Programs & service programs #5
Conversation
pykit/tests/test_program.py
Outdated
| def test_rainbow(self): | ||
| int = Integer("aint8", 3, 1) | ||
| float = Float("afloat", 4, 2, 5.55) | ||
| char = Character("achar", 32, "A") |
There was a problem hiding this comment.
int and float are built-in Python functions, which you've now just re-defined. Might want to use different variable names here.
pykit/types.py
Outdated
| self.name = str(name) | ||
| self.isReturn = False | ||
| self.byValue = False | ||
| self.payload = {"name":self.name} |
There was a problem hiding this comment.
These should all be prefixed with __ to ensure they are private only to the class/subclass. This will break ServiceProgram.add_return(), though.
| self.name = name | ||
| self.library = library | ||
| self.parameters = [] | ||
| self.payload = { |
There was a problem hiding this comment.
If these are going to be private fields, we should think about indicating as so with a preceding single or double underscore: https://docs.python.org/3/tutorial/classes.html#private-variables
Not something to hold this up, but something to think about.
| """ | ||
| Object for IBM i Decimal. | ||
| """ | ||
| def __init__(self, name, length, decimals, value, signed=False): |
There was a problem hiding this comment.
Huh, apparently indeed the s does stand for "signed". There's also an "l" and "r" types which have a leading or trailing sign digit separate from the numeric data (I wonder if anyone actually uses those?). I don't know of anybody that refers to them as "signed decimal" though (could be showing ignorance here) - I've always heard it referred to as zoned decimal. Indeed the the free-format names use ZONED and PACKED. Perhaps instead, we just flip the condition and use packed=True?
Alternatively instead of a parameter, use separate classes: ZonedDecimal and PackedDecimal (or Numeric and Decimal) You could use this class as an internal base class.
| """ | ||
| Object for IBM i Float. | ||
| """ | ||
| def __init__(self, name, length, precision, value): |
There was a problem hiding this comment.
Can a float have a precision? I think it has to be specified as 0. If so, just like the Integer class, leave it off the parameter list.
| @@ -0,0 +1,157 @@ | |||
| class Parameter: | |||
| def __init__(self, parameterType, direction='inout', byValue=False, isReturn=False): | |||
There was a problem hiding this comment.
The way dbsock handles return values is weird and unintuitive. A return value by definition cannot be an input, so why does it have a direction? I think there should have been a separate way to specify the return value instead of trying to shoe-horn it in as a "parameter".
I'd like to see us hide this quirk as much as possible. Can we split out the return value handling in to a separate class, maybe ReturnValue and then add a new method add_return?
| """ | ||
| Object for IBM i Character. | ||
| """ | ||
| def __init__(self, name, length, value, varying=''): |
There was a problem hiding this comment.
Maybe a separate VarChar class instead of a parameter?
Please Marco the code