|
[printable version of this document (pdf)]| [COMP 304B Home]
A spreadsheet Number is specified by the following regular expression:
D [0-9]
E [eE][+-]?({D})+
Number [({D}+{E}?)
({D}*'.'{D}+({E})?)
({D}+'.'{D}*({E})?)]
From this specification, we derive the following Finite State Machine
|
Note how the specification not only describes how characters in the input stream trigger automaton transitions. It also describes actions to be taken upon transition. In particular, these actions set/update self.value and self.exp attributes to hold the mantissa and exponent respectively of the recognized number.
The scanner is encoded in the class NumberScanner in scanner.py. This requires an input stream class CharacterStream found in charstream.py.
A spreadsheet CellRef is specified by the following regular expression:
'$'?[a-zA-Z][a-zA-Z]?'$'?[1-9][0-9]?[0-9]?[0-9]?
From this specification, we derive the following Finite State Machine
|
Note how the specification not only describes how characters in the input stream trigger automaton transitions. It also describes actions to be taken upon transition. In particular, these actions set/update self.row, self.rowIsAbsolute, self.column and self.columnIsAbsolute attributes to hold appropriate integer values.
The scanner is encoded in the class CellRefScanner in scanner.py. This requires an input stream class CharacterStream found in charstream.py.
The test script tests.py produces the following output when the __trace__ variable is set to False. It produces the following output when the __trace__ variable is set to True.
Note how the scanner will only commit the part of the input stream which was recognized. The remainder of the input stream remains ready for future scanning. This is necessary as the different scanners will be used by a parser which will drive the scanner which will try to recognize different tokens as it recognizes a grammar (in this case, the spreadsheet formula syntax).
