Bug 1039558

Description of problem:
New facts are assigned to variables before inserting into WM so that they can be initialized via setter methods. Variable names are generated by GRE and can be later changed by user. GRE generates names like fact0, fact1, etc.

The problem occurs after opening an existing rule that contains inserted fact assigned to fact0 variable. GRE doesn't recognize that fact0 is already used and starts generating variable names from fact0 again. When user adds another fact, it is assigned to fact0 and so fact0 is declared twice. This seems to be valid as long as the two facts are of the same type. However, if the *types are different*, the declaration is erroneous. Even so, nothing appears to be wrong until user validates the rule or project build is triggered. This can be easily fixed by assigning the new inserted fact to a different variable, but the validation (or build) error is very unexpected and confusing.

Version-Release number of selected component (if applicable):
ER5

How reproducible:
-

Steps to Reproduce:
1. create new rule
2. insert Fact1 and set value of one of its fields
3. save the rule and close it
4. reopen the rule
5. insert Fact2 and set value of one of its fields (notice that Fact1 and Fact2 are different types)
6. check rule source and validate it

Actual results:
Variable fact0 declared twice with different types. This is generated by GRE, user may save the rule without noticing if validation is skipped.

Expected results:
GRE should keep the rule valid when generating DRL source. It should detect that variable fact0 is already declared and choose next available name.

Additional info: