Certain number comparisons have inverted results

Python 3.7.3 hotxlfp 0.0.11

Under the following conditions:

Two numbers are being compared using >, >=, <=, or <
The number on the left side of the comparison is greater than the number on the right
The two numbers are different number types (e.g., float and int)

The following shows the behavior in action:

>>> from hotxlfp import Parser
>>> p = Parser()
>>> p.on('callCellValue', lambda *_: 3)
<hotxlfp.parser.Parser object at 0x10dd8ddd8>
>>> p.parse('A1 > 3.5')
{'result': False, 'error': None}
>>> p.parse('A1 < 3.5')
{'result': True, 'error': None}
>>> p.parse('A1 > 2.5')
{'result': False, 'error': None}
>>> p.parse('A1 < 2.5')
{'result': True, 'error': None}
>>> p.parse('4.5 > A1')
{'result': False, 'error': None}
>>> p.parse('4.5 < A1')
{'result': True, 'error': None}

# this particular case is an exception. unsure why.
>>> p.on('callCellValue', lambda *_: 3.5)
<hotxlfp.parser.Parser object at 0x10dd8ddd8>
>>> p.parse('4 < A1')
{'result': False, 'error': None}
>>> p.parse('4 > A1')
{'result': True, 'error': None}

# comparing hardcoded numbers exhibits the same behavior
>>> p.parse('3 < 2.25')
{'result': True, 'error': None}
>>> p.parse('3 > 2.25')
{'result': False, 'error': None}
>>> p.parse('3.5 > 2')
{'result': False, 'error': None}
>>> p.parse('3.5 < 2')
{'result': True, 'error': None}

# comparing the same type works fine
>>> p.parse('3.14 > 2.25')
{'result': True, 'error': None}
>>> p.parse('3.14 < 2.25')
{'result': False, 'error': None}
>>> p.parse('3 < 2')
{'result': False, 'error': None}
>>> p.parse('3 > 2')
{'result': True, 'error': None}

# same behavior applies to cell values
# same type compares fine
>>> values = [[1, 3]]
>>> p.on('callCellValue', lambda cell, done: done(values[cell.row.index][cell.col.index]))
<hotxlfp.parser.Parser object at 0x10dd8ddd8>
>>> p.parse('A1 < B1')  # 1 is less than 3
{'result': True, 'error': None}
>>> p.parse('A1 > B1')  # 1 is greater than 3
{'result': False, 'error': None}
>>> p.parse('B1 > A1')
{'result': True, 'error': None}
>>> p.parse('B1 < A1')
{'result': False, 'error': None}

# different types have inverted results
>>> values = [[1, 2.5]]
>>> p.on('callCellValue', lambda cell, done: done(values[cell.row.index][cell.col.index]))
<hotxlfp.parser.Parser object at 0x10dd8ddd8>
>>> p.parse('A1 > B1')
{'result': False, 'error': None}
>>> p.parse('A1 < B1')
{'result': True, 'error': None}
>>> p.parse('B1 > A1')
{'result': False, 'error': None}
>>> p.parse('B1 < A1')
{'result': True, 'error': None}

Additionally, in exploring this bug, I found that if a hardcoded floating point number less than 1 does not start with 0, it errors. I am unsure if this is intentional:

>>> p.parse('A1 > .5')
{'result': None, 'error': '#ERROR!'}

aidhound / hotxlfp

Certain number comparisons have inverted results #6