Re: Org as a static site generator

[Ian Barton <lists@wilkesley.net>]

On 06/04/13 16:15, Christopher Allan Webber wrote:
> Cool, thanks for that info!
>
> Ian Barton writes:
>
>> On Fri, Apr 05, 2013 at 11:02:56AM -0500, Christopher Allan Webber wrote:
>>> Ian Barton writes:
>>>
>>>> On 01/04/13 13:08, Vincent Beffara wrote:
>>>>
>>>>> Yes, I mean, I know which html you need for that, simply within o-blog you need to manage between relative paths, absolute paths, canonical paths and so on in the template, to match the right section,  - mainly it should be a matter of let-ing the right variable to the right value at the right point in the template and catching it when generating the toc, but I never took the time to get it right ...
>>>>>> I've also just found this, which uses Org only as a markup tool and
>>>>>> Jekyll to generate the site:
>>>>>>
>>>>>> http://orgmode.org/worg/org-tutorials/org-jekyll.html
>>>>> I had a look at the too, but it felt just a little bit too convoluted compared to managing everything from Org. Besides, it seems to lose fontification of code snippets and the like?
>>>>>
>>>>> /v
>>>>>
>>>> As the original author of that page, I agree that using Jekyll is
>>>> convoluted, but it gives you much more control. However I now use
>>>> Pelican: https://pelican.readthedocs.org/en/3.1.1/
>>>>
>>>> There are a few reasons for this. Pelican is written in Python, which I
>>>> find easier to hack on. It is more flexible than Jekyll, which I found
>>>> hard to get to work the way I wanted with categories and tags.
>>>>
>>>> I wrote a yaml importer for Pelican so I could use my old jekyll posts.
>>>> However, Pelican understands Markdown, which I think the new exporter
>>>> supports.
>>>>
>>>> So my work flow now is Emacs-> export as html -> run Jekyll
>>>>
>>>> Ian.
>>>
>>> Heya Ian,
>>>
>>> I've been planning to switch my blog over to pelican.  It's cool to hear
>>> you say this.
>>>
>>> Is there any special elisp you use for the export, including converting
>>> things like the title, etc?
>>>
>>> Thanks!
>>>   - Chris
>>
>>
>> Hi Chris,
>>
>> No, nothing special. I just use org's standard publish functions. However, I publish only the body part of the html and place the yaml tags in the org file. A typical org file for a blog post would look like:
>>
>>
>> #+STARTUP: showall indent
>> #+STARTUP: hidestars
>> #+OPTIONS: H:2 num:nil tags:nil toc:nil timestamps:nil
>> #+BEGIN_HTML
>> ---
>> title: My Fire Steel Crumbles to Dust.
>> date: 2013-02-17
>> tags: [gear]
>> category: blog
>>
>> ---
>> #+END_HTML
>>
>> After my walk over Moel Famau and Moel Arthur I was looking forward
>> to making a hot drink. My brew kit lives permanently in the boot of
>>
>>
>>
>> org pubish then creates a file with a yaml header and html body text. Then I just run Pelican to publish the post. I have written a Pelican yaml reader which converts the yaml files to allow Pelican to process them. I'll document the whole process over the next couple of days and put it on Worg. I keep meaning to contribute my yaml reader back to Pelican, but it's quite specific to publishing org-mode files and not really a general purpose yaml importer.
>>
>>
>> --
>>

I have just written a short post at: 
http://www.ian-barton.com/posts/2013/Apr/06/blogging-with-emacs-org-mode-and-pelican/ 
which describes how I use org-mode and Pelican.

I'll write a more comprehensive explanation for Worg.

Ian.