Merge pull request #664 from Gluejar/opds-json

add a json version of opds
pull/43/head
eshellman 2017-02-12 10:41:06 -05:00 committed by GitHub
commit 2ec1c8edd6
6 changed files with 383 additions and 16 deletions

287
api/opds_json.py Normal file
View File

@ -0,0 +1,287 @@
from itertools import islice
import datetime
import urlparse
from django.core.urlresolvers import reverse
from django.utils.http import urlquote
import json
import pytz
import logging
logger = logging.getLogger(__name__)
from regluit.core import models, facets
import regluit.core.cc as cc
from .opds import (
feeds,
get_facet_class,
add_query_component,
Facet,
get_facet_facet,
opds_feed_for_work,
)
licenses = cc.LICENSE_LIST
FORMAT_TO_MIMETYPE = {'pdf':"application/pdf",
'epub':"application/epub+zip",
'mobi':"application/x-mobipocket-ebook",
'html':"text/html",
'text':"text/html"}
UNGLUEIT_URL= 'https://unglue.it'
ACQUISITION = "application/vnd.opds.acquisition+json"
FACET_RELATION = "opds:facet"
def feeds():
for facet_path in facets.get_all_facets('Format'):
yield get_facet_facet(facet_path)
for facet_path in facets.get_all_facets('Keyword'):
yield get_facet_facet(facet_path)
def get_facet_class(name):
return get_facet_facet(name)
def text_node(tag, text):
return {tag:text}
def html_node(tag, html):
return {tag:html}
def isbn_node(isbn):
return 'urn:ISBN:'+ isbn
def work_node(work, facet=None):
content={}
# title
content["title"] = work.title
# id
content.update(text_node('id', "{base}{url}".format(base=UNGLUEIT_URL,url=reverse('work_identifier',kwargs={'work_id':work.id}))))
updated = None
# links for all ebooks
ebooks = facet.filter_model("Ebook",work.ebooks()) if facet else work.ebooks()
versions = set()
content['_links'] = links = {}
for ebook in ebooks:
if updated is None:
# most recent ebook, first ebook in loop
updated = ebook.created.isoformat()
content.update(text_node('updated', updated))
if not ebook.version_label in versions:
versions.add(ebook.version_label)
link_node_attrib = {}
ebookfiles = links.get("opds:acquisition:open-access",[])
ebookfiles.append(link_node_attrib)
# ebook.download_url is an absolute URL with the protocol, domain, and path baked in
link_node_attrib.update({"href":add_query_component(ebook.download_url, "feed=opds"),
"rights": str(ebook.rights)})
if ebook.is_direct():
link_node_attrib["type"] = FORMAT_TO_MIMETYPE.get(ebook.format, "")
else:
""" indirect acquisition, i.e. google books """
link_node_attrib["type"] = "text/html"
indirect_attrib = {}
indirect = {"indirectAcquisition":indirect_attrib}
indirect_attrib["type"] = FORMAT_TO_MIMETYPE.get(ebook.format, "")
link_node_attrib.update(indirect)
if ebook.version_label:
link_node_attrib.update({"version": ebook.version_label})
links["opds:acquisition:open-access"] = ebookfiles
# get the cover -- assume jpg?
if work.cover_image_small():
cover_node_attrib = {}
cover_node = {"opds:image:thumbnail": cover_node_attrib}
cover_node_attrib.update({"href":work.cover_image_small(),
"type":"image/"+work.cover_filetype(),
})
links.update(cover_node)
if work.cover_image_thumbnail():
cover_node2_attrib = {}
cover_node2 = {"opds:image": cover_node2_attrib}
cover_node2_attrib.update({"href":work.cover_image_thumbnail(),
"type":"image/"+work.cover_filetype(),
})
links.update(cover_node2)
# <dcterms:issued>2012</dcterms:issued>
content.update({"issued": work.publication_date})
# author
# TO DO: include all authors?
content["contributor"] = {"name": work.author()}
# publisher
#<dcterms:publisher>Open Book Publishers</dcterms:publisher>
if len(work.publishers()):
content["publishers"] = [{"publisher": publisher.name.name}
for publisher in work.publishers()]
# language
content["language"] = work.language
# description
content["summary"] = work.description
# identifiers
if work.identifiers.filter(type='isbn'):
content['identifers'] = [isbn_node(isbn.value)
for isbn in work.identifiers.filter(type='isbn')[0:9]] #10 should be more than enough
# subject tags
subjects = [subject.name for subject in work.subjects.all()]
if subjects:
content["category"] = subjects
# age level
# <category term="15-18" scheme="http://schema.org/typicalAgeRange" label="Teen - Grade 10-12, Age 15-18"/>
if work.age_level:
age_level_node_attrib = {}
age_level_node = {"category": age_level_node_attrib}
age_level_node_attrib["scheme"] = 'http://schema.org/typicalAgeRange'
age_level_node_attrib["term"] = work.age_level
age_level_node_attrib["label"] = work.get_age_level_display()
content.update(age_level_node)
# rating
content["Rating"] = {"ratingValue":"{:}".format(work.priority())}
return content
class Facet:
title = ''
works = None
feed_path = ''
description = ''
def feed(self, page=None, order_by='newest'):
self.works = self.works.order_by(*facets.get_order_by(order_by))
return opds_feed_for_works(self, page=page, order_by=order_by)
def updated(self):
# return the creation date for most recently added item
if not self.works:
return pytz.utc.localize(datetime.datetime.utcnow()).isoformat()
else:
return pytz.utc.localize(self.works[0].created).isoformat()
def get_facet_facet(facet_path):
class Facet_Facet(Facet):
def __init__(self, facet_path=facet_path):
self.feed_path = facet_path
self.facet_object = facets.get_facet_object(facet_path)
self.title = "Unglue.it"
for facet in self.facet_object.facets():
self.title = self.title + " " + facet.title
self.works = self.facet_object.get_query_set().distinct()
self.description = self.facet_object.description
return Facet_Facet
def opds_feed_for_work(work_id):
class single_work_facet:
def __init__(self, work_id):
try:
works=models.Work.objects.filter(id=work_id)
except models.Work.DoesNotExist:
works=models.Work.objects.none()
except ValueError:
# not a valid work_id
works=models.Work.objects.none()
self.works=works
self.title='Unglue.it work #%s' % work_id
self.feed_path=''
self.facet_object= facets.BaseFacet(None)
return opds_feed_for_works( single_work_facet(work_id) )
def opds_feed_for_works(the_facet, page=None, order_by='newest'):
books_per_page = 50
works = the_facet.works
feed_path = the_facet.feed_path
title = the_facet.title
feed = {'_type': ACQUISITION}
# add title
# TO DO: will need to calculate the number items and where in the feed we are
feed.update(text_node('title', title + ' - sorted by ' + order_by))
# id
feed.update(text_node('id', "{url}/api/opdsjson/{feed_path}/?order_by={order_by}".format(url=UNGLUEIT_URL,
feed_path=urlquote(feed_path), order_by=order_by)))
# updated
# TO DO: fix time zone?
# also use our wrapped datetime code
feed.update(text_node('updated',
pytz.utc.localize(datetime.datetime.utcnow()).isoformat()))
# author
author_node = {"author":{'name': 'unglue.it','uri': UNGLUEIT_URL}}
feed.update(author_node)
# links: start, self, next/prev (depending what's necessary -- to start with put all CC books)
feed['_links'] = {}
# start link
append_navlink(feed, 'start', feed_path, None , order_by, title="First {}".format(books_per_page))
# next link
if not page:
page =0
else:
try:
page=int(page)
except TypeError:
page=0
try:
works[books_per_page * page + books_per_page]
append_navlink(feed, 'next', feed_path, page+1 , order_by, title="Next {}".format(books_per_page))
except IndexError:
pass
# sort facets
append_navlink(feed, FACET_RELATION, feed_path, None, 'popular', group="Order", active = order_by=='popular', title="Sorted by popularity")
append_navlink(feed, FACET_RELATION, feed_path, None, 'newest', group="Order", active = order_by=='newest', title="Sorted by newest")
#other facets
for other_group in the_facet.facet_object.get_other_groups():
for facet_object in other_group.get_facets():
append_navlink(feed, FACET_RELATION, feed_path + '/' + facet_object.facet_name, None, order_by, group=other_group.title, title=facet_object.title)
works = islice(works, books_per_page * page, books_per_page * page + books_per_page)
if page > 0:
append_navlink(feed, 'previous', feed_path, page-1, order_by, title="Previous {}".format(books_per_page))
feedlist = []
feed['publications'] = feedlist
for work in works:
node = work_node(work, facet=the_facet.facet_object)
feedlist.append(node)
return json.dumps(feed,indent=2, separators=(',', ': '), sort_keys=False)
def append_navlink(feed, rel, path, page, order_by, group=None, active=None , title=""):
if page==None:
return
link = {
"href": UNGLUEIT_URL + "/api/opdsjson/" + urlquote(path) + '/?order_by=' + order_by + ('&page=' + unicode(page) ),
"type": ACQUISITION,
"title": title,
}
if rel == FACET_RELATION:
if group:
link['facetGroup'] = group
if active:
link['activeFacet'] = 'true'
feed['_links'][rel] = link

22
api/templates/opds.json Normal file
View File

@ -0,0 +1,22 @@
{
"_type": "application/vnd.opds.navigation+json",
"title": "Unglue.it Catalog",
"_links": {
"start": { "href": "https://unglue.it{% url 'opds' %}", "type": "application/vnd.opds.navigation+json" },
"opds:featured": {"href": "{{ feed.feed_path }}/?order_by=popular", "type": "application/vnd.opds.acquisition+json" }
},
"navigation": [{
"title": {{ feed.title }} - Popular,
"_links": {"opds:featured": { "href": "{{ feed.feed_path }}/?order_by=popular", "profile": "http://opds-spec.org/navigation", "type": "application/vnd.opds.acquisition+json"}}
},{
"title": "{{ feed.title }} - New",
"_links": {"opds:new": { "href": "{{ feed.feed_path }}/?order_by=newest", "profile": "http://opds-spec.org/navigation", "type": "application/vnd.opds.acquisition+json" }}
}
{% for feed in feeds %},{
"title": "{{ feed.title }}",
"_links": {"opds:new": { "href": "{{ feed.feed_path }}", "profile": "http://opds-spec.org/navigation", "type": "application/vnd.opds.acquisition+json" }}
}
{% endfor %}]
}

View File

@ -18,7 +18,7 @@
type="application/atom+xml;profile=opds-catalog;kind=acquisition" />
<entry>
<title>{{ feed.title }} - Popular</title>
<id>https://unglue.it{% url 'opds_acqusition' feed.feed_path %}?order_by=popular</id>
<id>https://unglue.it{% url 'opds_acqusition' feed.feed_path %}/?order_by=popular</id>
<updated>{{ feed.updated }}</updated>
<link href="{{ feed.feed_path }}/?order_by=popular"
rel="http://opds-spec.org/sort/popular"
@ -27,9 +27,9 @@
</entry>
<entry>
<title>{{ feed.title }} - New</title>
<id>https://unglue.it{% url 'opds_acqusition' feed.feed_path %}?order_by=newest</id>
<id>https://unglue.it{% url 'opds_acqusition' feed.feed_path %}/?order_by=newest</id>
<updated>{{ feed.updated }}</updated>
<link href="{{ feed.feed_path }}/?order_by=popular"
<link href="{{ feed.feed_path }}/?order_by=newest"
rel="http://opds-spec.org/sort/new"
type="application/atom+xml;profile=opds-catalog;kind=acquisition" />
<content type="text">{{ feed.description }} - ordered by newest</content>

View File

@ -32,7 +32,9 @@ urlpatterns = [
url(r'^featured_cover$', featured_cover, name="featured_cover"),
url(r'^featured_url$', featured_url, name="featured_url"),
url(r"^opds/$", OPDSNavigationView.as_view(template_name="opds.xml"), name="opds"),
url(r"^opdsjson/$", OPDSNavigationView.as_view(template_name="opds.json", json=True), name="opdsjson"),
url(r"^opds/(?P<facet>.*)/$", OPDSAcquisitionView.as_view(), name="opds_acqusition"),
url(r"^opdsjson/(?P<facet>.*)/$", OPDSAcquisitionView.as_view(json=True), name="opdsjson_acqusition"),
url(r"^onix/(?P<facet>.*)/$", OnixView.as_view(), name="onix"),
url(r"^onix/$", OnixView.as_view(), name="onix_all"),
url(r'^id/work/(?P<work_id>\w+)/$', negotiate_content, name="work_identifier"),

View File

@ -23,7 +23,7 @@ from django.http import (
import regluit.core.isbn
from regluit.core.bookloader import load_from_yaml
from regluit.api import opds, onix
from regluit.api import opds, onix, opds_json
from regluit.api.models import repo_allowed
from regluit.core import models
@ -166,26 +166,37 @@ class ApiHelpView(TemplateView):
return context
class OPDSNavigationView(TemplateView):
json=False
# http://stackoverflow.com/a/6867976: secret to how to change content-type
def render_to_response(self, context, **response_kwargs):
response_kwargs['content_type'] = "application/atom+xml;profile=opds-catalog;kind=navigation"
if json:
response_kwargs['content_type'] = "application/vnd.opds.navigation+json"
else:
response_kwargs['content_type'] = "application/atom+xml;profile=opds-catalog;kind=navigation"
return super(TemplateView, self).render_to_response(context, **response_kwargs)
def get_context_data(self, **kwargs):
context = super(OPDSNavigationView, self).get_context_data(**kwargs)
context["feeds"] = opds.feeds()
context["feed"] = opds.get_facet_facet('all')
if json:
context["feeds"] = opds_json.feeds()
context["feed"] = opds_json.get_facet_facet('all')
else:
context["feeds"] = opds.feeds()
context["feed"] = opds.get_facet_facet('all')
return context
class OPDSAcquisitionView(View):
json = False
def get(self, request, *args, **kwargs):
work = request.GET.get('work', None)
if work:
return HttpResponse(opds.opds_feed_for_work(work),
content_type="application/atom+xml;profile=opds-catalog;kind=acquisition")
if self.json:
return HttpResponse(opds_json.opds_feed_for_work(work),
content_type="application/json;profile=opds-catalog;kind=acquisition")
else:
return HttpResponse(opds.opds_feed_for_work(work),
content_type="application/atom+xml;profile=opds-catalog;kind=acquisition")
facet = kwargs.get('facet')
page = request.GET.get('page', None)
order_by = request.GET.get('order_by', 'newest')
@ -193,9 +204,14 @@ class OPDSAcquisitionView(View):
page = int(page)
except:
page = None
facet_class = opds.get_facet_class(facet)()
return HttpResponse(facet_class.feed(page,order_by),
content_type="application/atom+xml;profile=opds-catalog;kind=acquisition")
if self.json:
facet_class = opds_json.get_facet_class(facet)()
return HttpResponse(facet_class.feed(page,order_by),
content_type="application/vnd.opds.acquisition+json")
else:
facet_class = opds.get_facet_class(facet)()
return HttpResponse(facet_class.feed(page,order_by),
content_type="application/atom+xml;profile=opds-catalog;kind=acquisition")
class OnixView(View):

View File

@ -1,6 +1,6 @@
from django.apps import apps
from django.db.models import Q
from regluit.core import cc
class BaseFacet(object):
facet_name = 'all'
@ -236,7 +236,47 @@ class KeywordFacetGroup(FacetGroup):
def description(self):
return "%s eBooks" % self.keyword
return KeywordFacet
class SearchFacetGroup(FacetGroup):
def __init__(self):
super(FacetGroup,self).__init__()
self.title = 'Search Term'
# make facets in TOPKW available for display
self.facets = []
self.label = '{} is ...'.format(self.title)
def has_facet(self, facet_name):
# recognize any facet_name that starts with "s." as a valid facet name
return facet_name.startswith('s.')
def get_facet_class(self, facet_name):
class KeywordFacet(NamedFacet):
def set_name(self):
self.facet_name=facet_name
# facet_names of the form 's.SUBJECT' and SUBJECT is therefore the 3rd character on
self.term=self.facet_name[2:]
def get_query_set(self):
return self._get_query_set().filter(
Q(title__icontains=self.term) |
Q(editions__authors__name__icontains=self.term) |
Q(subjects__name__iexact=self.term)
)
def template(self):
return 'facets/search.html'
@property
def title(self):
return self.term
@property
def label(self):
return self.term
@property
def description(self):
return "eBooks for {}".format(self.term)
return KeywordFacet
class PublisherFacetGroup(FacetGroup):
def __init__(self):
@ -285,7 +325,7 @@ class PublisherFacetGroup(FacetGroup):
return PublisherFacet
# order of groups in facet_groups determines order of display on /free/
facet_groups = [KeywordFacetGroup(), FormatFacetGroup(), LicenseFacetGroup(), PublisherFacetGroup(), IdFacetGroup()]
facet_groups = [KeywordFacetGroup(), FormatFacetGroup(), LicenseFacetGroup(), PublisherFacetGroup(), IdFacetGroup(), SearchFacetGroup()]
def get_facet(facet_name):
for facet_group in facet_groups: