You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This modifies two existing test cases to include an assertion about
the length. These test cases are retained as there is value in
testing on output from a real git command rather than only with
test doubles.
More importantly, this also adds a parameterized test method to
check parsing of:
- All numeric, shorter than the limit - all fields used.
- All numeric, at the limit - all fields used.
- All numeric, longer than the limit - extra fields dropped.
- Has unambiguous non-numeric - dropped from there on.
- Has ambiguous non-numeric, negative int - dropped from there on.
- Has ambiguous non-numeric, number+letter - dropped from there on.
The cases for parsing when a field is not numeric (or not fully or
unambiguously numeric) currently all fail, because the existing
logic drops intermediate non-numeric fields (gitpython-developers#1833).
Parsing should instead stop at (or, *perhaps* in cases like "2a",
after) such fields. When the code is changed to stop at them
rather than dropping them and presenting the subsequent field as
though it were a previous field, these test cases should also pass.
0 commit comments