Original website (i.e. Stew's version): http://web.uvic.ca/~lang02/bailey/
New development website (note: requires Netlink login): http://web.uvic.ca/~lang02/bailey_v2/
The data has gone through a few schema changes, all documented on the blog. See, in chronological order (i.e. earliest first):
For each set of data that SD sends, I do the following:
Once I get the file back from SD, I do the following:
As noted at the top of this document, the schema file is at http://web.uvic.ca/~lang02/bailey/schema/bailey_trialfile_proofing.rng . SD validates against this file so it always needs to be current.
On the search form, the Jquery plugin bsmSelect is used to make multiple select fields more user-friendly. The plugin homepage is here: http://plugins.jquery.com/project/bsmSelect , and the blog post is here: http://hcmc.uvic.ca/blogs/index.php?blog=36&p=7864&more=1&c=1&tb=1&pb=1
The site search – both the form and the results – is handled by the Search suite of classes located at includes/classes (Search.php and the Search subdirectory). The Search_Form class, located at Search/Form.php, is responsible for displaying all of the search fields and populating them with the proper values. All of the classes are fully documented.
The site uses the Zend Framework's DB library for database interaction. The classes are located at includes/classes/Zend . All of the classes used in the site follow the PEAR class naming convention for easy autoloading. The PEAR convention dictates the following:
Each subdirectory that a class is in should be represented in the name, with single underscores separating each level of the hierarchy. For example, if a class lives in Search/Form.php, its class name would be Search_Form . Similarly, if a class lives in My/Name/Is/Bob.php, its class name would be My_Name_Is_Bob . This convention has two primary benefits:
The search results pages – both the results table and a single record – use the Model_ series of classes to display information, as well as either Search_Result (for the results summary) or Search_Single (for a detailed record view). Each row of a table (trials, trial_files, etc.) is represented by a class, which allows for easy placement of formatting functions and the like.
The Chart / Data Visualization
The chart to visualize results uses Flot, a Jquery-based plotting tool:
To accomplish the 'stacking' of the results, I used the stacking plugin, which is included wtih Flot:
The Flot chart gets its information from a dynamic JSON dataset, which is based on a submitted search form. When a search is submitted, it's stored in the user's session for easy re-use on either the search page (to modify results) or the chart page. This is also how the “passing” of data from the chart to the results table (and vice versa) is accomplished – both just read and parse any search information in the session. The results themselves are not saved in the session, just the search query parameters.
The Flot initialization code is in js/init-flot.js – it's almost all 'stock' and taken from the examples, with the minor exception of a modification I had to make to get the totals for each year display under the year name in on the X axis of the chart.
I'm overall very happy with Flot, as is SD. It's fairly simple, decently fast, and, unlike some alternatives, still under active development.
No Pingbacks for this post yet...
Simon Devereaux has approximately 10,000 records of people convicted in potentially capital cases between 1710 and 1840 in London heard at the Old Bailey court. This project will create a web-based database which will allow interested researchers and members of the public to compose queries on that data (e.g. women charged with robbery 1710-1720). It must be able to support a range of queries and produce output allowing researchers to identify trends in judicial practice over that time.
|<< <||> >>|