uk-gov-mirror / openregister.kibitz Goto Github PK

  kibitz - gov.uk Picker Configuration Tools
  ==========================================

  Copyright (C) 2017, Andy Bennett, Crown Copyright (Government Digital Service).

 1. Introduction

	gov.uk Open Registers are a way of expressing an authoritative list
	that you can trust. This distribution is a set of tools for combining
	Register data with other data in order to produce a configration file
	that can be used to drive an interactive, multi-lingual, Location
	Picker that a user can use to easily select a valid Location from a
	list.

	There are a number of files supplied:

	+ picker-input.sample.scm

	    This file documents the format and structure of the configuration
	    file produced. The file starts with some description and then
	    provides some examples. Finally, the file includes some code so
	    that the whole thing can be interpreted by "csi" in order to
	    produce the examples in JSON format.

	+ picker-input.sample.json

	    This file shows the examples from picker-input.sample.scm in JSON
	    format. This is an example of the actual configuration file that
	    you would provide to the Location Picker.

	+ create-picker-input.scm

	    This is the main event! This script takes the Registers in .rsf
	    format and the spreadsheet in .csv format and produces a Location
	    Picker Configuration in .scm format.
	    This script does all the hard work to transform the data into the
	    format described in picker-input.sample.scm.

	+ picker-input-to-json.scm

	    This file is a copy of the code in picker-input.sample.scm that can
	    be used to convert a configuration file in .scm format to a
	    configuration file in .json format.

	+ visualise-picker-input.scm

	    This script can be used to convert a Location Picker Configuration
	    in .scm format into a .pdf file that visualises the configuration
	    for human consumption. The files produced by this are suitable for
	    printing on printers that can use incredibly large paper. This tool
	    is useful for debugging the decisions made by the Location Picker.

	+ fetch-register-feeds.sh

	    This script updates the .rsf files in the data/ directory.
	    The .rsf files contain a copy of the Register data.

	+ data/Location picker data - Data.csv

	    This file contains the non-Register data required to build a usable
	    picker. It contains things such as synonyms, endonyms and links
	    between the entites described by the Register.

	+ data/country.rsf

	    This is the Country Register in .rsf format.

	+ data/territory.rsf

	    This is the Territory Register in .rsf format.

	+ data/uk.rsf

	    This is the UK Register in .rsf format.

	+ data/FEEDS

	    This file documents where all the other files in data/ come from.

	+ Makefile

	    Here we script the tools to produce the outputs.

	The whole package is distributed under an MIT license and as such is
	free to use and modify as long as you agree to its terms.

	Note: Should you have any trouble in setting up and using these tools,
	please feel free to contact me, Andy Bennett at either
	[email protected] or, alternatively at
	[email protected].


 2. Installation

	The tools in this package can be run straight away; they do not need to
	be built or installed. However, they do have a number of dependencies
	which need to be satisfied.

	The tools in this package requires GNU Make and a recent CHICKEN. morc
	(from the orc distribution) is also required in order to process the
	Register data. morc has the same requirements: GNU Make and a recent
	CHICKEN.

	CHICKEN can be found at http://www.call-cc.org/ and can be build on
	Linux, MacOS X and Windows.

	The tools rely on a number of "eggs" from the CHICKEN ecosystem.
	Install them thus:

	    chicken-install medea csv-xml clojurian list-utils vector-lib utf8

	The visualisation tools require Graphviz.


 3. Usage

	In order to build a Location Picker Configuration using the supplied
	data, invoke make thus:

	    make all

	This will produce three files:

	+ out.scm

	    This is the Location Picker Configuration in .scm format.

	+ out.json

	    This is the Location Picker Configuration in .json format. This
	    file is suitable for loading into the Javascript part of the
	    Location Picker.

	+ out.pdf

	    This is the Location Picker Configuration rendered into a graphical
	    format that can be understood by a human. Use this for debugging
	    the decisions made by the Location Picker.

	The Location Picker Configuration uses a local copy of data from three
	registers: the Country Register, the Territory Register and the UK
	Register. You can update the local copies of the Registers like this:

	    ./fetch-register-feeds.sh

	This will update the .rsf files in the data/ directory. Once you have
	updated the Register data, you can regenerate the Location Picker
	Configuration by using "make" as described above.

	The Location Picker Configuration is customised by a spreadsheet
	containing all of the non-Register data such as synonyms, endonyms and
	mappings between locations. For example, Antarctica is a location in
	the Territory Register but it is actually claimed by a number of Nation
	States. The spreadsheet documents these claims such that when a user
	types "Antarctica" into the Location Picker they are offered a choice
	of one of the locations that exist there as well as an opporunity to
	choose the Nation States themselves. A link to the original version of
	this spreadsheet can be found in the data/FEEDS file. Alternatively,
	you can customise the local copy in data/Location picker data -
	Data.csv to satisfy your own tastes. If you add columns to this file
	then you will need to customise the parser in create-picker.input.scm:
	start with the definition of "col-spec" which declares the name and
	type of each column as well as what to do with it.


 4. Compatibility notes

	See the documentation in picker-input.sample.scm to learn about how the
	data has been structured such that it is possible to record telemetry
	about how the picker is used and then compare the data between
	different configurations that are produced by the tool. In particular,
	the "stable-name" property is supplied in order to make these
	comparisons both possible and easy.


 5. What's next?

        First and foremost, enjoy the tools and use and extend them to build
        your own pickers that incoporate register data.

	Please feel free to send tools that you build with this so that they
	can be integrated and distributed with this package.

        Suggestion, extensions and patches are welcome.

        If you have any questions or problems (even the slightest problems, or
        the most stupid questions), then please feel free to get in touch with
        me directly using the addresses above. I will try to help you, get you
        going or point you in the right direction.