
    bugs this way.
 
 3. Create an abstraction layer for data sources.
    What is data sources? Could mean the text files in you directories,
    could also be the emails in you `Maildir/` folders. I'm banking on
    Abe to give me a nice abstraction layer for his
    [http://www.fettig.net/index.cgi/2002/12/#Heppinings messaging
    library]. I can then see if there's a need to extend it, and I'll
    work on the blosxom format stuff as well. I do hope that Abe has got
    an indexing mechanism for his `Maildir/` stuff as parsing mail files
    can be real slow.
 
 4. Create an abstraction layer for the templates renderer.
    While the blosxom templates (flavours) serves it's purpose well, it
    can be quite simplistic for some. People may want to work on
    Albatross, Cheetah or empy or what have you. This should allow for
    people more creative than me :)

 5. Comments, trackback and feedback.
    Ah yes, something that I personally need. Before I can migrate my
    real site http://roughingit.subtlehints.net/ to pyblosxom, I want to
    keep in the comments as well. I can probably plug some external
    tools for comments (like what I did with trackback) but really, it's
    not the same. But there's one question here. How can I store
    comments? In a text file? Or could be in an embeded database like
    sqlite, or gadfly, etc. Looks like I need some form of abstraction
    here as well. Then again, how can I know the number of comments or
    trackbacks on the fly? That'll lead to point 6.
 
 6. Dynamic variables in templates.
    Sort of destroying point number 4. (Or could work together) I wish
    that I've got a $variable that I need that can call a function to do
    some calculations for me. Like the Archives portion. If you don't
    use it, CPU cycles will not be wasted for it. If you do, I just call
    up the function, simple as that. This is useful for getting the
    number of comments, etc.

Well looks like there's lots to do, and I'll have to keep this in mind.
For all of you reading this, have a Merry Chrismas and a Happy New Year.

-- 
Regards: Wari Wahab
Senior R&D Engineer
Celestix Networks
http://www.celestix.com/

Re: [pyblosxom-users] preformatter

[Wari W. <wa...@ce...>]

For the latest release (rev3), all the preformatters are located in the
libs/preformatters/ directory, included in the distribution are wiki
(requires MoinMoin), genericwiki (quite expandable and useful too,
written by jo...@ou...) and some others that may or may not be
useful to you. They do provide a good example on how to code your own
parser.

Then older (rev2) requires a file called preformatter_xxxx.py in your
libs directory.

Anyway, preformatters is the reason pyblosxom was born :)

On Mon, 2002-12-16 at 23:46, Kai Hendry wrote:
> Just realized how to make use of the preformatter in p.ini by setting:
> ; Default Parser
> parser = preformatter
> 
> I want some sort of formatter, as when I was typing paragraphs in my blog.txt
> it was not wrapping the paragraphs in p tags.
> 
> However since setting it, it has made no changes. Is there any docs to make use
> of these preformatters available in the disb?
> 
> http://db.cs.helsinki.fi/~hendry/log/
> 
> And my last post is attached.
> 
> Regards
> -Kai

[pyblosxom-users] preformatter

[Kai H. <he...@cs...>]

Just realized how to make use of the preformatter in p.ini by setting:
; Default Parser
parser = preformatter

I want some sort of formatter, as when I was typing paragraphs in my blog.txt
it was not wrapping the paragraphs in p tags.

However since setting it, it has made no changes. Is there any docs to make use
of these preformatters available in the disb?

http://db.cs.helsinki.fi/~hendry/log/

And my last post is attached.

Regards
-Kai

Re: [pyblosxom-users] Using mod_python and pyblosxom?

[Wari W. <wa...@ce...>]

On Mon, 2002-12-16 at 02:39, er...@sp... wrote:
> I'm tinkering with the idea of using pyblosxom with mod_python in apache to
> avoid the overhead involved when using cgi.

Having no mod_python on my installation, I have no clue how to test such
implementation.

> I realize the same probably applies to mod_perl and blosxom, but I prefer
> using python.

Me too :)

> I haven't got a clue about how to do it, since I'm relatively inexperienced 
> when it comes to web-programming (actually programming at large :)

My understanding is that for a mod_python type of programs, pyblosxom
should never exit when called, currently of course, it exits :)

The easiest way I can think of, is to use jonpy CGI modules, which is
great, coz you write the code once, and subclass the proper CGI classes,
and it's mod_python, or fastCGI or CGi compatible (at least in theory).

> Anyway, I'm going to try to see if I can make it work and I will keep you 
> posted on the progress. It probably won't hurt to try ;-)

Thanks, I really do urge you to use jonpy (http://jonpy.sf.net/) as it
keeps code changes to a minimal. I was planning to use it myself, but
have no places to test it.

-- 
Regards: Wari Wahab
Senior R&D Engineer
Celestix Networks
http://www.celestix.com/

[pyblosxom-users] Using mod_python and pyblosxom?

[<er...@sp...>]

Hello,

I'm tinkering with the idea of using pyblosxom with mod_python in apache =
to
avoid the overhead involved when using cgi.

I realize the same probably applies to mod_perl and blosxom, but I prefer
using python.

I haven't got a clue about how to do it, since I'm relatively inexperienc=
ed=20
when it comes to web-programming (actually programming at large :)

Anyway, I'm going to try to see if I can make it work and I will keep you=
=20
posted on the progress. It probably won't hurt to try ;-)

Kind regards,

Rune

--=20
"If we knew what we were doing, we couldn't call it research, now could w=
e?"
- A. Einstein, 1941

[pyblosxom-users] Re: .htaccess

[Wari W. <wa...@ce...>]

I'm cc'ing this to the mailing list as I think it contains useful
information.

* Kai Hendry <he...@cs...> [021125 22:19]:
> You changed the .ini ref to p.ini. That caught me off guard. :)

Hehe, have to force users to change that. INI files are not meant to be
seen by the webserver. Better to get you to be aware of it then to be
open to unsecure possibilities.

> Including the .htaccess should do no harm to the installation, and it
> makes it easier for me to "play".

Different people will need different .htaccess, mine for example looks
like:

<Files pyblosxom>
ForceType application/cgi-script
SetHandler cgi-script
</Files>

Better to give examples. Really. Other people may thank me for it
instead of me forcing a change for them.

> The release numbering could be better. I do not like revX notation so
> much.

Once I got pyblosxom to work radically different from Rael's blosxom, I
will change the version to my/our liking, like 1.0.01, etc..

> Also the head is not in pyarchives. I guess I need to edit
> pyblosxom.py again!

Like I put in my previous mail, it can slow down rendering, so I put it
last. For the next release, I'm trying to load this automagically, that
is, when a flavour needs it then it'll be called, that's why it's in a
plugin directory now, and I'm going to make an auto registering plugin
architechure for template. Once that happens you do not need to change
the pyarchives line anymore as it will not be there at all.

-- 
Regards: Wari Wahab
Senior R&D Engineer
Celestix Networks
http://www.celestix.com/

Re: [pyblosxom-users] first post

[Wari W. <wa...@ce...>]

* Kai Hendry <he...@cs...> [021124 19:01]:
> When I am writing blogs, I find it really tiresome to write <li> </li>
> type tags.

Yes, me too, it's the exact reason why I created pyblosxom, so that I do
not have to do HTML editing.

> Is it possible to convert:
> *	a point
> *	another point
> 
> to 
> <ul>
> <li>a point</li>
> <li>another point</li>
> </ul>

BTW, I use MoinMoin http://wiki.subtlehints.net/moin/MoinMoin to help me
translate all the text markups to HTML markups. This is what
preformatters are for. Check out
http://wiki.subtlehints.net/moin/PyBlosxom/ExtendingPyBlosxom/PreFormatters
and
http://wiki.subtlehints.net/moin/PyBlosxom/ExtendingPyBlosxom/PreFormatters/WikiFormat
on how to extend your pyblosxom. Actually, all the codes are there, and
in the latest release, there's a mini version called genericwiki.py
which will help you do simple markups simpler :) BTW, preformatters
interface is changed on the new release, do read the changelog for
details.

To use preformatters, you either need to modify the INI file for a
default prefomatter, or create it with your own blog entry. Here's an
example of genericwiki (You should change the filename so that it's
smaller and easier to type.

------------
This is a new blog entry
#parser genericwiki
This entry should have '''bold''' and ''italics'' text.
[http://roughingit.subtlehints.net/pyblosxom Click here for details].
This demo good 'nuff?
------------

This will output:

This entry should have <b>bold</b> and <i>italics</i> text. <a
href="http://roughingit.subtlehints.net/pyblosxom">Click here for
details</a>. This demo good 'nuff?

---------------------

So what we see here is that there's no HTML in the original text. If you
look at the previous email anouncing the latest pyblosxom, that is
MoinMoin wiki markup, the same text is used on the pyblosxom website.

> Or is that not the "blosxom" way? Maybe someone can suggest a vim macro
> or something. :)

It's not an original blosxom way.

> Another thing I have noticed is that filling directories with
> blogs*.txt(and templates) leaves it all a little cluttered for my
> liking. I am not sure what the ideal solution would be really.
> Maybe converting them to dotfiles after a month?

Do what you like, that's up to the user, the entries are not touched by
pyblosxom at all, just read. You might want to try out dotfiling them.
Try it out first and see if it works, if it does, set up a cron job to
help doing the work for you.

> I also tend to update various blogs, and I wish the blogger was clever
> enough to know I merely updated it, and maybe link back to previous
> versions. Like say I just wanted to add an update to the progress of
> the story I reported on earlier.

How do you propose this to be done? There are some kind of limitations
in blosxom style blogs (I believe it's actually a feature I liked), when
you edit a file, it immediately goes to the top, being marked as new.

Unless I have a proper database somewhere, this will not happen yet.

-- 
Regards: Wari Wahab
Senior R&D Engineer
Celestix Networks
http://www.celestix.com/

[pyblosxom-users] [ANN] PyBlosxom 0+5i_rev3 released

[Wari W. <wa...@ce...>]

I am proud to announce the latest version of pyblosxom. You are advised
to upgrade to this version as it fixes a rendering of blog entries when
you query via a /YYYY/MM combination.

Also in this version, I've decided to organize a few things here and
there, one of it is the change of how Preformatters work. I've now
placed them in the `libs/preformatters/` directory, so those of you with
custom parsers might want to have a look at the other files as examples.
Changes are rather non destructive, I just require you to subclass your
Preformatter class with Preformatterbase. This should help future
upgrades if we need to modify the interface.

Etags and Last-Modified are added in this release, though I've not made
it a user option yet. The pyblosxom for windows is not fixed yet, so for
Windows users, please hang-on. There's still lots to do.

This release is quite stable. Better than rev2. I've placed the files in
the download area of Sourceforge.net
[http://sourceforge.net/project/showfiles.php?group_id=67445 here].
That's one option, or you can go ahead and download using CVS (the
better option IMHO).  CVS instructions are located
[http://sourceforge.net/cvs/?group_id=67445 here].

Anyway, do view the
[http://roughingit.subtlehints.net/code/ChangeLog.html ChangeLog] for
more details of the changes made.

Most of all, have fun :)

[pyblosxom-users] first post

[Kai H. <he...@cs...>]

When I am writing blogs, I find it really tiresome to write 
<li> </li> type tags.

Is it possible to convert:
*	a point
*	another point

to 
<ul>
<li>a point</li>
<li>another point</li>
</ul>

Or is that not the "blosxom" way? Maybe someone can suggest a vim macro
or something. :)

Another thing I have noticed is that filling directories with
blogs*.txt(and templates) leaves it all a little cluttered for my
liking. I am not sure what the ideal solution would be really.
Maybe converting them to dotfiles after a month?

I also tend to update various blogs, and I wish the blogger was clever
enough to know I merely updated it, and maybe link back to previous
versions. Like say I just wanted to add an update to the progress of the
story I reported on earlier.

-Kai
--
http://db.cs.helsinki.fi/~hendry/log/

[pyblosxom-users] This is a test post

[Wari W. <wa...@ho...>]

To show that this mailing list works.

Plese ignore.