- Feb 17, 2020
-
-
Xun Yang authored
-
- Feb 16, 2020
-
-
Xun Yang authored
-
- Feb 12, 2020
-
-
Nicholas Robinson authored
- can now successfully check that class has at least one constructor - added rule fixing `this.x = 123;` in `J1_publicfields.java` - cleaned up Test.py a bit more
-
Nicholas Robinson authored
-
Nicholas Robinson authored
- added AstBuilding.py - new AstNodes.py that stores all Nodes and different types that we'll create - cleaned up Parsing.py heavily, now returns the parse tree - cleaned up Test.py as best I could to accomodate everyone - removed not needed file utils.py - commented out my code in Weeding.py
-
- Feb 11, 2020
-
-
Nicholas Robinson authored
-
- Feb 10, 2020
-
-
Nicholas Robinson authored
-
Nicholas Robinson authored
-
pycsham authored
-
pycsham authored
-
pycsham authored
-
pycsham authored
-
Xun Yang authored
-
pycsham authored
-
Xun Yang authored
-
Xun Yang authored
-
Xun Yang authored
-
pycsham authored
-
Xun Yang authored
-
Xun Yang authored
-
pycsham authored
-
pycsham authored
-
Xun Yang authored
-
Nicholas Robinson authored
-
pycsham authored
-
pycsham authored
-
pycsham authored
changed cfg to handle error cases: array intialize to empty brackets, and clean up cfg for methodmod
-
Nicholas Robinson authored
- casts & castExprType - instanceof - method and field invocs - params separated by commas - public abstract class
-
Xun Yang authored
-
Xun Yang authored
-
pycsham authored
-
Xun Yang authored
-
-
pycsham authored
-
Nicholas Robinson authored
- added new methodOrField thing that lets you invoc methods and fields recursively infinitely - statement can now declare and define any type - removed arrayID nonterminal - adjusted arrayAccess because of arrayID removal - added a cmd in README to make my workflow more convenient - commented out all the many many many prints in Test.py
-
pycsham authored
Made changes to scanner and grammar after going over around 3/4 of error public test cases. (e.g. bitwise operations, integer range checking)
-
Nicholas Robinson authored
- methodDcl is now one rule (rather than splitting into abstractMethodDcl & normalMethodDcl) had to do this because they were too similar, getting conflicts as abstractMethodDcl ends with SEMICO and normalMethodDcl ends with methodBody which has a rule `methodBody SEMICO` - added rule `statement primitiveType variableDcl SEMICO` because we didn't have a way to parse 'int x = 0;' or similars, it should actually be `statement type variableDcl SEMICO` but that causes conflicts in the grammer with `arrayAccess`
-
- Feb 09, 2020
-
-
Nicholas Robinson authored
- condensed getting the last rule, no more need for copying the stack - children are now right way round, not reversed
-