DM-42257: Make the text type un-sized so length is not required

[JeremyMcCormick]

The length should not be required when defining a text column, since it represents an unbounded string. The is_sized flag in Felis forces a column with that type to require a length, which we do not want for this datatype.

[JeremyMcCormick]

@gpdf proposes changing this line to "char" instead of "unicodeChar" on this PR.

class Unicode(FelisType, felis_name="unicode", votable_name="unicodeChar", is_sized=True):

[JeremyMcCormick]

@gpdf proposes changing this line to "char" instead of "unicodeChar" on this PR.

class Unicode(FelisType, felis_name="unicode", votable_name="unicodeChar", is_sized=True):

@gpdf I've changed this to "char" on this PR for the Felis text type.

[JeremyMcCormick]

@andy-slac I incorporated a number of your unrelated changes (though not all) from #51 in this.

[andy-slac]

Looks great! And I mentioned it already - we need unit tests for everything to cover all type-related quirks.

[JeremyMcCormick]

we need unit tests for everything to cover all type-related quirks

I'll keep this in mind and add more on another ticket/branch.

There are a decent number of tests for the type system now but maybe not covering all corner cases.

[gpdf]

@gpdf proposes changing this line to "char" instead of "unicodeChar" on this PR.

class Unicode(FelisType, felis_name="unicode", votable_name="unicodeChar", is_sized=True):

@gpdf I've changed this to "char" on this PR for the Felis text type.

The actual change you made was on the class Text() line, not on class Unicode(), and that's correct, that's exactly what I had in mind: fix Text() and leave Unicode() alone.

[gpdf]

Thanks, looks good.

[JeremyMcCormick]

@gpdf proposes changing this line to "char" instead of "unicodeChar" on this PR.

class Unicode(FelisType, felis_name="unicode", votable_name="unicodeChar", is_sized=True):

@gpdf I've changed this to "char" on this PR for the Felis text type.

The actual change you made was on the class Text() line, not on class Unicode(), and that's correct, that's exactly what I had in mind: fix Text() and leave Unicode() alone.

My mistake on this comment - I actually changed the line with class Text and not class Unicode.