✅ Add more tests for the behaviour of rich_markup_mode
#964
Conversation
| @pytest.mark.parametrize( | ||
| "mode,lines", | ||
| [ | ||
| ("markdown", ["First line", "Line 1 Line 2 Line 3", ""]), | ||
| ("rich", ["First line", "Line 1", "", "Line 2", "", "Line 3", ""]), | ||
| ("none", ["First line", "Line 1", "Line 2", "Line 3", ""]), | ||
| ], | ||
| ) | ||
| def test_markup_mode_newline_pr815(mode: str, lines: List[str]): |
There was a problem hiding this comment.
Here, the question is mostly about the newlines: should there be one after "First line" or not?
Further, should the markdown parser output "Line 1 Line 2 Line 3" as one string, or split them into multiple strings, similar to the current behaviour for the option none?
| ), | ||
| ], | ||
| ) | ||
| def test_markup_mode_newline_issue447(mode: str, lines: List[str]): |
There was a problem hiding this comment.
Here, the splitting of the strings is different for all 3 options, with rich keep each input line as such, markdown merging everything that is not header, and none keeping all double newlines as separators.
| pytest.param( | ||
| "markdown", | ||
| ["First line", "", "• 1", "• 2", "• 3", ""], | ||
| marks=pytest.mark.xfail(), | ||
| ), |
There was a problem hiding this comment.
This test currently fails on master, with the output being
['First line', '', '• 1 - 2 - 3', '']
-> this happens mostly because the newlines are not parsed correctly, after which the bullet points aren't recognized correctly. PR #815 attempts to fix this.
| pytest.param( | ||
| "markdown", | ||
| ["First line", "", "• 1", "• 2", "• a", "• b", "• 3", ""], | ||
| marks=pytest.mark.xfail(), | ||
| ), |
There was a problem hiding this comment.
This test currently fails on master, with the output being
['First line', '', '• 1 - 2 - a - b - 3', '']
-> this happens mostly because the newlines are not parsed correctly, after which the bullet points aren't recognized correctly. PR #815 attempts to fix this.
There's a few open PRs around markdown formatting and parsing that I want to review, but first I would like to establish what exactly is the desired outcome for different docstrings and different settings of
rich_markup_mode.This PR is supposed to capture the ideal output for each test, using a
pytest.mark.xfailmarker for those cases that currently fail onmaster. Any potential fixes will then be contributed in subsequent PRs.