babu-thomas/calculator-parser

Parser for simple arithmetic calculator

calculator-parser's Issues

Only expose Parser object to parse any input

Currently to parse a string the following needs to be done -

• Create InputStream object from newline delimited input string.
• Create TokenStream object from the InputStream object.
• Create Parser object from the TokenStream object.
• Call parse() function of Parser object.

It would be better to just expose a Parser object to the caller which takes just the input string and takes care of all other object creations inside it.

Don't make the caller pass newline delimited string

Currently to correctly parse an input string it needs to be delimited with a newline character. If the caller doesn't pass a correctly delimited string, the parser fails. Take care of delimiting input string inside InputStream class instead of relying on the user.

Successive operator tokens without space is lexed as single token

Running

parser = Parser(TokenStream(InputStream('  ***  ')))
print(parser.parse())

gives

Expected token of type: num, found token: {'type': 'op', 'value': '***'} (column: 5)
None

The 3 multiplication operators are lexed as one token. Expected output is

Expected token of type: num, found token: {'type': 'op', 'value': '*'} (column: 3)
None

where the operators are lexed individually.

Add tests

Currently the parser behaviour is only tested manually by trying different inputs. Adding unit tests for different cases will ensure the parser behaves correctly and continues to do so in the future.