Wiki page for notes and development of the BRICCS Study Onyx questionnaire.
The production version of the BRICCS Study Onyx Questionnaire has the following stages:
1 - Acute Verbal Consent for samples collection 2 - Participant Consent 3 - Paper Consent 4 - Recruitment Context 5 - Risk Factor 6 - Patient-reported Medical History 7 - Samples Preliminary 8 - Blood Samples Collection 9 - Urine Sample Collection 10 - End of Patient Contact 11 - Notes-recorded Medical History 12 - Data Submission 13 - Conclusion
Versions are managed on the OnyxVersions page. Each version should have a trac ticket created to list and link to the changes (each change has its own ticket) scheduled for that version.
Questionnaire is edited using eclipse and svn. We are currently editing the language files in src at the same time as editing the questionnaire building java files. After building the questionnaire, the language file in the target is not automatically overwritten by the one we edit in source. This means we have a mismatch - the language files in source are correct, those in target are not, while the questionnaire.xml in target are correct, and those in source are not.
Important to test with html output from questionnaire building process before doing anything else. Currently we are rebuilding the entire onyx artifact each time we create new questionnaire files. To do this, after building the questionnaire, copy each questionnaire.xml file from the relevant /target/questionnaire/ folder to the relevant /src/ folder, and then run 'mvn clean install' in top level of the project workspace. This builds locally to the maven repository on your own machine. Then run the shell script to deploy the artifact. Note that mvn must be able to ssh to the mvn.briccs.org.uk server, which means that 'mvn.briccs.org.uk' must be in the ssh known_hosts.
Need to consider refining the process to separate the questionnaire from the code, in the way suggested on the Obiba wiki. Then versioning each questionnaire stage separately, and the code separately again.
Need to begin testing the questionnaire for 'healthy control' recruitment. Test onyx deployment is established. Jason has supplied a list of 'patient IDs' that are current in the test PMI so that we can use them. The list is in the OnyxTestSystem page.
There is a process for generating an html output of the questionnaire using the QuestionnaireRenderer class, but it is messy to say the least.
As a minimum, you need to have the following project checked out (or an equivalent for a subsequent version of Onyx):
Then the following *should* result in an html version of the questionnaires being generated. The first arg is the folder containing the questionnaire xml and the language files, and the second arg is the destination directory.
mvn -e exec:java -Dexec.classpathScope=test \ -Dexec.mainClass=org.obiba.onyx.quartz.core.engine.questionnaire.util.QuestionnaireRenderer \ -Dexec.args="/home/nick/workspace2/onyx-questionnaire/src/main/webapp/WEB-INF/config/quartz/resources/questionnaires /home/nick/briccs/html/"
A zipped file containing HTML prints of the questionnaire (v.1.9.2) is available here.