yab2web

yab2web is yet another BibTeX-to-HTML converter, along the lines of (but naturally better than :-) various similar tools.

As shown in this example, yab2web takes as input a BibTeX file, and generates as output a static HTML page that shows the bibliographic items, along with a simple query/filter mechanism.

Note that yab2web is not a tool for editing/maintaining BibTeX files. (If that's what you want, try some of these.)

Download

yab2web is a 500-odd line Perl script -- you can download it
HERE.

Instructions

yab2web depends on two third-party pieces of software:

• L^ATEX2HTML does most of the heavy lifting. I think that many Linux distributions come with L^ATEX2HTML pre-installed.
• btOOL's Text::BibTeX Perl module is used to parse the BibTeX. You'll probably need to install this yourself.

You tell yab2web what else to add to the generated page using an HTML template. Here's mine: nk.yab. This file is just plain HTML, with one exception: your template must contain a line of the form:
      yab2web bibfile
When it generates its output, yab2web will replace this line with a large chunk of HTML.

You run yab2web with a command such as the following:
      ./yab2web < nk.yab
If all goes well, you will find the output in a file called "yab2web-output.html".

yab2web will print out a few informative messages as it proceeds. You may also see many many warning messages of the form
      lexical buffer overflowed (reallocating to 4000 bytes)
You can safely ignore these messages; they are generated by btOOL with large BibTeX files.

To check whether your BibTeX file contains error, look at /tmp/yab2web-latex.log.

yab2web uses several special (ie, nonstandard) BibTeX fields. All of these fields are optional.

• topic is a semicolon-delimited list of topics/areas/keywords, such as:

              topic = "machine learning; information retrieval"
  

• acceptance can be anything you want; it is intended to hold some sort of indication of the venue acceptance rtate, such as

              acceptance = "30%"
  

• pdf, ps, psgz and html generate links to to PDF, Postscript, gzip'ed Postscript and HTML, respectively. For example,

              psgz = "http://www.blah.com/foo/fizz.ps.gz
  

  (If you need more, it's simple to add more formats; see %filetypes in yab2web.)

Limitations

Due to a bug/issue in btOOL, yab2web doesn't work quite right with DOS-style line termination (\r\n). If you notice peculiar behavior (Javascript complaints about non-terminated strings, and/or strange people such as "and Fred Smith" in the author menu), then your BibTeX file might suffer from DOS-style line termination. Try converting your file to Unix-style termination (\n). (For example, in Emacs, use "ctrl-X RET f unix RET".)

The query/filter mechanism seems to work fine in Firefox and MSIE, but I think it doesn't work in Safari and I am not sure about other browsers. I'm working on it! In the meantime, please tell me of any strange behavior you might discover.

Nicholas Kushmerick / 22 May 2005