feat: support for relative (percent) metrics in data contract checks.

[HenrikSpiegel]

• Tests pass (uv run pytest)
• Code formatted (uv run ruff check --fix && uv run ruff format)
• README.md updated (if relevant)
• CHANGELOG.md entry added

#1228

[HenrikSpiegel]

discussion:

I initially implemented this going for minimal changes and ensuring the changes matched the style of the repository.

I would however personally prefer to re-write the function using a operator mapping:

arg_to_op = {
  'mustBe' : '=',
  'mustNotBe': '!='
}
for arg, op in arg_to_op.items():
  if val := getattr(arg, quality) is not none:
    return f"{op} {val}{threshold_suffix}"

[jschoedl]

I get your point, but this won't work out for the 'between' cases, right?
The reason for this repetitive style is that it is easier to understand the code with all relevant cases, whereas a mapping as you proposed has more mental overhead to read. Of course this is a trade-off, since copying things like the {threshold_suffix} you added can be error-prone as well. But in this case, I'd vote for keeping it as is.

[jschoedl]

Thank you for the PR! A few notes, and please add test cases for this in the tests directory - e.g.:

• test that percent really appends %
• test that no % exists for the default case or rows
• test that e.g. check_property_null_values with percent produces the correct sodacl check
• test that a percent sign in check_row_count behaves correctly

[jschoedl]

After the warning, we should not create the check (Soda would ignore the percent sign and only use the integer, which is not what the user is intending)

[jschoedl]

Suggested change

	f"Unsupported quality.unit ={quality.unit} in quality check, must be 'rows' or 'percent' or None"
	)
	f"Unsupported quality.unit '{quality.unit}' in quality check, must be 'rows' (default) or 'percent'"
	)

None basically means that the value is not present and 'rows' will be used.

[jschoedl]

Suggested change

	if quality.unit is not None and quality.unit not in ("rows", "percent"):
	logger.warning(
	if quality.unit is not None and quality.unit.lower() not in ("rows", "percent"):

[jschoedl]

Suggested change

	threshold_suffix = "%" if quality.unit == "percent" else ""
	threshold_suffix = "%" if quality.unit.lower() == "percent" else ""

[jschoedl]

This still includes duplicate_count. (5 similar locations need to be fixed)