Description:

The Unicode codepoints U 2028 (Line Separator) and U 2029 (Paragraph Separator) are kind of treated as line breaks by Text.pack() but as normal zero-width symbols by Text.render(), instead of omitting the codepoints, breaking the text and padding the line as necessary.

I said "kind of" because pack() returns the width as though the text is broken by the codepoint(s) but the height is always 1, regardless of how many of these codepoints there are in the text.

By the way, urwid.calc_width() treats them as zero-width symbols.

Affected versions (if applicable)

• master branch (specify commit)
• Latest stable version from pypi
• Other (specify source)

Steps to reproduce (if applicable)

Text.pack():

>>> urwid.Text("123\u202812345").pack()
(5, 1)
>>> urwid.Text("123\u20281234").pack()
(4, 1)
>>> urwid.Text("1234\u20281234").pack()
(4, 1)
>>> urwid.Text("12345\u20281234").pack()
(5, 1)
>>> urwid.Text("123\u20281234\u202912").pack()
(4, 1)
>>> urwid.Text("123\u202812\u20291234").pack()
(4, 1)

Text.render():

>>> urwid.Text("123\u202812345").render(()).text
[b'123\xe2\x80\xa812', b'345  ']
>>> urwid.Text("123\u202812\u20291234").render(()).text
[b'123\xe2\x80\xa81', b'2\xe2\x80\xa9123', b'4   ']

Expected/actual outcome

I honestly care less whether they're treated as actual line breaks or just plain zero-width symbols and escaped (replaced with "?") like other whitespace codepoints such as U 0009 (Horizontal Tab, \t), all I care about is that they're handled consistently.

Thank you 😃