Last weekend, I published a new version of bib2tpl, a PHP class to convert BibTeX to anything I wrote a while ago. There certainly were some bumps to iron out; these are the major changes made:

• Entry filtering now works on all fields.
• Templates can contain multiple occurrences of group and entry subtemplates now.
• Condition tags can perform more detailed comparisons now.
• Improved control over entry ordering.
• Grouping tags are no longer necessary if grouping is turned off.
• It is now possible to reuse parsed BibTeX for multiple conversions.
• @entryid@ has been removed; use @entrykey@ instead, which is now guaranteed to be unique.

You can find a more complete list of changes here, and more information on bib2tpl here.

Fancy Example

I have been charged lately with setting up a new homepage for my group at university. An integral part of any scientist’s professional homepage is his list of publications which basically represents his success. Despite its importance and therewith the wish that it better be up to date and look nice, most researchers I have met do not want to spend their time on maintaining lists or any part of any homepage. Therefore, a manually maintained HTML list or anything like that was out of question immediately. Other solutions like pulling from DBLP or embedding publicationslist.org have their own problems.

In the end, I decided on using BibTeX directly since most people in computer science use LaTeX and BibTeX anyway (or really should). I intended to provide a solution that would integrate well with any website, could use naked BibTeX as input and offered flexibility for individual adaptions as well as creating multiple, disjoint lists from one file. Given that I chose to build the website with a wiki, it should also be possible to store both BibTeX and template in the wiki, allowing for dynamic and easy adaptions.

Sadly, I could not find any software that met my needs. Consequently, dutiful computer nerd that I am, I set out to build my own solution. I am really bad at making up names, so I went with bib2tpl. Since Google does return exactly zero results I should not step on any toes. It is written in PHP, of course, and can transform BibTeX to anything you build a template for. I did find a BibTeX parser in PEAR so I was spared that part. The simple template language it uses I built myself in order to have a minimal yet sufficient amount of power in the given scenario. We will see wether or not this was a good decision.

You can find bib2tpl on its own pages. Please try it out and tell me what you think!