def dataproxy_resource_update(context, data_dict=None):
"""
Intercepts default resource_update action and encrypts password for dataproxy type resources
Args:
context: Request context.
data_dict: Parsed request parameters.
Returns:
see get_action('resource_update').
Raises:
Exception: if ckan.dataproxy.secret configuration not set.
"""
#If not set, default to empty string
data_dict['url_type'] = data_dict.get('url_type', '')
url_type = data_dict['url_type']
if url_type == 'dataproxy':
secret = config.get('ckan.dataproxy.secret', False)
if not secret:
raise Exception('ckan.dataproxy.secret must be defined to encrypt passwords')
#replace password with a _password_ placeholder
password = data_dict.get('db_password', '')
if password == '':
#we don't want to overwrite existing password with empty string
resource = Resource.get(data_dict['id'])
data_dict['db_password'] = resource.extras['db_password']
else:
data_dict['url'] = data_dict['url'].replace(password, '_password_')
#encrypt db_password
data_dict['db_password'] = hexlify(encrypt(secret, password))
site_url = config.get('ckan.site_url', '127.0.0.1')
data_dict['url'] = '{0}/api/3/action/datastore_search?resource_id={1}&downloaded=true'.format(site_url, data_dict['id'])
return orig_resource_update(context, data_dict)
def datastore_search(original_action, context, data_dict):
'''
Execute a datastore search using the custom method in dataproxy
'''
log.debug("datastore_search: context: {0} data_dict: {1}".format(context, data_dict))
logic.check_access('datastore_search', context)
resource = Resource.get(data_dict['resource_id'])
if resource is not None and resource.url_type == 'dataproxy':
# Discard sorting and only return 10 rows cuz it's a data preview
data_dict['sort'] = None
data_dict['limit'] = 10
# execute search
results=json.loads(SearchController().dataproxy_search(data_dict, resource))
log.debug("datastore_search: results: {0}".format(results))
# set total to limit
results['result']['total']=data_dict['limit']
# Hack in an _id field because datatables expects it. Add a _id value to each row
results['result']['fields']=[{"id":"_id", "type": "BIGINT"}] + results['result']['fields']
for i in range(0, len(results['result']['records'])):
results['result']['records'][i]['_id'] = i+1
return results['result']
#Default action otherwise
return original_action(context, data_dict)
def dataproxy_resource_update(context, data_dict=None):
"""
Intercepts default resource_update action and encrypts password for dataproxy type resources
Args:
context: Request context.
data_dict: Parsed request parameters.
Returns:
see get_action('resource_update').
Raises:
Exception: if ckan.dataproxy.secret configuration not set.
"""
#If not set, default to empty string
data_dict['url_type'] = data_dict.get('url_type', '')
url_type = data_dict['url_type']
if url_type == 'dataproxy':
secret = config.get('ckan.dataproxy.secret', False)
if not secret:
raise Exception(
'ckan.dataproxy.secret must be defined to encrypt passwords')
#replace password with a _password_ placeholder
password = data_dict.get('db_password', '')
if password == '':
#we don't want to overwrite existing password with empty string
resource = Resource.get(data_dict['id'])
data_dict['db_password'] = resource.extras['db_password']
else:
data_dict['url'] = data_dict['url'].replace(password, '_password_')
#encrypt db_password
data_dict['db_password'] = hexlify(encrypt(secret, password))
return orig_resource_update(context, data_dict)
def import_collection_to_package(params, id):
"""
Import a collection to dataset. Does not import whole file data but
rather the metadata.
"""
from irods import irodsCollection
path = params['path']
pkg = Package.get(id)
conn = get_connection_from_params(params)
if (conn):
coll = irodsCollection(conn, path)
from irods import iRodsOpen
rev = model.repo.new_revision()
i = 0
for obj in coll.getObjects():
extras = {}
fname, _ = obj
fpath = "%s/%s" % (coll.getCollName(), fname)
f = iRodsOpen(conn, fpath, 'r')
if f:
i += 1
res = Resource.by_name(fname)
if not res:
res = Resource(url = '', name=fname, extras=extras, \
resource_type='file')
for met in f.getUserMetadata():
key, value, _ = met
extras[key] = value
res.extras = extras
resgrp = pkg.resource_groups[0]
resgrp.resources.append(res)
Session.add(res)
Session.add(resgrp)
rev.message = "Update from iRODS, matched file %s" % fname
for met in coll.getUserMetadata():
key, value, _ = met
pkg.extras[key] = value
Session.add(pkg)
model.repo.commit()
conn.disconnect()
h.flash_success("iRODS import to dataset OK! Imported %s resources." %
i)
else:
h.flash_error("Could not connect to iRODS!")
h.redirect_to(controller='package', action='read', id=id)
def import_collection_to_package(params, id):
"""
Import a collection to dataset. Does not import whole file data but
rather the metadata.
"""
from irods import irodsCollection
path = params['path']
pkg = Package.get(id)
conn = get_connection_from_params(params)
if (conn):
coll = irodsCollection(conn, path)
from irods import iRodsOpen
rev = model.repo.new_revision()
i = 0
for obj in coll.getObjects():
extras = {}
fname, _ = obj
fpath = "%s/%s" % (coll.getCollName(), fname)
f = iRodsOpen(conn, fpath, 'r')
if f:
i += 1
res = Resource.by_name(fname)
if not res:
res = Resource(url = '', name=fname, extras=extras, \
resource_type='file')
for met in f.getUserMetadata():
key, value, _ = met
extras[key] = value
res.extras = extras
resgrp = pkg.resource_groups[0]
resgrp.resources.append(res)
Session.add(res)
Session.add(resgrp)
rev.message = "Update from iRODS, matched file %s" % fname
for met in coll.getUserMetadata():
key, value, _ = met
pkg.extras[key] = value
Session.add(pkg)
model.repo.commit()
conn.disconnect()
h.flash_success("iRODS import to dataset OK! Imported %s resources." % i)
else:
h.flash_error("Could not connect to iRODS!")
h.redirect_to(controller='package', action='read', id=id)
def search_action(self):
"""Routes dataproxy type resources to dataproxy_search method, else performs 'datastore_search' action"""
#TODO: No access control checks for dataproxy resources!
request_data = self._get_request_data(try_url_params=True)
resource = Resource.get(request_data['resource_id'])
if resource is not None and resource.url_type == 'dataproxy':
pylons.response.headers['Content-Type'] = 'application/json;charset=utf-8'
return self.dataproxy_search(request_data, resource)
#Default action otherwise
return self.action('datastore_search', ver=3)
def search_action(self):
"""Routes dataproxy type resources to dataproxy_search method, else performs 'datastore_search' action"""
#TODO: No access control checks for dataproxy resources!
request_data = self._get_request_data(try_url_params=True)
downloaded = False
if 'downloaded' in request_data:
downloaded = str(request_data['downloaded']).upper() == 'TRUE'
resource = Resource.get(request_data['resource_id'])
if resource is not None and resource.url_type == 'dataproxy':
if (downloaded):
pylons.response.headers['Content-Type'] = 'text/csv'
pylons.response.headers[
'Content-Disposition'] = 'attachment;filename="{0}.{1}"'.format(
resource.name, resource.format)
datas = json.loads(
self.dataproxy_search(request_data, resource))
result = datas['result']
tmp = ''
fields = result['fields']
for val in fields:
field = val['id']
if isinstance(field, unicode):
tmp += field.encode('utf-8') + ','
else:
tmp += str(field) + ','
records = result['records']
for row in records:
tmp += '\n'
for val in fields:
if val['id'] in row:
cell = row[val['id']]
if cell is not None:
if isinstance(cell, unicode):
tmp += '"' + cell.encode('utf-8') + '",'
else:
tmp += '"' + str(cell) + '",'
else:
tmp += ','
else:
tmp += ','
return str(tmp)
else:
pylons.response.headers[
'Content-Type'] = 'application/json;charset=utf-8'
return self.dataproxy_search(request_data, resource)
#Default action otherwise
return self.action('datastore_search', ver=3)
def search_action(self):
"""Routes dataproxy type resources to dataproxy_search method, else performs 'datastore_search' action"""
#TODO: No access control checks for dataproxy resources!
request_data = self._get_request_data(try_url_params=True)
log.info('{}'.format(request_data))
if 'resource_id' in request_data:
resource = Resource.get(request_data['resource_id'])
if resource is not None and resource.url_type == 'dataproxy':
pylons.response.headers['Content-Type'] = 'application/json;charset=utf-8'
return self.dataproxy_search(request_data, resource)
#Default action otherwise
return self.action('datastore_search', ver=3)
def update_res_license(context, res_dict, license_id):
session = context['session']
resource_id = res_dict['id']
log.debug(u'Updating license for resource {}: {}'.format(resource_id, license_id))
resource = Resource.get(res_dict['id'])
res_license = session.query(CeonResourceLicense).filter(CeonResourceLicense.resource_id == resource.id).first()
if res_license:
log.debug(u'Deleting license res_license: {}'.format(res_license))
session.delete(res_license)
new_res_license = CeonResourceLicense(resource_id = resource.id, license_id = license_id)
session.merge(new_res_license)
log.debug(u'Created license res_license: {}'.format(new_res_license))
return new_res_license
def search_action(self):
"""Routes dataproxy type resources to dataproxy_search method, else performs 'datastore_search' action"""
#TODO: No access control checks for dataproxy resources!
request_data = self._get_request_data(try_url_params=True)
downloaded = False
if 'downloaded' in request_data:
downloaded = str(request_data['downloaded']).upper() == 'TRUE'
resource = Resource.get(request_data['resource_id'])
if resource is not None and resource.url_type == 'dataproxy':
if(downloaded):
pylons.response.headers['Content-Type'] = 'text/csv'
pylons.response.headers['Content-Disposition'] = 'attachment;filename="{0}.{1}"'.format(resource.name, resource.format)
datas = json.loads(self.dataproxy_search(request_data, resource))
result = datas['result']
tmp = ''
fields = result['fields']
for val in fields:
field = val['id']
if isinstance(field, unicode):
tmp += field.encode('utf-8') + ','
else:
tmp += str(field) + ','
records = result['records']
for row in records:
tmp += '\n'
for val in fields:
if val['id'] in row:
cell = row[val['id']]
if cell is not None:
if isinstance(cell, unicode):
tmp += '"' + cell.encode('utf-8') + '",'
else:
tmp += '"' + str(cell) + '",'
else:
tmp += ','
else:
tmp += ','
return str(tmp)
else:
pylons.response.headers['Content-Type'] = 'application/json;charset=utf-8'
return self.dataproxy_search(request_data, resource)
#Default action otherwise
return self.action('datastore_search', ver=3)
def view(self, id):
"""
Renders a form for iRODS resource import, if it receives a POST request,
it simply calls sync_irods function in order to do the actual import.
"""
res = Resource.get(id)
context = {'model':model, 'user': c.user or c.author, 'resource': res }
try:
check_access('resource_update', context, {'id' : id})
except NotAuthorized:
abort(401, _('Not authorized to see this page'))
if ('save' in request.params):
sync_irods(request.params, id)
c.resource_name = res.name
c.resource_id = res.id
return render('ckanext/irods/irods.html')
def update_res_license(context, res_dict, license_id):
session = context['session']
resource_id = res_dict['id']
log.debug(u'Updating license for resource {}: {}'.format(resource_id, license_id))
resource = Resource.get(res_dict['id'])
res_license = session.query(CeonResourceLicense).filter(CeonResourceLicense.resource_id == resource.id).first()
if res_license:
res_license.license_id = license_id
log.debug(u'Updated license res_license: {}'.format(res_license))
session.merge(res_license)
return res_license
else:
new_res_license = CeonResourceLicense(resource_id = resource.id, license_id = license_id)
session.merge(new_res_license)
log.debug(u'Created license res_license: {}'.format(new_res_license))
return new_res_license
def view(self, id):
"""
Renders a form for iRODS resource import, if it receives a POST request,
it simply calls sync_irods function in order to do the actual import.
"""
res = Resource.get(id)
context = {'model': model, 'user': c.user or c.author, 'resource': res}
try:
check_access('resource_update', context, {'id': id})
except NotAuthorized:
abort(401, _('Not authorized to see this page'))
if ('save' in request.params):
sync_irods(request.params, id)
c.resource_name = res.name
c.resource_id = res.id
return render('ckanext/irods/irods.html')
def test_extractor_extract_update_ignored_format(self, send_task):
"""
extractor_extract for a resource with updated, ignored format.
"""
res_dict = factories.Resource(format='pdf')
send_task.reset_mock()
fake_process(res_dict)
resource = Resource.get(res_dict['id'])
resource.format = 'foo'
resource.save()
result = call_action('extractor_extract', id=res_dict['id'])
assert_equal(result['status'], 'ignored', 'Wrong state')
assert_true(result['task_id'] is None, 'Unexpected task ID')
assert_equal(send_task.call_count, 0,
'Wrong number of extraction tasks.')
assert_no_metadata(res_dict)
def test_extractor_extract_update_ignored_format(self, send_task):
"""
extractor_extract for a resource with updated, ignored format.
"""
res_dict = factories.Resource(format='pdf')
send_task.reset_mock()
fake_process(res_dict)
resource = Resource.get(res_dict['id'])
resource.format = 'foo'
resource.save()
result = call_action('extractor_extract', id=res_dict['id'])
assert_equal(result['status'], 'ignored', 'Wrong state')
assert_true(result['task_id'] is None, 'Unexpected task ID')
assert_equal(send_task.call_count, 0,
'Wrong number of extraction tasks.')
assert_no_metadata(res_dict)
def test_extractor_extract_update_indexed_format(self, send_task):
"""
extractor_extract for a resource with updated, indexed format.
"""
res_dict = factories.Resource(format='pdf')
send_task.reset_mock()
fake_process(res_dict)
resource = Resource.get(res_dict['id'])
resource.format = 'doc'
resource.save()
result = call_action('extractor_extract', id=res_dict['id'])
assert_equal(result['status'], 'update', 'Wrong state')
assert_false(result['task_id'] is None, 'Missing task ID')
assert_equal(result['task_id'],
get_metadata(res_dict).task_id, 'Task IDs differ.')
assert_equal(send_task.call_count, 1,
'Wrong number of extraction tasks.')
def test_extractor_extract_update_indexed_format(self, send_task):
"""
extractor_extract for a resource with updated, indexed format.
"""
res_dict = factories.Resource(format='pdf')
send_task.reset_mock()
fake_process(res_dict)
resource = Resource.get(res_dict['id'])
resource.format = 'doc'
resource.save()
result = call_action('extractor_extract', id=res_dict['id'])
assert_equal(result['status'], 'update', 'Wrong state')
assert_false(result['task_id'] is None, 'Missing task ID')
assert_equal(result['task_id'], get_metadata(res_dict).task_id,
'Task IDs differ.')
assert_equal(send_task.call_count, 1,
'Wrong number of extraction tasks.')
def update_resource_url(context, res_dict):
if not 'url' in res_dict or not res_dict['url']:
return res_dict
if not 'url_type' in res_dict or 'upload' != res_dict['url_type']:
return res_dict
log.debug(u"Updating resource {} url {}".format(res_dict['id'], res_dict['url']))
res_url = remove_locales_from_url(res_dict['url'])
log.debug(u"new url {}".format(res_url))
if res_dict['url'] != res_url:
log.debug(u"here 1")
res_dict['url'] = res_url
session = context['session']
res = Resource.get(res_dict['id'])
log.debug(u"here 2 {}".format(res))
if not res:
raise Exception(u'Resource "{}" not found'.format(res_dict['id']))
res.url = res_url
session.merge(res)
log.debug(u"here 3 {}".format(res))
return res_dict
def _get_ids(self, only_with_metadata=False):
"""
Get list of resource IDs from command line arguments.
Returns the specific IDs listed or all IDs if ``all`` was passed.
If ``only_with_metadata`` is true and ``all`` was passed then only
IDs of resources which have metadata are returned.
"""
from ckan.plugins import toolkit
if len(self.args) < 1:
_error('Missing argument. Specify one or more resource IDs '
+ 'or "all".')
if len(self.args) == 1 and self.args[0].lower() == 'all':
if only_with_metadata:
return sorted(toolkit.get_action('extractor_list')({}, {}))
else:
from ckan.model import Resource
return sorted(r.id for r in Resource.active())
else:
return self.args[:]
def test_link_and_map_shown(self):
name = u'annakarenina'
wms_url = 'http://maps.bgs.ac.uk/ArcGIS/services/BGS_Detailed_Geology/MapServer/WMSServer?'
rev = model.repo.new_revision()
pkg = Package.get(name)
pr = Resource(url=wms_url, format='WMS')
pkg.resources.append(pr)
pkg.save()
model.repo.commit_and_remove()
# Load the dataset page and check if link appears
offset = url_for(controller='package', action='read', id=name)
res = self.app.get(offset)
assert 'View available WMS layers' in res, res
# Load the dataset map preview page and check if libraries are loaded
offset = '/dataset/%s/map' % name
res = self.app.get(offset)
assert '<script type="text/javascript" src="/ckanext/spatial/js/wms_preview.js"></script>' in res, res
assert 'CKAN.WMSPreview.setup("%s");' % wms_url.split('?')[0] in res
def sync_irods(params, id):
"""
Fetches a resource from database with the same path as user specified and
that matches an existing resource in CKAN.
"""
from irods import getFileUserMetadata, rcModAccessControl
rev = model.repo.new_revision()
conn = get_connection_from_params(params)
resource = Resource.get(id)
path = params['path']
extras = {}
# Lets handle only resources with file names
if resource.name:
fname = "%s/%s" % (path, resource.name.split('/')[-1])
log.debug(fname)
i = 0
access = rcModAccessControl()
log.debug(access.getPath())
if conn:
for met in getFileUserMetadata(conn, fname):
i += 1
key, value, _ = met
extras[key] = value
resource.extras = extras
Session.add(resource)
conn.disconnect()
model.repo.commit()
rev.message = "Update from iRODS, matched file %s" % fname
h.flash_success(
"iRODS import to resource OK! Imported %s metadatas" % i)
else:
h.flash_error("Could not connect to iRODS!")
else:
h.flash_error("Resource is an URL, cannot import!")
h.redirect_to(controller='package', action='resource_read', \
id=resource.resource_group.package.name, \
resource_id=resource.id)
def sync_irods(params, id):
"""
Fetches a resource from database with the same path as user specified and
that matches an existing resource in CKAN.
"""
from irods import getFileUserMetadata, rcModAccessControl
rev = model.repo.new_revision()
conn = get_connection_from_params(params)
resource = Resource.get(id)
path = params['path']
extras = {}
# Lets handle only resources with file names
if resource.name:
fname = "%s/%s" % (path, resource.name.split('/')[-1])
log.debug(fname)
i = 0
access = rcModAccessControl()
log.debug(access.getPath())
if conn:
for met in getFileUserMetadata(conn, fname):
i += 1
key, value, _ = met
extras[key] = value
resource.extras = extras
Session.add(resource)
conn.disconnect()
model.repo.commit()
rev.message = "Update from iRODS, matched file %s" % fname
h.flash_success("iRODS import to resource OK! Imported %s metadatas" % i)
else:
h.flash_error("Could not connect to iRODS!")
else:
h.flash_error("Resource is an URL, cannot import!")
h.redirect_to(controller='package', action='resource_read', \
id=resource.resource_group.package.name, \
resource_id=resource.id)
def read_data(self, id, resource_id):
res = Resource.get(resource_id)
pkg = Package.get(id)
c.pkg_dict = pkg.as_dict()
c.package = pkg
c.resource = get_action('resource_show')({'model': model},
{'id': resource_id})
label = res.url.split(config.get('ckan.site_url') + '/storage/f/')[-1]
label = urllib2.unquote(label)
ofs = get_ofs()
try:
furl = ofs.get_url(BUCKET, label).split('file://')[-1]
except FileNotFoundException:
h.flash_error(_('Cannot do data mining on remote resource!'))
url = h.url_for(controller='package', action='resource_read',
id=id, resource_id=resource_id)
return redirect(url)
wordstats = {}
ret = {}
if res.format in ('TXT', 'txt'):
wdsf, wdspath = tempfile.mkstemp()
os.write(wdsf, "%s\nmetadata description title information" % furl)
with os.fdopen(wdsf, 'r') as wordfile:
preproc = orngText.Preprocess()
table = orngText.loadFromListWithCategories(wdspath)
data = orngText.bagOfWords(table, preprocessor=preproc)
words = orngText.extractWordNGram(data, threshold=10.0, measure='MI')
for i in range(len(words)):
d = words[i]
wordstats = d.get_metas(str)
for k, v in wordstats.items():
if v.value > 10.0:
ret[unicode(k, 'utf8')] = v.value
from operator import itemgetter
c.data_tags = sorted(ret.iteritems(), key=itemgetter(1), reverse=True)[:30]
os.remove(wdspath)
for i in range(len(data)):
d = words[i]
wordstats = d.get_metas(str)
words = []
for k, v in wordstats.items():
words.append(k)
model.repo.new_revision()
if not 'autoextracted_description' in pkg.extras:
pkg.extras['autoextracted_description'] = ' '.join(words)
pkg.save()
return render('datamining/read.html')
elif res.format in ('odt', 'doc', 'xls', 'ods', 'odp', 'ppt', 'doc', 'html'):
textfd, textpath = convert_to_text(res, furl)
if not textpath:
h.flash_error(_('This file could not be mined for any data!'))
os.close(textfd)
return render('datamining/read.html')
else:
wdsf, wdspath = tempfile.mkstemp()
os.write(wdsf, "%s\nmetadata description title information" % textpath)
preproc = orngText.Preprocess()
table = orngText.loadFromListWithCategories(wdspath)
data = orngText.bagOfWords(table, preprocessor=preproc)
words = orngText.extractWordNGram(data, threshold=10.0, measure='MI')
for i in range(len(words)):
d = words[i]
wordstats = d.get_metas(str)
for k, v in wordstats.items():
if v.value > 10.0:
ret[unicode(k, 'utf8')] = v.value
from operator import itemgetter
c.data_tags = sorted(ret.iteritems(), key=itemgetter(1), reverse=True)[:30]
os.close(textfd)
os.close(wdsf)
os.remove(wdspath)
os.remove(textpath)
for i in range(len(data)):
d = words[i]
wordstats = d.get_metas(str)
words = []
for k, v in wordstats.items():
log.debug(k)
words.append(substitute_ascii_equivalents(k))
model.repo.new_revision()
if not 'autoextracted_description' in pkg.extras:
pkg.extras['autoextracted_description'] = ' '.join(words)
pkg.save()
return render('datamining/read.html')
else:
h.flash_error(_('This metadata document is not in proper format for data mining!'))
url = h.url_for(controller='package', action='resource_read',
id=id, resource_id=resource_id)
return redirect(url)
