问题描述
我正在使用的ctypes实现，并且一切正常好吧，除了我无法弄清楚如何处理此部分：
I am using a ctypes implementation of CreateProcessWithLogonW, and everything works well except I cannot figure out how to handle this section:
一个环境块由一个以空终止的，以空终止的字符串组成的块组成。每个字符串的格式如下：
An environment block consists of a null-terminated block of null-terminated strings. Each string is in the following form:
name = value\0
name=value\0
要生成原始字符串，请执行以下操作：
To generated the raw string I execute the following:
lpEnvironment = '\0'.join(['%s=%s' % (k, os.environ[k]) for k in os.environ]) + '\0\0'
print lpEnvironment
'XAUTHORITY=/home/username/.Xauthority\x00MUTT_DIR=/home/username/.mutt\x00LASTDIRFILE=/home/username/.lastpwd-geany\x00LOGNAME=username\...\x00\x00'
但是我还是用它做一个ctypes变量，它会截断信息：
However run I make a ctypes variable out of it, it truncates the information:
ctypes.c_wchar_p(lpEnvironment)
c_wchar_p(u'XAUTHORITY=/home/username/.Xauthority')
如何正确传递lpEnvironment信息？
How can I pass the lpEnvironment information correctly?
推荐答案
据我所知，整个字符串确实在一个方向上正确地通过了 ctypes 边界，但是在返回的途中被截断了。
As far as I can tell, the whole string does get passed across the ctypes boundary correctly in one direction, but gets truncated on the way back.
>>> ctypes.create_string_buffer('abc\0def').value
'abc'
>>> ctypes.create_string_buffer('abc\0def').raw
'abc\x00def'
不幸的是（至少对我来说，Linux上的Python 2.6.5） create_unicode_buffer 的结果没有 .raw 属性。但是，
Unfortunately (at least for me, Python 2.6.5 on Linux) the result of create_unicode_buffer doesn't have a .raw property. However,
>>> ctypes.wstring_at(ctypes.create_unicode_buffer(u'abc\0def), 7)
u'abc\x00def'
工作正常。
这篇关于Python Ctypes空终止字符串块的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！