Adding LLM Naming strategy

[Pgarrett]

Test cases can now be named using an LLM, EM will have to be set up using LLM config. Naming strategy involves prompting the LLM stating it should behave as a software engineer specializing in test naming. The prompt will include:

• Programming language being used (in order to follow naming conventions)
• Max amount of characters the new name can have
• Already generated test case names (to avoid repetition)
• The test case source code

Upon LLM response, a simple validation takes place in which we verify the response does not contain any illegal character, whitespaces and if it honours the max amount of characters the test name should have. If any of those validations fails, the LLM is re-prompted to provide a new name which follows the guidelines.

No disambiguation phase is added since that's already handled when prompting the LLM for the test case name.

Test case source code is generated at this stage only to provide the LLM with context. Since generating source code is just concatenating strings, we can accept the cost of it being called here and later on for the real test case writing.