Original Research
- A database and suite for handling original research leads
* The program as stands is NOT safe for multi-user use *
* and does NOT have tests *
Design
- database (currently text files)
- classes (object-orientation used)
- api (expose manipulation to command line)
- automated_scripts (relieving some manual work)
- php files (manual interaction via dialog(1) and
shell_exec() calls to the api)
- or.sh (main entry point)
- /tmp/effort02/ Variety files used for transferring state
- pics/ Pictures for the manual
- ~/OR/scrapes/ Raw & processed data scraped from sites
- gexf/ Network maps in XML form of node & edges
- manual/ Documentation
or.sh <project name>
Main entry point for using Original Research. The project name if not
previously used will cause new files to be written in the data
directory.
Ctrl-C to exit
/data
Project files go in here, one subdirectory each.
/data/<project>/db
The db directory contains continuing state: a number of flat files
that contained serialised objects. These are used instead of an SQL
database at this point.
- things.serialised
The Things are entities that we're keeping data about: people,
podcasts, etc. We also keep data about who uploaded it (only me
at this point), timestamp of upload etc.
The tag is a per-file unique identifier for each entity.
- automated_cribs.serialised
Placeholder file for the moment: this data would be used to
keep track of when was the last time that a Thing was automatically
processed, as if it hasn't changed since, there may be no need to
process it again.
- reviews.serialised
As we research we might reach our own opinion on something: the
reviews maintain those notes. We can also leave an empty review
which merely marks the timestamp it was done.
Reviews answer the question: when was the last time I considered X?
- links.serialised
Two different entities might be linked: for instance Alice appeared
in a YouTube video. The tags for Alice and the video appear
together in this file. Data is stored as a triple: subject, predicate and
object.
/api
Helper files to interact with the database. Their expectation is to
be called from the command line, and so assist in automation.
/classes
The database is effected using object-oriented classes, and they
reside here.
/tmp/effort02/
Variety temporary files used to track state around between bash
and PHP. Each file contains one tag.
- were_looking_at (Last tag looked at)
- what_to_do_next (Next tag to show)
- break_link_from (Tags that should have any linkage be
- break_link_to removed)