more imdb refinement

ox/web/imdb.py

@@ -7,12 +7,13 @@ import os
 import time
 
 import ox
-from ox import findRe
+from ox import findRe, stripTags
 from ox.normalize import normalizeTitle, normalizeImdbId
 
 from siteparser import SiteParser
 import google
 
+
 class Imdb(SiteParser):
     regex =  {
         'alternative_titles': {
@@ -25,13 +26,17 @@ class Imdb(SiteParser):
         
         },
         'cast': {
-            'page': 'combined',
+            'page': 'fullcredits',
-            're': '<td class="nm">.*?>(.*?)</a>.*?<td class="char">(.*?)</td>',
+            're': [
+                '<td class="nm">.*?>(.*?)</a>.*?<td class="char">(.*?)</td>',
+                lambda ll: [stripTags(l) for l in ll]
+             ],
             'type': 'list'
         },
         'cinematographers': {
-            'page': 'combined',
+            'page': 'fullcredits',
             're': [
+                lambda data: data.split('Series Crew')[0],
                 'Cinematography by</a>(.*?)</table>',
                 '<a href="/name/.*?/">(.*?)</a>'
             ],
@@ -48,16 +53,18 @@ class Imdb(SiteParser):
             'type': 'list'
         },
         'directors': {
-            'page': 'combined',
+            'page': 'fullcredits',
             're': [
+                lambda data: data.split('Series Crew')[0],
                 'Directed by</a>(.*?)</table>',
                 '<a href="/name/.*?/">(.*?)</a>'
             ],
             'type': 'list'
         },
         'editors': {
-            'page': 'combined',
+            'page': 'fullcredits',
             're': [
+                lambda data: data.split('Series Crew')[0],
                 'Film Editing by</a>(.*?)</table>',
                 '<a href="/name/.*?/">(.*?)</a>'
             ],
@@ -108,7 +115,7 @@ class Imdb(SiteParser):
         },
         'rating': {
             'page': 'combined',
-            're': '<div class="starbar-meta">.*?<b>(.*?)/10</b>',
+            're': '<div class="starbar-meta">.*?<b>([\d,.]?)/10</b>',
             'type': 'float'
         },
         'release_date': {
@@ -141,12 +148,13 @@ class Imdb(SiteParser):
         },
         'votes': {
             'page': 'combined',
-            're': '<a href="ratings" class="tn15more">(.*?) votes</a>',
+            're': '<a href="ratings" class="tn15more">([\d,]*?) votes</a>',
             'type': 'string'
         },
         'writers': {
-            'page': 'combined',
+            'page': 'fullcredits',
             're': [
+                lambda data: data.split('Series Crew')[0],
                 'Writing credits</a>(.*?)</table>',
                 '<a href="/name/.*?/">(.*?)</a>'
             ],
@@ -167,8 +175,8 @@ class Imdb(SiteParser):
             if 'min' in self['runtime']: base=60
             else: base=1
             self['runtime'] = int(findRe(self['runtime'], '([0-9]+)')) * base
-        else:
+        if 'runtime' in self and not self['runtime']:
-            self['runtime'] = 0
+            del self['runtime']
         if 'votes' in self: self['votes'] = self['votes'].replace(',', '')
         if 'connections' in self:
             cc={}
@@ -179,6 +187,7 @@ class Imdb(SiteParser):
             self['connections'] = cc
 
         for key in ('countries', 'genres'):
+            if key in self:
                 self[key] = filter(lambda x: x.lower() != 'home', self[key])
 
 
@@ -192,7 +201,7 @@ def guess(title, director='', timeout=google.DEFAULT_TIMEOUT):
     return_url = ''
 
     #lest first try google
-    #i.e. site:imdb.com Michael Stevens Sin
+    #i.e. site:imdb.com Michael Stevens "Sin"
     if director:
         search = 'site:imdb.com %s "%s"' % (director, title)
     else:

ox/web/siteparser.py

@@ -15,7 +15,7 @@ def cleanup(key, data, data_type):
             data = [decodeHtml(p).strip() for p in data]
         elif isinstance(data[0], list) or isinstance(data[0], tuple):
             data = [cleanup(key, p, data_type) for p in data]
-        while len(data) == 1:
+        while len(data) == 1 and not isinstance(data, basestring):
             data = data[0]
         if data_type == 'list' and isinstance(data, basestring):
             data = [data, ]
@@ -37,12 +37,18 @@ class SiteParser(dict):
             if isinstance(self.regex[key]['re'], basestring):
                 data = re.compile(self.regex[key]['re'], re.DOTALL).findall(data)
                 data = cleanup(key, data, self.regex[key]['type'])
+            elif callable(self.regex[key]['re']):
+                data = self.regex[key]['re'](data)
             else:
                 for r in self.regex[key]['re']:
-                    if isinstance(data, basestring):
+                    if callable(r):
-                        data = re.compile(r, re.DOTALL).findall(data)
+                        f = r
                     else:
-                        data = [re.compile(r, re.DOTALL).findall(d) for d in data]
+                        f = re.compile(r, re.DOTALL).findall
+                    if isinstance(data, basestring):
+                        data = f(data)
+                    else:
+                        data = [f(d) for d in data]
                         data = cleanup(key, data, self.regex[key]['type'])
             def apply_f(f, data):
                 if data and isinstance(data[0], list):
@@ -50,12 +56,13 @@ class SiteParser(dict):
                 else:
                     data = f(data)
                 return data
-            if self.regex[key]['type'] == 'float':
+            if self.regex[key]['type'] == 'float' and data:
                 data = apply_f(float, data)
             elif self.regex[key]['type'] == 'int':
                 data = apply_f(int, data)
             elif self.regex[key]['type'] == 'date':
                 parse_date = lambda d: d and datetime.strptime('-'.join(d), '%m-%d-%Y').strftime('%Y-%m-%d')
                 data = apply_f(parse_date, data)
+            if data:
                 self[key] = data