diSimplex Citation Manager

A citation manager for the diSimplex projects which understands our NikolaBase citation system.

Top level tools

• capture to capture raw RIS formatted citation from Zotero/BibItNow into our new style database.

• scanner to scan one or more (LaTeX) *.aux files for a list of citations. The scanner creates a BibTeX file for use by bibtex/latex for a given project.

• websiteBuilder to export our internal new style database into external static HTML website. (Create an external version which only contains a sub-site specified by a simple list of citations (and their associated authors))

Solution

Reference extraction (for a latex document)

1. Extract "bib" item from our citation bank

2. Use Jinja2 or Pybtex! to format into a specific bbl (or bib) format.

Reference entry

1. use Zotero and/or bibitnow to capture a reference from a give webpage.

2. Copy the reference in RIS format from Zotero and/or bibitnow into our own NiceGUI based tool to add into our own database. We can adjust the reference information (citeKey) to suit our needs in the NiceGUI editor before saving.

Reference database

Our reference database will be:

1. A primary collection of author and citation YAML databases (more than one author/citation per file). This will be a collection of human readable text files which can be version controlled.

   We will need a mapping from RIS to our form of YAML (possibly a 1-1 mapping).

2. A primary collection of author/citation Markdown files containing extra RIS notes. The markdown files will be named using the primary collection's keys.

3. A secondary SQLite3 database (re)built from the primary collection, used for searching for existing matching authors and/or citations.

Notes: the primary collections will have a multiple directory structure using the first two characters of each author/citation keys.

Online access

We will using Jinja2 templates to expand our reference database into a collection of static HTML files to be used as our online reference system.

We can have two different "versions" of the templates to provide an external as well as an internal variant.

Tasks

Reference extraction

1. build a (sorted) .bib file from our reference system (our own python script)

2. create a .bbl file from our .bib file (bibtex (no utf-8) or bibulous(python) (utf-8)?)

3. use natbib/amsref/raw-bbl-file

Phase update of existing citation system

1. Start using Zotero/bibItNow.

   Need to create an ultra simple NiceGui capture tool, possibly simply saving the raw RIS as a file for each citation.

2. Write collector to collect our existing data from old citation system and transform into new style database.

3. Merge recently captured citations into new style database.

4. Expand NiceGui capture tool to use new style database.

Resources

LaTeX citation systems

• natbib (does not create a "Mathematical" citation)

• amsref (Not used by TAC, not allowed in ACM, but is probably allowed in AMS journals; this is the format used in most mathematical citations)

• Embedded bbl-files (\bibitems)

TeX Tools

• bibtex (old useful reference but we won't use)

• biblatex / biber (Not allowed in ACM)

Python tools

• MrTango/rispy: Python RIS files parser, provides RIS files as dictionary via generator.

• Pybtex can be used to create a .bbl which can then be included in most journals

• pybtexris: A pybtex plugin for inputting and outputting RIS files

• nicegui: Create web-based user interfaces with Python. The nice way.

Browser tools

• Zotero Style Repository

• Zotero | Your personal research assistant

• bibitnow: Site adjustors for browser plugin "bibitnow"

Questions

• how easy would it be to combine our own python script with parts from pybtex/pybtexris/rispy?

• how easy would it be to hand-translate an existing .bst file to a pybtex style file?