D120.de/forum

I noticed that the argument for CFG generation is now called -cfg instead of the old -cfgfile. Does this mean that our tool should still support generating CFGs for programs in the "old" language, i.e. not enclosed in "begin" .. "end" and not including procedures? At least in my implementation, this would cause significant redundancy in the code, since otherwise I'd just modify CFG generation to support only the new program format and not care about the old CFG generator. However, I can see no benefit in supporting this, since the old versions of our submissions can still generate the CFG for such programs.

Similarly, is it okay to drop support for -ids03j from the program to avoid having to adapt the data sanitization analysis as well?

DanielSchoepe hat geschrieben:I noticed that the argument for CFG generation is now called -cfg instead of the old -cfgfile.

No, the command line parameters are as in Assignment 2 (Task 2), cf. also the section "Tool Usage" on the web page of the lab course.

DanielSchoepe hat geschrieben:Does this mean that our tool should still support generating CFGs for programs in the "old" language, i.e. not enclosed in "begin" .. "end" and not including procedures?

Your current version of the tool does not need to support the "old" language.

DanielSchoepe hat geschrieben:Similarly, is it okay to drop support for -ids03j from the program to avoid having to adapt the data sanitization analysis as well?

Yes, it's OK to remove the data sanitization analysis.