Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Amazon plugin: Handle the case of authors unknown to amazon and dont set the cover URL to the no cover available image

Browse Source
This commit is contained in:
Kovid Goyal 2011-03-16 23:44:48 -06:00
parent 2aee9ec321
commit 49e4f2cf28
2 changed files with 42 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
src/calibre/ebooks/metadata/sources/amazon.py
View File

@ -62,6 +62,7 @@ class Worker(Thread): # {{{
raw = xml_to_unicode(raw, strip_encoding_pats=True, raw = xml_to_unicode(raw, strip_encoding_pats=True,
resolve_entities=True)[0] resolve_entities=True)[0]
# open('/t/t.html', 'wb').write(raw)
if '<title>404 - ' in raw: if '<title>404 - ' in raw:
self.log.error('URL malformed: %r'%self.url) self.log.error('URL malformed: %r'%self.url)
@ -127,6 +128,7 @@ class Worker(Thread): # {{{
self.cover_url = self.parse_cover(root) self.cover_url = self.parse_cover(root)
except: except:
self.log.exception('Error parsing cover for url: %r'%self.url) self.log.exception('Error parsing cover for url: %r'%self.url)
mi.has_cover = bool(self.cover_url)
pd = root.xpath('//h2[text()="Product Details"]/../div[@class="content"]') pd = root.xpath('//h2[text()="Product Details"]/../div[@class="content"]')
if pd: if pd:
@ -177,7 +179,10 @@ class Worker(Thread): # {{{
return re.sub(r'[(\[].*[)\]]', '', title).strip() return re.sub(r'[(\[].*[)\]]', '', title).strip()
def parse_authors(self, root): def parse_authors(self, root):
aname = root.xpath('//span[@class="contributorNameTrigger"]') x = '//h1[@class="parseasinTitle"]/following-sibling::span/*[(name()="a" and @href) or (name()="span" and @class="contributorNameTrigger")]'
aname = root.xpath(x)
for x in aname:
x.tail = ''
authors = [tostring(x, encoding=unicode, method='text').strip() for x authors = [tostring(x, encoding=unicode, method='text').strip() for x
in aname] in aname]
return authors return authors
@ -215,6 +220,7 @@ class Worker(Thread): # {{{
imgs = root.xpath('//img[@id="prodImage" and @src]') imgs = root.xpath('//img[@id="prodImage" and @src]')
if imgs: if imgs:
src = imgs[0].get('src') src = imgs[0].get('src')
if '/no-image-avail' not in src:
parts = src.split('/') parts = src.split('/')
if len(parts) > 3: if len(parts) > 3:
bn = parts[-1] bn = parts[-1]
@ -424,30 +430,44 @@ if __name__ == '__main__':
# To run these test use: calibre-debug -e # To run these test use: calibre-debug -e
# src/calibre/ebooks/metadata/sources/amazon.py # src/calibre/ebooks/metadata/sources/amazon.py
from calibre.ebooks.metadata.sources.test import (test_identify_plugin, from calibre.ebooks.metadata.sources.test import (test_identify_plugin,
title_test) title_test, authors_test)
test_identify_plugin(Amazon.name, test_identify_plugin(Amazon.name,
[ [
( # An e-book ISBN not on Amazon, one of the authors is
# unknown to Amazon, so no popup wrapper
{'identifiers':{'isbn': '0307459671'},
'title':'Invisible Gorilla', 'authors':['Christopher Chabris']},
[title_test('The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us',
exact=True), authors_test(['Christopher Chabris', 'Daniel Simons'])]
),
( # This isbn not on amazon ( # This isbn not on amazon
{'identifiers':{'isbn': '8324616489'}, 'title':'Learning Python', {'identifiers':{'isbn': '8324616489'}, 'title':'Learning Python',
'authors':['Lutz']}, 'authors':['Lutz']},
[title_test('Learning Python: Powerful Object-Oriented Programming', exact=True)] [title_test('Learning Python: Powerful Object-Oriented Programming',
exact=True), authors_test(['Mark Lutz'])
]
), ),
( # Sophisticated comment formatting ( # Sophisticated comment formatting
{'identifiers':{'isbn': '9781416580829'}}, {'identifiers':{'isbn': '9781416580829'}},
[title_test('Angels & Demons - Movie Tie-In: A Novel', exact=True)] [title_test('Angels & Demons - Movie Tie-In: A Novel',
exact=True), authors_test(['Dan Brown'])]
), ),
( # No specific problems ( # No specific problems
{'identifiers':{'isbn': '0743273567'}}, {'identifiers':{'isbn': '0743273567'}},
[title_test('The great gatsby', exact=True)] [title_test('The great gatsby', exact=True),
authors_test(['F. Scott Fitzgerald'])]
), ),
( # A newer book ( # A newer book
{'identifiers':{'isbn': '9780316044981'}}, {'identifiers':{'isbn': '9780316044981'}},
[title_test('The Heroes', exact=True)] [title_test('The Heroes', exact=True),
authors_test(['Joe Abercrombie'])]
), ),

9
src/calibre/ebooks/metadata/sources/test.py
View File

@ -37,6 +37,15 @@ def title_test(title, exact=False):
return test return test
def authors_test(authors):
authors = set([x.lower() for x in authors])
def test(mi):
au = set([x.lower() for x in mi.authors])
return au == authors
return test
def test_identify_plugin(name, tests): def test_identify_plugin(name, tests):
''' '''
:param name: Plugin name :param name: Plugin name