This is at proof of concept stage (quick and dirty), with these objectives
- auto generate a directory/package structure that the backend will happily parse
- auto generate the test steps as acucurately as possible (*)
- ensure all the relevant information makes it into the files for easier manual intervention
- ensure a 1-1 test step generation (or as close as possible) even if names are meaningless
##Issues
- Inconsistent numbering of test steps - requires manual regex changes
- Inconsistent table structure - requires corrective work on input files
If test cases were machine readable, the task of autogeneration would be trivial They could also be used to generate test plan docs
The dir_gen.py script was used to generate directories tcdd, tcdm and tcsc
as well as the subdirectories and package glue for each test case.
It also produces file templates for each test case
As an example to create the tcsc directory run
cat tcsc.txt | ./dir_gen.py
tcsc.txt is generated by cutting and pasting the test case list from the rendered version of Test Case List
The test_gen.py script generates the create_test_steps method and the body of the execute method. It's prone to failure because of the inconsistent nature of the test procedures
tcsc/TCSC4_8/tcsc4_8.py was generated by
cat TCSC4_8.txt | ./test_gen.py
TCSC4_8.txt was created by cutting and pasting the rendered version of Test Procedure into a text file
The command will print out create_test_steps and the body of the execute method which can then be cut and pasted into the correct test.
##Other issues Many manual steps creates the opportunity for mistakes The python is hacked together POC - python experts should be able to do it better. As a basis for progress it may be best to start anew