Steps to Reproduce:
# Go to an item (eg. [[https://www.wikidata.org/wiki/Q4115189?uselang=cs|sandbox]]) with Czech (cs) as the interface language.
# Add a new property with time datatype (eg. [[https://www.wikidata.org/wiki/Property:P585|P585]]).
# Write a month in Czech (see below) and a year (eg. 2019).
Actual Results:
- For some months, the preview will show the first day of the month:
- Namely "březen" (3), "květen" (5), "červen" (6), "červenec" (7), "září" (9), "říjen" (10).
- The rest is parsed correctly: "leden" (1), "únor" (2), "duben" (4), "srpen" (8), "listopad" (11), "prosinec" (12).
- Possible explanation: All problematic months contain characters with [[ https://www.wikidata.org/wiki/Q26948 | caron ]]. It's possible that the parser doesn't handle special characters correctly.
- It could be caused by `strcasecmp` which is used in [[ https://github.com/wmde/Time/blob/0ce1e1fc10fbf1da232f88235adb609c924f3e79/src/ValueParsers/YearMonthTimeParser.php#L101 | YearMonthTimeParser ]]. But I wasn't able to confirm this.
- The same problem happens with all English months (still with the interface in Czech).
{F30380987}
{F30380993}
{F30381005}
Expected Results:
- The value is always parsed to the "month" precision.