python-urwid/python-urwid-test_vterm-NUL.patch
David Cantrell 24b99e9bcf Apply upstream fix for EINTR handling in test_vterm test case
Apply upstream fix for test_remove_watch_file flakiness
Apply upstream fix for test_vterm handling of NUL characters
2017-11-15 14:27:31 -05:00

48 lines
1.7 KiB
Diff

commit 701138a380fac06023e5915448af92ba13614cb9
Author: aszlig <aszlig@redmoonstudios.org>
Date: Thu Feb 11 03:14:07 2016 +0100
vterm: Fix handling of NUL characters
According to the VT100 programmers manual, the NUL character has to be
ignored (at least on our side, because we are not a printer):
http://vt100.net/docs/tp83/appendixb.html
According to the bug reporter the VMS console driver inserts NUL
characters after line feeds and our implementation prints those as "?".
Tested against Python 2.7, 3.2, 3.3, 3.4 and 3.5.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Reported-by: Robert Urban <urban@unix-beratung.de>
diff --git a/urwid/tests/test_vterm.py b/urwid/tests/test_vterm.py
index 59fe166..4dadfcc 100644
--- a/urwid/tests/test_vterm.py
+++ b/urwid/tests/test_vterm.py
@@ -143,6 +143,10 @@ class TermTest(unittest.TestCase):
self.write('1\n2\n3\n4\e[2;1f\e[2M')
self.expect('1\n4')
+ def test_nul(self):
+ self.write('a\0b')
+ self.expect('ab')
+
def test_movement(self):
self.write('\e[10;20H11\e[10;0f\e[20C\e[K')
self.expect('\n' * 9 + ' ' * 19 + '1')
diff --git a/urwid/vterm.py b/urwid/vterm.py
index cc4eb7f..0f091ea 100644
--- a/urwid/vterm.py
+++ b/urwid/vterm.py
@@ -671,7 +671,7 @@ class TermCanvas(Canvas):
self.widget.beep()
elif not dc and char in B("\x18\x1a"): # CAN/SUB
self.leave_escape()
- elif not dc and char == B("\x7f"): # DEL
+ elif not dc and char in B("\x00\x7f"): # NUL/DEL
pass # this is ignored
elif self.within_escape:
self.parse_escape(char)