Source Files for Personal Website of Thomas Graf

This is the repository for my personal website. The Master branch holds the source files, the gh-pages branch is used for the output files. The source files are simple markdown files that are automatically converted to HTML5 via the Python-based Pelican and then uploaded to Github with ghp-import.

The setup is pretty vanilla in that I do not use any plugins or fancy extensions, except for tagcloud. I made some modifications to the blueidea theme, and I rolled my own shell scripts to automatically convert my bibtex bibliography into 1) a nicely formatted list of publications and 2) individual blog entries for each bibtex entry. Those scripts are in the bib folder, and their basic usage is explained farther down.

System Requirements

If you're affiliated with Stony Brook University, you can download a virtual machine image that comes with all necessary packages pre-installed.

1. Recent version of Python2.

2. Pelican 3.6 or higher. Under Linux, this might already be in your repository.

   • In Debian Jessie you have to activate the backports repository to get a recent version.
   • In Ubuntu 14.04 (LTS) you'll need to find a ppa or use the Mac/Windows method.
   • On OSX and Windows (and Linux, if necessary) you can install Pelican via Python's package manager pip.

   pip install pelican
   

3. The ghp-import script. Again you can install it via pip

   pip install ghp-import
   

   or use your package manager if you're on Linux.

4. Python's typogrify library. As before, use pip or your package manager.

5. A recent version of pandoc-citeproc (1.17 or newer) if you want to use my bibtex converter.

File Structure

Once you've cloned the repository, you'll see a couple of files and folders. The most important ones are:

• content: the markdown files for the website
• output: the actual website generated by Pelican
• pelicanconf.py: contains various settings for site creation; for your own site, you'll want to change all the variables at the top, plus LINKS and SOCIAL further down.
• publishconf.py: allows you to override certain settings in the final creation phase; for your own site, change at least the value for SITEURL.
• Makefile: set of instructions for building the website; you might want to change the value for GITHUB_PAGES_BRANCH depending on your Pelican workflow

In addition, the bib folder contains a number of shell scripts (fully sh-compatible), and a subfolder with specific info on each bibtex entry.

Shell Scripts

All the shell scripts use mostly standard commands such as cat, grep, sed, and awk, and they do not use any bashisms. The one exception is bib2mdown, which uses pandoc-citeproc to convert the bibtex file into a markdown file using the style sheet mylanguage.cls. Since bib2mdown is called in one way or another by every other script, you absolutely need to have pandoc installed.

Here's a description of each script:

1. bib2mdown takes a bibtex file as input and returns a markdown file. The output produced by pandoc is pretty bad, so a major dose of sed magic is applied after the initial pandoc-conversion step. The following tweaks are applied:

   • add doi links
   • rip out all other links since doi is more reliable
   • add link to pdf, if it exists
   • add link to detailed blog entry
   • list can be bulleted, numbered, or reverse numbered
   • remove remnants of Latex commands
   • fix some formatting errors

2. bib2blog produces a blog entry for a given entry in a bibtex file (with the help of bib2mdown). It automatically assigns it the right category (books, talks, or papers) and saves it to the corresponding subfolder (talks for talks, and papers for books and papers). It also checks whether there is a folder with the same bibkey in auxfiles, and loads the abstract, tags, and date from the files in this folder, if it exists. It also looks for an archive with the source code in doc/talks or doc/papers and adds the corresponding link. All of that is then put together into a nice entry

3. create_bibliography takes two bibtex files as input, one for publications and one for presentations. It then uses bib2mdown to produce the Output page of the website.

4. create_bibkey runs bib2blog for every bibtex key in a bibtex file.

5. compile_references runs create_bibliography and create_bibkey on my bibtex files. If everything is configured correctly, this is the only script that needs to be run, all others will be called as needed.

All scripts except compile_references take command line arguments and can be altered via various options. Just run them with --help to learn more. You can also change the variables at the top of each script to adapt it to your use case.