Improve field boundary detection

[software/subdivxget.git] / subdivxget

diff --git a/subdivxget b/subdivxget
index 5c5cc7a49cfab50003860880a380ff6737f67fdd..4ebcca5a71abf2bd94aa04b9c93b1003af36f240 100755 (executable)
--- a/subdivxget
+++ b/subdivxget
@@ -7,12 +7,13 @@ import subprocess
 import HTMLParser
 
 class SubDivXQuery:
-       def __init__(self, to_search):
+       def __init__(self, to_search, page_number):
                self.host = "www.subdivx.com"
                self.page = "/index.php"
                self.down_page = "/bajar.php"
                self.query = dict(
                        buscar = to_search,
+                       pg = page_number,
                        accion = 5,
                        masdesc = '',
                        subtitulos = 1,
@@ -29,8 +30,6 @@ class SubDivXQuery:
        def down_uri(self):
                return 'http://' + self.host + self.down_page
 
-class Subtitle:
-       pass
 
 class SubDivXHTMLParser(HTMLParser.HTMLParser):
 
@@ -44,70 +43,110 @@ class SubDivXHTMLParser(HTMLParser.HTMLParser):
                self.parsing = False
                self.subs = []
                self.attr = None
+               self.attr_depth = 0
+               self.cur = None
+               self.in_script_style = False
 
        def handle_starttag(self, tag, attrs):
                attrs = dict(attrs)
                if tag == 'div' and attrs.get('id') == 'menu_detalle_buscador':
-                       #self.cur = Subtitle()
                        self.cur = dict()
                        self.subs.append(self.cur)
                        self.parsing = True
                if not self.parsing:
                        return
+               if tag == 'script' or tag == 'style':
+                       self.in_script_style = True
+                       return
                if tag == 'div':
                        if attrs.get('id') == 'buscador_detalle':
                                self.parsing = True
                        elif attrs.get('id') == 'buscador_detalle_sub':
                                self.attr = 'desc'
+                               self.attr_depth = self.depth + 1
+                               self.cur[self.attr] = ''
                elif tag == 'a':
                        if attrs.get('class') == 'titulo_menu_izq':
-                               self.attr = 'title'
+                               self.attr = 'titulo'
+                               self.attr_depth = self.depth + 1
+                               self.cur[self.attr] = ''
                        elif attrs.get('href', '').startswith(self.down_uri):
                                self.cur['url'] = attrs['href']
-               if self.parsing:
+               # br are usually not closed, so ignore them in depth calculation
+               if self.parsing and tag != 'br':
                        self.depth += 1
 
        def handle_endtag(self, tag):
                if self.parsing:
-                       self.depth -= 1
+                       if tag == 'script' or tag == 'style':
+                               self.in_script_style = False
+                               return
+                       if self.depth == self.attr_depth:
+                               self.attr = None
+                               self.attr_depth = 0
+                       # see comment in handle_starttag()
+                       if tag != 'br':
+                               self.depth -= 1
                if self.depth == 0:
                        self.parsing = False
 
        def handle_data(self, data):
-               if self.parsing:
-                       data = data.strip()
-                       if self.attr is not None and data:
-                               self.cur[self.attr] = data
+               if not self.parsing:
+                       return
+               data = data.strip()
+               # Hack to handle comments in <script> <style> which don't end
+               # up in handle_comment(), so we just ignore the whole tags
+               if self.in_script_style:
+                       return
+               if self.attr is not None and data:
+                       self.cur[self.attr] += ' ' + data
+                       if self.attr_depth == 0:
+                               self.cur[self.attr] = self.cur[self.attr].strip()
                                self.attr = None
-                               #self.cur[self.attr] = self.cur.get(self.attr, '') + data.strip()
-                               #setattr(self.cur, self.attr, data.strip())
-                       elif data in ('Downloads:', 'Cds:', 'Comentarios:',
-                                       'Formato:'):
-                               self.attr = data[:-1].lower()
-                       elif data == 'Subido por:':
-                               self.attr = 'autor'
-                       elif data == 'el':
-                               self.attr = 'fecha'
+                               self.attr_depth = 0
+               elif data in ('Downloads:', 'Cds:', 'Comentarios:', 'Formato:'):
+                       self.attr = data[:-1].lower()
+                       self.attr_depth = 0
+                       self.cur[self.attr] = ''
+               elif data == 'Subido por:':
+                       self.attr = 'autor'
+                       self.attr_depth = 0
+                       self.cur[self.attr] = ''
+               elif data == 'el':
+                       self.attr = 'fecha'
+                       self.attr_depth = 0
+                       self.cur[self.attr] = ''
 
 
-def get_subs(query_str):
-       query = SubDivXQuery(query_str)
+def subdivx_get_subs(query_str):
+       page_number = 1
+       subs = []
+       while True:
+               query = SubDivXQuery(query_str, page_number)
+               url = urllib.urlopen(query.url)
+               parser = SubDivXHTMLParser(query.down_uri)
+
+               for line in url:
+                       parser.feed(line)
 
-       url = urllib.urlopen(query.url)
+               url.close()
 
-       parser = SubDivXHTMLParser(query.down_uri)
+               if not parser.subs:
+                       break
 
-       for line in url:
-               parser.feed(line)
+               subs.extend(parser.subs)
+               page_number += 1
 
-       url.close()
+       return sorted(subs, key=lambda s: int(s['downloads']), reverse=True)
 
+
+def get_subs(query_str):
        zip_exts = ('application/zip',)
        rar_exts = ('application/rar', 'application/x-rar-compressed')
 
-       for sub in sorted(parser.subs, key=lambda s: int(s['downloads']), reverse=True):
+       for sub in subdivx_get_subs(query_str):
                print '''\
-       - %(title)s (%(autor)s - %(fecha)s - %(downloads)s - %(comentarios)s)
+       - %(titulo)s (%(autor)s - %(fecha)s - %(downloads)s - %(comentarios)s)
          %(desc)s
                DOWNLOADING ...
        ''' % sub