Use html5-parser for all metadata download sources

src/calibre/ebooks/metadata/sources/amazon.py

@@ -30,10 +30,20 @@ class SearchFailed(ValueError):
 ua_index = -1
 
 
+def parse_html(raw):
+    try:
+        from html5_parser import parse
+    except ImportError:
+        # Old versions of calibre
+        import html5lib
+        return html5lib.parse(raw, treebuilder='lxml', namespaceHTMLElements=False)
+    else:
+        return parse(raw)
+
+
 def parse_details_page(url, log, timeout, browser, domain):
     from calibre.utils.cleantext import clean_ascii_chars
     from calibre.ebooks.chardet import xml_to_unicode
-    import html5lib
     from lxml.html import tostring
     log('Getting details from:', url)
     try:
@@ -65,9 +75,8 @@ def parse_details_page(url, log, timeout, browser, domain):
         raise ValueError('No cached entry for %s found' % url)
 
     try:
-        root = html5lib.parse(clean_ascii_chars(raw), treebuilder='lxml',
+        root = parse_html(clean_ascii_chars(raw))
-                              namespaceHTMLElements=False)
+    except Exception:
-    except:
         msg = 'Failed to parse amazon details page: %r' % url
         log.exception(msg)
         return
@@ -589,8 +598,7 @@ class Worker(Thread):  # Get details {{{
             if m is not None:
                 try:
                     text = unquote(m.group(1)).decode('utf-8')
-                    nr = html5lib.parse(
+                    nr = parse_html(text)
-                        text, treebuilder='lxml', namespaceHTMLElements=False)
                     desc = nr.xpath(
                         '//div[@id="productDescription"]/*[@class="content"]')
                     if desc:
@@ -1201,7 +1209,6 @@ class Amazon(Source):
     # }}}
 
     def search_amazon(self, br, testing, log, abort, title, authors, identifiers, timeout):  # {{{
-        import html5lib
         from calibre.utils.cleantext import clean_ascii_chars
         from calibre.ebooks.chardet import xml_to_unicode
         matches = []
@@ -1242,8 +1249,7 @@ class Amazon(Source):
 
         if found:
             try:
-                root = html5lib.parse(raw, treebuilder='lxml',
+                root = parse_html(raw)
-                                      namespaceHTMLElements=False)
             except Exception:
                 msg = 'Failed to parse amazon page for query: %r' % query
                 log.exception(msg)

src/calibre/ebooks/metadata/sources/edelweiss.py

@@ -16,14 +16,23 @@ from calibre.ebooks.metadata import check_isbn
 from calibre.ebooks.metadata.sources.base import Source
 
 
-def parse_html(raw):
+def clean_html(raw):
-    import html5lib
     from calibre.ebooks.chardet import xml_to_unicode
     from calibre.utils.cleantext import clean_ascii_chars
-    raw = clean_ascii_chars(xml_to_unicode(raw, strip_encoding_pats=True,
+    return clean_ascii_chars(xml_to_unicode(raw, strip_encoding_pats=True,
                                 resolve_entities=True, assume_utf8=True)[0])
-    return html5lib.parse(raw, treebuilder='lxml',
+
-                              namespaceHTMLElements=False).getroot()
+
+def parse_html(raw):
+    raw = clean_html(raw)
+    try:
+        from html5_parser import parse
+    except ImportError:
+        # Old versions of calibre
+        import html5lib
+        return html5lib.parse(raw, treebuilder='lxml', namespaceHTMLElements=False)
+    else:
+        return parse(raw)
 
 
 def astext(node):

src/calibre/ebooks/metadata/sources/google_images.py

@@ -13,6 +13,17 @@ from calibre import random_user_agent
 from calibre.ebooks.metadata.sources.base import Source, Option
 
 
+def parse_html(raw):
+    try:
+        from html5_parser import parse
+    except ImportError:
+        # Old versions of calibre
+        import html5lib
+        return html5lib.parse(raw, treebuilder='lxml', namespaceHTMLElements=False)
+    else:
+        return parse(raw)
+
+
 class GoogleImages(Source):
 
     name = 'Google Images'
@@ -55,7 +66,6 @@ class GoogleImages(Source):
     def get_image_urls(self, title, author, log, abort, timeout):
         from calibre.utils.cleantext import clean_ascii_chars
         from urllib import urlencode
-        import html5lib
         import json
         from collections import OrderedDict
         ans = OrderedDict()
@@ -72,8 +82,8 @@ class GoogleImages(Source):
         # URL scheme
         url = 'https://www.google.com/search?as_st=y&tbm=isch&{}&as_epq=&as_oq=&as_eq=&cr=&as_sitesearch=&safe=images&tbs={}iar:t,ift:jpg'.format(q, sz)
         log('Search URL: ' + url)
-        raw = br.open(url).read().decode('utf-8')
+        raw = clean_ascii_chars(br.open(url).read().decode('utf-8'))
-        root = html5lib.parse(clean_ascii_chars(raw), treebuilder='lxml', namespaceHTMLElements=False)
+        root = parse_html(raw)
         for div in root.xpath('//div[@class="rg_meta notranslate"]'):
             try:
                 data = json.loads(div.text)

src/calibre/ebooks/metadata/sources/search_engines.py

@@ -14,7 +14,6 @@ from urlparse import parse_qs
 
 from lxml import etree
 
-import html5lib
 from calibre import browser as _browser, prints, random_user_agent
 from calibre.utils.monotonic import monotonic
 from calibre.utils.random_ua import accept_header_for_ua
@@ -48,7 +47,14 @@ def encode_query(**query):
 
 
 def parse_html(raw):
+    try:
+        from html5_parser import parse
+    except ImportError:
+        # Old versions of calibre
+        import html5lib
         return html5lib.parse(raw, treebuilder='lxml', namespaceHTMLElements=False)
+    else:
+        return parse(raw)
 
 
 def query(br, url, key, dump_raw=None, limit=1, parser=parse_html, timeout=60):