cfa151f032
* Archived many refs from OMEGA.bib into archiv.bib
* Those archived included many refs of that were of interest in my previous scientific research, but that are less pertinent to my current research. So the big single database file blob was parsed down to a smaller one (maybe 30-50% less lines).
* Performed manually, over course of a workday, but this was sped up immensely with a help of a keyboard shortcut associated to a vim macro that with a single keystroke cuts and appends the bib reference to an external file and then sets cursor to the next ref. Had worked out this macro functionality this past weekend. This made navigating from reference to reference, for a quick scan of the title and abstract, and making a decision, sometimes in less than 0.5s, other times taking several secs to make a decision.
* Used the following mapping in vim:
:map <F2> :norm@a<CR>:delete<CR>:silent !./cutbib<CR>:norm@b<CR>
where cutbib is a file containing `wl-paste >> archiv.bib`
and @a=v% and @b=n so that search next leads to a visual
delimiter-to-delimiter highlight, giving the lines for the cut
(delete and cutbib) operation.
bibd : A bibliographic reference database repo
omega.bib : A bibtex data file containing scientific source literature curated by the ackmanlab
History
Long ago some of this data started as a sequence of non-open library files associated with proprietary programs (endnote, sente in the early-mid 2000s). Once in awhile some of the data would explore other possibilities (papers, citeulike, mendeley, bookends, zotero). But nothing quite fit the combination of truly open/non-proprietary and flexible/hackable/future proofable.
Then the data was freed. It turned into a plain text/open/future proof file of a bibtex type format. And there the data has lived happily for ten years, being used with python routines or pandoc for conversion, integration to document prepartion formats intended for web or print publishing. It has enjoyed being managed with the excellent open source BibDesk program long developed for mac osx/macos. It has been also managed by the equally powerful and feature rich jabref program, which is also open source but runs on any platform (as a java application) including linux. In recent times the management has been minimzized (simplified meaning increasingly less external dependencies) with fuzzy-search and bash scripts at the command line.
Methods for transforming the file url links from BibDesk (which were macos specific) are detailed in a separate document '2018-11-01-bibdesk-convert.md'.