CUCC Expedition Handbook

Troggle Architecture

Troggle overall architecture

Troggle is made up of three large chunks:

• The website pages: the handbook, instructions, indexes, reference material, walking guides, logbook typescripts, reports to funding agencies, copies of expo reports in magazines, instruction guides for how to run an expo, accounting protocols for subsidising transport, rope maintenance standards: over 900 pages in all. What you are reading now is part of this.
  
  
• The file archive: the photo galleries, the copies of PDF reports, the scans of original survey notebooks, GPS tracks, maps, aerial photos, scanned images of our handwritten expo logbooks. There are 40 GB of files.
  
  
• The troggle-generated reports: the tables of survey trips for each cave, the lists of scanned surveys for each survey trip, who was on what trip, calendar lists of logbook entries corrleated with recorded survey trips, who surveyed what caves, what source material was used to create the final cave survey images, who went on which expeditions (and who didn't go caving). There are over 2,000 webpages generated using over 6,000 lines of bespoke python code from over 90 different urls.

Troggle data architecture

The core of the troggle software is the data architecture: the set of tables into which all the cave survey and expo data is poured and stored. These tables are what enables us to produce a large number of different but consistent reports and views.

Also there have been tables added to the core representation which are not included in this old diagram, e.g. Scannedimage, Survexdirectory, Survexscansfolder, Survexscansingle, Tunnelfile, TunnelfileSurvexscansfolders, Survey. See Troggle data model python code (15 May 2020) and click on the Class Diagram below on the right.

Purpose

The reasons why we have an online system at all are described in our website history.

There is an introductory article "Troggle: a revised system for cave data management".

Implementation in software

Troggle is written in Python (about 6,400 lines excluding comments) and is built on the Django framework. Before starting to work on Troggle it might be a good idea to run through an initial install and exploration of a tutorial Django project to get the Django concepts bedded down - which are not at all obvious and some exist only within Django.

Django is the thing that puts the survey data in a database in a way that helps us write far less code to get it in and out again, and provides templates which make it quicker to maintain all the webpages. See the django design philosophy for why we chose it: while django comes with a full stack (db, request/response, URL mapping, HTML templates) the layers of the stack are independent and individually replaceable.

We have to keep up to date with new rleases of django, see Upgrading Django for Troggle.

Troggle parsers and input files

To understand how troggle imports the data from the survex files, tunnel files, logbooks etc., see the troggle import (databaseReset.py) documentation.

The following separate import operations are managed by the import utility (databaseReset.py):

• expo logbooks
• folk (people)
• wallets of notes & scans (surveyscans)
• cave survey data - survex files
• QMs - question marks
• drawings - tunnel & therion files

Files generated by troggle

There are only two places where this happens. This is where online forms are used to create cave entrance records and cave records. These are created in the database but also exported as files so that when troggle is rebuilt and data reimported the new cave data is there.

Helpful tools and scripts

The public server uses a MariaDB SQL database and development is usually done using a single-user sqlite database which is a standard django option.

You will find it very useful to see what is going on if you look directly at the data in the database (just a single file in the sqlite case) and browse the data in the tables. A light-weight, simple db browser is DB Browser for SQLite. Connecting directly the the MariaDB database with a control panel or workbench gives even more tools and documentation capabilities.