尝试用python自动翻译babelfish

Question

我修改了一个python bablizer来帮助我把英语翻译成中文。在

## {{{ http://code.activestate.com/recipes/64937/ (r4)
# babelizer.py - API for simple access to babelfish.altavista.com.
#                Requires python 2.0 or better.
#
# See it in use at http://babel.MrFeinberg.com/

"""API for simple access to babelfish.altavista.com.

Summary:

    import babelizer

    print ' '.join(babelizer.available_languages)

    print babelizer.translate( 'How much is that doggie in the window?',
                               'English', 'French' )

    def babel_callback(phrase):
        print phrase
        sys.stdout.flush()

    babelizer.babelize( 'I love a reigning knight.',
                        'English', 'German',
                        callback = babel_callback )

available_languages
    A list of languages available for use with babelfish.

translate( phrase, from_lang, to_lang )
    Uses babelfish to translate phrase from from_lang to to_lang.

babelize(phrase, from_lang, through_lang, limit = 12, callback = None)
    Uses babelfish to translate back and forth between from_lang and
    through_lang until either no more changes occur in translation or
    limit iterations have been reached, whichever comes first.  Takes
    an optional callback function which should receive a single
    parameter, being the next translation.  Without the callback
    returns a list of successive translations.

It's only guaranteed to work if 'english' is one of the two languages
given to either of the translation methods.

Both translation methods throw exceptions which are all subclasses of
BabelizerError.  They include

LanguageNotAvailableError
    Thrown on an attempt to use an unknown language.

BabelfishChangedError
    Thrown when babelfish.altavista.com changes some detail of their
    layout, and babelizer can no longer parse the results or submit
    the correct form (a not infrequent occurance).

BabelizerIOError
    Thrown for various networking and IO errors.

Version: $Id: babelizer.py,v 1.4 2001/06/04 21:25:09 Administrator Exp $
Author: Jonathan Feinberg <jdf@pobox.com>
"""
import re, string, urllib
import httplib, urllib
import sys

"""
Various patterns I have encountered in looking for the babelfish result.
We try each of them in turn, based on the relative number of times I've
seen each of these patterns.  $1.00 to anyone who can provide a heuristic
for knowing which one to use.   This includes AltaVista employees.
"""
__where = [ re.compile(r'name=\"q\">([^<]*)'),
            re.compile(r'td bgcolor=white>([^<]*)'),
            re.compile(r'<\/strong><br>([^<]*)')
          ]
# <div id="result"><div style="padding:0.6em;">??</div></div>
__where = [ re.compile(r'<div id=\"result\"><div style=\"padding\:0\.6em\;\">(.*)<\/div><\/div>', re.U) ]

__languages = { 'english'   : 'en',
                'french'    : 'fr',
                'spanish'   : 'es',
                'german'    : 'de',
                'italian'   : 'it',
                'portugese' : 'pt',
                'chinese' : 'zh'
              }

"""
  All of the available language names.
"""
available_languages = [ x.title() for x in __languages.keys() ]

"""
  Calling translate() or babelize() can raise a BabelizerError
"""
class BabelizerError(Exception):
    pass

class LanguageNotAvailableError(BabelizerError):
    pass
class BabelfishChangedError(BabelizerError):
    pass
class BabelizerIOError(BabelizerError):
    pass

def saveHTML(txt):
    f = open('page.html', 'wb')
    f.write(txt)
    f.close()

def clean(text):
    return ' '.join(string.replace(text.strip(), "\n", ' ').split())

def translate(phrase, from_lang, to_lang):
    phrase = clean(phrase)
    try:
        from_code = __languages[from_lang.lower()]
        to_code = __languages[to_lang.lower()]
    except KeyError, lang:
        raise LanguageNotAvailableError(lang)
    html = ""
    try:
        params = urllib.urlencode({'ei':'UTF-8', 'doit':'done', 'fr':'bf-res', 'intl':'1' , 'tt':'urltext', 'trtext':phrase, 'lp' : from_code + '_' + to_code , 'btnTrTxt':'Translate'})
        headers = {"Content-type": "application/x-www-form-urlencoded","Accept": "text/plain"}
        conn = httplib.HTTPConnection("babelfish.yahoo.com")
        conn.request("POST", "http://babelfish.yahoo.com/translate_txt", params, headers)
        response = conn.getresponse()
        html = response.read()
        saveHTML(html)
        conn.close()
        #response = urllib.urlopen('http://babelfish.yahoo.com/translate_txt', params)
    except IOError, what:
        raise BabelizerIOError("Couldn't talk to server: %s" % what)
    #print html
    for regex in __where:
        match = regex.search(html)
        if match: 
            break
    if not match: 
        raise BabelfishChangedError("Can't recognize translated string.")
    return match.group(1)
    #return clean(match.group(1))

def babelize(phrase, from_language, through_language, limit = 12, callback = None):
    phrase = clean(phrase)
    seen = { phrase: 1 }
    if callback:
        callback(phrase)
    else:
        results = [ phrase ]
    flip = { from_language: through_language, through_language: from_language }
    next = from_language
    for i in range(limit):
        phrase = translate(phrase, next, flip[next])
        if seen.has_key(phrase): break
        seen[phrase] = 1
        if callback:
            callback(phrase)
        else:
            results.append(phrase)
        next = flip[next]
    if not callback: return results

if __name__ == '__main__':
    import sys
    def printer(x):
        print x
        sys.stdout.flush();


    babelize("I won't take that sort of treatment from you, or from your doggie!",
             'english', 'french', callback = printer)
## end of http://code.activestate.com/recipes/64937/ }}}

测试代码是

^{pr2}$

结果在div块中是预期的。我修改了脚本以保存html响应。我发现所有的utf8字符都变成了nul。治疗utf8反应需要特别小心吗？在

Answer 1

我认为你需要使用

import codecs
codecs.open

在你的

^{pr2}$

方法，以处理utf-8文档。有关完整的解释，请参见Python Unicode Howto。在

希望这有帮助。在