Improve bin/wikitrans

* bin/wikitrans (getwiki): Take options as second argument and
modify it directly. Deduce options.lang, options.itype, and
options.kwdict['html_base'] from the URL, when possible.
* README.rst: Update description of bin/wikitrans

2 files changed, 15 insertions, 13 deletions

diff --git a/README.rst b/README.rst
index b6863bc..d189c98 100644
--- a/README.rst
+++ b/README.rst
@@ -167,8 +167,10 @@ Options are:
   Set debug level (0..100)
 ``-D``, ``--dump``
   Dump parse tree and exit; same as ``--type=dump``.
 ``-b URL``, ``--base-url=URL``
   Set base url. 
 
-
+Note: when using ``--base-url`` or passing URL as an argument (2nd and 3rd
+use cases above), if the URL is in 'wikipedia.org' or 'wiktionary.org'
+domain, the options ``--input-type``, and ``--lang`` are set automatically.
 
diff --git a/bin/wikitrans b/bin/wikitrans
index e9ab81f..caaa885 100755
--- a/bin/wikitrans
+++ b/bin/wikitrans
@@ -67,13 +67,13 @@ def setkw(option, opt, value, parser):
     if not parser.values.kwdict:
         parser.values.kwdict = {}
     (kw,sep,val) = value.partition('=')
     if val:
         parser.values.kwdict[kw] = val
 
-def getwiki(url):
+def getwiki(url, options):
     tmp = tempfile.NamedTemporaryFile()            
     if sys.version_info[0] > 2:
         import urllib.request
         with urllib.request.urlopen(url) as u:
             root = etree.fromstring(u.read())
     else:
@@ -84,13 +84,21 @@ def getwiki(url):
     if 'version' in root.attrib:
         ns['wiki'] = 'http://www.mediawiki.org/xml/export-%s/' % root.attrib['version']
         text = root.find('wiki:page/wiki:revision/wiki:text',ns)
     if text is None:
         print("no page/revision/text element in the downloaded page")
         exit(0)
-    return text.text.encode()
+
+    m = re.match('(?P<url>(?:.+://)(?P<lang>.+?)\.(?P<root>wik(?:ipedia|tionary))\.org)', url)
+    if m:
+        options.lang = m.group('lang')
+        options.kwdict['html_base'] = m.group('url') + '/wiki/'
+        if m.group('root') == 'wiktionary':
+            options.itype = 'wiktionary'
+        
+    options.kwdict['text'] = text.text.encode()
         
 def main():
     usage = '%prog [OPTIONS] ARG'
     version = '%prog 1.0'
     description = """Translates MediaWiki documents markup to various other formats.
 If ARG looks like a URL, the wiki text to be converted will be downloaded
@@ -138,25 +146,17 @@ Otherwise, ARG is name of the file to read wiki material from.
                       help='set base url')
     
 
     (options, args) = parser.parse_args()
     if len(args) == 1:
         if options.base_url:
-            options.kwdict['text'] = getwiki(options.base_url
-                                             + '/wiki/Special:Export/' + args[0])
-            m = re.match('(?:.+://)(.+?)\.(wik(?:ipedia|tionary))\.org',
-                         options.base_url)
-            if m:
-                options.lang = m.group(1)
-                options.kwdict['html_base'] = options.base_url + '/wiki/'
-                if m.group(2) == 'wiktionary':
-                    options.itype = 'wiktionary'
+            getwiki(options.base_url + '/wiki/Special:Export/' + args[0], options)
         elif args[0] == '-':
             options.kwdict['file'] = sys.stdin
         elif re.match('^(http|ftp)s?://',args[0]):
-            options.kwdict['text'] = getwiki(args[0])
+            getwiki(args[0], options)
         else:
             options.kwdict['filename'] = args[0]
     else:
         parser.error("bad number of arguments")
         
     options.kwdict['lang'] = options.lang # FIXME