def nypl_identify_object(body, ctype, list_sets=None):
try:
data = json.loads(body)
except:
response.code = 500
response.add_header('content-type', 'text/plain')
return "Unable to parse body as JSON"
H = httplib2.Http('/tmp/.cache')
H.force_exception_as_status_code = True
resp, content = H.request(list_sets)
if not resp[u'status'].startswith('2'):
logger.error(' HTTP error (' + resp[u'status'] + ') resolving URL: ' + list_sets)
return body
content_dict = xmltodict.parse(content, xml_attribs=True, attr_prefix='', force_cdata=False, ignore_whitespace_cdata=True)
sets = content_dict["nyplAPI"]["response"]
for r in sets:
if "collection" == r:
for coll_dict in sets[r]:
if "uuid" in coll_dict and "title" in coll_dict and (coll_dict["uuid"] == data["title"] or coll_dict["uuid"] in data["@id"]):
data["title"] = coll_dict["title"]
return json.dumps(data)
def mix_freemix(body, ctype):
#See: http://foundry.zepheira.com/issues/137#note-10
'''
{
"datasets": {
"dataset1": "http://recollection.zepheira.com/data/guide/data-profile-arthur-y-ford-photograph-albums-reprise-for-demo/data.json",
"dataset2": "http://recollection.zepheira.com/data/guide/jean-thomas-collection/data.json"
},
"alignProperties": {
"label": "Name",
"dataset1": "Surname",
"dataset2": "Name_of_Candidate"
}
}
'''
USER, PASSWD = "loc", "recollection"
cache_dir = make_named_cache('mix.freemix.json')
H = httplib2.Http(cache_dir)
if USER:
H.add_credentials(USER, PASSWD)
request = json.loads(body)
datasets = request['datasets']
alignments = request.get('alignProperties')
if len(datasets) != 2:
raise ValueError('You must provide Mixer exactly 2 data sets')
if alignments:
mixed = []
for dataset in datasets:
#Replace the data set URL with the content
logger.debug("Processing dataset: %s" % (datasets[dataset]))
resp, content = H.request(datasets[dataset])
items = json.loads(content)[u'items']
prop = alignments[dataset]
newprop_label = alignments['label']
#Potluck (the usual mixer client) seems to generate property names such as
#"Activity / Activity" which Exhibit cannot handle. Work around that.
#See: http://foundry.zepheira.com/issues/334
newprop = UNSUPPORTED_IN_EXHIBITKEY.sub('_', newprop_label)
logger.debug("Mapping: %s -> %s" % (prop, newprop))
for item in items:
if prop in item:
item[newprop] = item[prop]
mixed.append(item)
else:
mixed = []
for dataset in datasets:
#Replace the data set URL with the content
logger.debug("Processing dataset: %s" % (datasets[dataset]))
resp, content = H.request(datasets[dataset])
items = json.loads(content)[u'items']
mixed += items
for (counter, item) in enumerate(mixed):
item[u'id'] = u'_%i' % counter
result = json.dumps({'items': mixed}, indent=4)
return result
def __init__(self, root, logger=logging, cachedir='/tmp/.cache'):
'''
root - root of the OAI service endpoint, e.g. http://dspace.mit.edu/oai/request
'''
self.root = root
self.logger = logger
self.h = httplib2.Http(cachedir)
return
def listrecords(limit=100):
import httplib
h = httplib2.Http()
h.force_exception_as_status_code = True
url = join(COUCH_DATABASE, '_design', VIEW_APP, '_view', VIEW_NAME)
url += '?limit=' + str(limit)
logger.debug(url)
resp, content = h.request(url, "GET", headers=COUCH_AUTH_HEADER)
logger.debug("Content: " + content)
if str(resp.status).startswith('2'):
return content
else:
logger.error("Couldn't get documents via: " + repr(resp))
def oaisetname(body, ctype, sets_service=None):
'''
Service that accepts a JSON document and sets the "name" property based on looking up
the set in the HTTP_CONTEXT using the service passed in the 'sets_service' parameter.
Assumes that the set_service returns a JSON array of two-element arrays, where the first
element is the id and the second element the complete name.
'''
if not sets_service:
response.code = 500
response.add_header('content-type', 'text/plain')
return "No set service has been selected"
try:
data = json.loads(body)
except:
response.code = 500
response.add_header('content-type', 'text/plain')
return "Unable to parse body as JSON"
if not is_absolute(sets_service):
prefix = request.environ['wsgi.url_scheme'] + '://'
prefix += request.environ['HTTP_HOST'] if request.environ.get(
'HTTP_HOST') else request.environ['SERVER_NAME']
sets_service = prefix + sets_service
H = httplib2.Http('/tmp/.cache')
H.force_exception_as_status_code = True
resp, content = H.request(sets_service)
if not resp[u'status'].startswith('2'):
print >> sys.stderr, ' HTTP error (' + resp[
u'status'] + ') resolving URL: ' + sets_service
try:
sets = json.loads(content)
except:
response.code = 500
response.add_header('content-type', 'text/plain')
return "Unable to parse sets service result as JSON: " + repr(content)
setpos = data['_id'].find('--')
match = data['_id'][setpos + 2:] if setpos > -1 else data['_id']
for s in sets:
if match == s['setSpec']:
data[u'title'] = s['setName']
if s['setDescription']:
data[u'description'] = s['setDescription'].strip()
break
return json.dumps(data)
def update_document(body, ctype):
logger.debug(body)
from StringIO import StringIO
io = StringIO(body)
parsed_doc = json.load(io)
document_id = parsed_doc[u"id"]
document = body
logger.debug("Storing the document: " + document_id)
import httplib
h = httplib2.Http()
h.force_exception_as_status_code = True
url = join(COUCH_DATABASE, document_id)
resp, content = h.request(url, 'PUT', body=document, headers=COUCH_AUTH_HEADER)
if str(resp.status).startswith('2'):
return content
else:
logger.error("Couldn't store the document %s with the id: %s. " % (document, document_id, ) )
def nypl_identify_object(body, ctype, list_sets=None):
try:
assert ctype.lower() == HTTP_TYPE_JSON, "%s is not %s" % (
HTTP_HEADER_TYPE, HTTP_TYPE_JSON)
data = json.loads(body)
except Exception as e:
error_text = "Bad JSON: %s: %s" % (e.__class__.__name__, str(e))
logger.exception(error_text)
response.code = HTTP_INTERNAL_SERVER_ERROR
response.add_header(HTTP_HEADER_TYPE, HTTP_TYPE_TEXT)
return error_text
H = httplib2.Http('/tmp/.cache')
H.force_exception_as_status_code = True
resp, content = H.request(list_sets)
if not resp[u'status'].startswith('2'):
logger.error(' HTTP error (' + resp[u'status'] + ') resolving URL: ' +
list_sets)
return body
content_dict = xmltodict.parse(content,
xml_attribs=True,
attr_prefix='',
force_cdata=False,
ignore_whitespace_cdata=True)
sets = content_dict["nyplAPI"]["response"]
for r in sets:
if "collection" == r:
for coll_dict in sets[r]:
if "uuid" in coll_dict and "title" in coll_dict and (
coll_dict["uuid"] == data["title"]
or coll_dict["uuid"] in data["@id"]):
data["title"] = coll_dict["title"]
return json.dumps(data)
import sys
from server_support import server, print_error_log
from amara.thirdparty import httplib2
from amara.thirdparty import json
CT_JSON = {"Content-Type": "application/json"}
H = httplib2.Http()
def _get_server_response(body, prop=None):
url = server() + "decode_html?prop=%s" % prop
return H.request(url, "POST", body=body, headers=CT_JSON)
def test_decode_html():
"""Should decode ", &, <, and >"""
INPUT = {
"subject": ['a', 'b', '"&', 'c; <', 'd', '>;e']
}
EXPECTED = {
"subject": ['a', 'b', '\"&', 'c; <', 'd', '>;e']
}
resp, content = _get_server_response(json.dumps(INPUT), prop="subject")
assert resp.status == 200
assert json.loads(content) == EXPECTED
if __name__ == "__main__":
raise SystemExit("Use nosetest")
def post(body, sink):
headers = {'Content-type': 'application/json'}
h = httplib2.Http()
resp, content = h.request(sink, "POST", body=body, headers=headers)
return resp, content
def request(self, url, method, body, headers=HEADERS):
from amara.thirdparty import httplib2
h = httplib2.Http()
return h.request(url, method, body, headers)
