[Xcb] [PATCH:libxcb] Fix inconsistent use of tabs vs. space.

[Thomas Klausner]

Needed for at least python-3.5.x.

Signed-off-by: Thomas Klausner <***@NetBSD.org>
---
src/c_client.py | 14 +++++++-------
1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/src/c_client.py b/src/c_client.py
index 57de3fb..043338d 100644
--- a/src/c_client.py
+++ b/src/c_client.py
@@ -1364,7 +1364,7 @@ def _c_serialize(context, self):
_c(' unsigned int xcb_align_to = 0;')
if self.is_switch:
_c(' unsigned int xcb_padding_offset = %d;',
- self.get_align_offset() )
+ self.get_align_offset() )
prefix = [('_aux', '->', self)]
aux_ptr = 'xcb_out'

@@ -1390,7 +1390,7 @@ def _c_serialize(context, self):
_c(' unsigned int xcb_align_to = 0;')
if self.is_switch:
_c(' unsigned int xcb_padding_offset = %d;',
- self.get_align_offset() )
+ self.get_align_offset() )

elif 'sizeof' == context:
param_names = [p[2] for p in params]
@@ -1930,14 +1930,14 @@ def _c_accessors_list(self, field):
# from the request size and divide that by the member size
return '(((R->length * 4) - sizeof('+ self.c_type + '))/'+'sizeof('+field.type.member.c_wiretype+'))'
else:
- # use the accessor to get the start of the list, then
- # compute the length of it by subtracting it from
+ # use the accessor to get the start of the list, then
+ # compute the length of it by subtracting it from
# the adress of the first byte after the end of the
# request
- after_end_of_request = '(((char*)R) + R->length * 4)'
- start_of_list = '%s(R)' % (field.c_accessor_name)
+ after_end_of_request = '(((char*)R) + R->length * 4)'
+ start_of_list = '%s(R)' % (field.c_accessor_name)
bytesize_of_list = '%s - (char*)(%s)' % (after_end_of_request, start_of_list)
- return '(%s) / sizeof(%s)' % (bytesize_of_list, field.type.member.c_wiretype)
+ return '(%s) / sizeof(%s)' % (bytesize_of_list, field.type.member.c_wiretype)
else:
raise Exception(
"lengthless lists with varsized members are not supported. Fieldname '%s'"

[Uli Schlachter]

Post by Thomas Klausner
Needed for at least python-3.5.x.

Same as the other two patches: Merged.

Cheers,
Uli

[Ingo Bürk]

Hi Uli,

haven't looked at the patches, but the rationale is certainly correct.
Those are some differences between Python 2 and 3. So should be fine.



Ingo

Post by Uli Schlachter

Post by Thomas Klausner
Needed for at least python-3.5.x.

Same as the other two patches: Merged.
Cheers,
Uli