blocks|key|158033|text|注意：OP为解决这个问题所做的就是将encoding='UTF-8'传递给open调用。如果你遇到这个问题，只是想找一个修复方法，这是可行的。这篇文章的其余部分是强调为什么。|type|unstyled|depth|inlineStyleRanges|offset|length|style|BOLD|CODE|entityRanges|data|158034|158035|会发生什么？|158036|正如卢卡斯所说，文档规定：|158037|158038|在某些系统上，需要调用setlocale()来获取用户首选项|blockquote|158039|158040|158041|最初，string.letters设置为返回lowercase+%2B+uppercase|158042|lowercase+=+'abcdefghijklmnopqrstuvwxyz'
uppercase+=+'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
letters+=+lowercase+%2B+uppercase|code-block|syntax|javascript|158043|但是，当您调用getpreferredencoding()时，_locale模块会在使用以下代码在fixup_ulcase(void)中生成它们之后，通过调用PyDict_SetItemString(string,+"letters",+ulo);来覆盖它：|158044|/*+create+letters+string+*/
n+=+0;
for+(c+=+0;+c+<+256;+c%2B%2B)+{
++++if+(isalpha(c))
++++++++ul[n%2B%2B]+=+c;
}
ulo+=+PyString_FromStringAndSize((const+char+*)ul,+n);
if+(!ulo)
++++return;
if+(string)
++++PyDict_SetItemString(string,+"letters",+ulo);
Py_DECREF(ulo);|158045|反过来，这是在PyLocale_setlocale中调用的，它实际上是setlocale，这里的getpreferredencoding代码调用http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l612：|158046|++def+getpreferredencoding(do_setlocale+=+True):
++++++++"""Return+the+charset+that+the+user+is+likely+using,
++++++++according+to+the+system+configuration."""
++++++++if+do_setlocale:
++++++++++++oldloc+=+setlocale(LC_CTYPE)
++++++++++++try:
++++++++++++++++setlocale(LC_CTYPE,+"")
++++++++++++except+Error:
++++++++++++++++pass
++++++++++++result+=+nl_langinfo(CODESET)
++++++++++++setlocale(LC_CTYPE,+oldloc)
++++++++++++return+result
++++++++else:
++++++++++++return+nl_langinfo(CODESET)|158047|我该如何避免呢？|158048|试试getpreferredencoding(False)|158049|为什么在windows中不会发生这种情况？|158050|Windows使用不同的代码来获取区域设置，正如您可以看到的here。|158051|在Python+3中|158052|在Python3中，可以看到here，getdefaultlocale不接受布尔型setlocale变量，也不调用setlocale本身。|158053|entityMap|0|LINK|mutability|MUTABLE|url|http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l612|1|http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l593|2|http://hg.python.org/cpython/file/d1bf37def4fd/Lib/locale.py#l506^0|0|2|I|G|11|4|0|0|0|0|0|0|0|0|M|L|0|0|7|M|V|7|1D|I|27|19|0|0|7|I|Z|9|1C|K|20|1T|0|0|0|0|2|R|0|0|U|4|1|0|0|J|G|E|4|2|0^^$0|@$1|2|3|4|5|6|7|1U|8|@$9|1V|A|1W|B|C]|$9|1X|A|1Y|B|D]|$9|1Z|A|20|B|D]]|E|@]|F|$]]|$1|G|3|-4|5|6|7|21|8|@]|E|@]|F|$]]|$1|H|3|I|5|6|7|22|8|@]|E|@]|F|$]]|$1|J|3|K|5|6|7|23|8|@]|E|@]|F|$]]|$1|L|3|-4|5|6|7|24|8|@]|E|@]|F|$]]|$1|M|3|N|5|O|7|25|8|@]|E|@]|F|$]]|$1|P|3|-4|5|6|7|26|8|@]|E|@]|F|$]]|$1|Q|3|-4|5|6|7|27|8|@]|E|@]|F|$]]|$1|R|3|S|5|6|7|28|8|@$9|29|A|2A|B|D]]|E|@]|F|$]]|$1|T|3|U|5|V|7|2B|8|@]|E|@]|F|$W|X]]|$1|Y|3|Z|5|6|7|2C|8|@$9|2D|A|2E|B|D]|$9|2F|A|2G|B|D]|$9|2H|A|2I|B|D]|$9|2J|A|2K|B|D]]|E|@]|F|$]]|$1|10|3|11|5|V|7|2L|8|@]|E|@]|F|$W|X]]|$1|12|3|13|5|6|7|2M|8|@$9|2N|A|2O|B|D]|$9|2P|A|2Q|B|D]|$9|2R|A|2S|B|D]]|E|@$9|2T|A|2U|1|2V]]|F|$]]|$1|14|3|15|5|V|7|2W|8|@]|E|@]|F|$W|X]]|$1|16|3|17|5|6|7|2X|8|@]|E|@]|F|$]]|$1|18|3|19|5|6|7|2Y|8|@$9|2Z|A|30|B|D]]|E|@]|F|$]]|$1|1A|3|1B|5|6|7|31|8|@]|E|@]|F|$]]|$1|1C|3|1D|5|6|7|32|8|@]|E|@$9|33|A|34|1|35]]|F|$]]|$1|1E|3|1F|5|6|7|36|8|@]|E|@]|F|$]]|$1|1G|3|1H|5|6|7|37|8|@$9|38|A|39|B|D]]|E|@$9|3A|A|3B|1|3C]]|F|$]]|$1|1I|3|-4|5|6|7|3D|8|@]|E|@]|F|$]]]|1J|$1K|$5|1L|1M|1N|F|$1O|1P]]|1Q|$5|1L|1M|1N|F|$1O|1R]]|1S|$5|1L|1M|1N|F|$1O|1T]]]]

<p><strong>Note</strong>: what OP did to solve the issue is to pass <code>encoding='UTF-8'</code> to the <code>open</code> call. If you run into this issue and are just looking for a fix this works. The rest of the post is an emphasis on <em>why</em>.</p>

<hr>

<h3>What happens</h3>

<p>As Lukas said, the docs specify:</p>

<blockquote>
  <p>On some systems, it is necessary to invoke setlocale() to obtain the user preferences</p>
</blockquote>

<p>Initially, string.letters is set to returning <code>lowercase + uppercase</code>:</p>

<pre><code>lowercase = 'abcdefghijklmnopqrstuvwxyz'
uppercase = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
letters = lowercase + uppercase
</code></pre>

<p>However, when you call <code>getpreferredencoding()</code>, the <code>_locale</code> module overrides it by calling <code>PyDict_SetItemString(string, "letters", ulo);</code> after it generates them inside <code>fixup_ulcase(void)</code> with the following:</p>

<pre><code>/* create letters string */
n = 0;
for (c = 0; c &lt; 256; c++) {
    if (isalpha(c))
        ul[n++] = c;
}
ulo = PyString_FromStringAndSize((const char *)ul, n);
if (!ulo)
    return;
if (string)
    PyDict_SetItemString(string, "letters", ulo);
Py_DECREF(ulo);
</code></pre>

<p>In turn, this is called in <code>PyLocale_setlocale</code> which is indeed <code>setlocale</code>, which is called by <code>getpreferredencoding</code> - code here <a href="http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l612" rel="noreferrer">http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l612</a> :</p>

<pre><code>  def getpreferredencoding(do_setlocale = True):
        """Return the charset that the user is likely using,
        according to the system configuration."""
        if do_setlocale:
            oldloc = setlocale(LC_CTYPE)
            try:
                setlocale(LC_CTYPE, "")
            except Error:
                pass
            result = nl_langinfo(CODESET)
            setlocale(LC_CTYPE, oldloc)
            return result
        else:
            return nl_langinfo(CODESET)
</code></pre>

<h3>How do I avoid it?</h3>

<p>Try <code>getpreferredencoding(False)</code></p>

<h3>Why does it not happen in windows?</h3>

<p>Windows uses different code for getting the locale, as you can see <a href="http://hg.python.org/cpython/file/07a6fca7ff42/Lib/locale.py#l593" rel="noreferrer">here</a>.</p>

<h3>In Python 3</h3>

<p>In Python 3, <code>getdefaultlocale</code> does not accept a boolean setlocale variable and does not call setlocale itself as you can see <a href="http://hg.python.org/cpython/file/d1bf37def4fd/Lib/locale.py#l506" rel="noreferrer">here</a>.</p>


<pre><code>&gt;&gt;&gt; import string
&gt;&gt;&gt; import locale
&gt;&gt;&gt; string.letters
'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
&gt;&gt;&gt; locale.getpreferredencoding()
'UTF-8'
&gt;&gt;&gt; string.letters
'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'
</code></pre>

<p>Any workarounds for this?</p>

<p>Platform: Linux
Python2.6.7 and Python2.7.3 seem to be affected, Works fine in Python3 (with <code>ascii_letters</code>)</p>


locale.getpreferredencoding() - why does this reset string.letters?

Windows

Python

Linux

>>> import string>>> import locale>>> string.letters'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'>>> locale.getpreferredencoding()'UTF-8'>>> string.let...

问locale.getpreferredencoding() -为什么要重置string.letters？
EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问locale.getpreferredencoding() -为什么要重置string.letters？EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问locale.getpreferredencoding() -为什么要重置string.letters？
EN