Newer
Older
#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
pybis.py
"""
Swen Vermeul
committed
from __future__ import print_function
import os
Chandrasekhar Ramakrishnan
committed
import random
from requests.packages.urllib3.exceptions import InsecureRequestWarning
requests.packages.urllib3.disable_warnings(InsecureRequestWarning)
Swen Vermeul
committed
import time
from urllib.parse import urlparse, urljoin, quote
from collections import namedtuple
from texttable import Texttable
from tabulate import tabulate
Chandrasekhar Ramakrishnan
committed
from . import data_set as pbds
Swen Vermeul
committed
from .utils import parse_jackson, check_datatype, split_identifier, format_timestamp, is_identifier, is_permid, nvl, VERBOSE
from .property import PropertyHolder, PropertyAssignments
from .masterdata import Vocabulary
from .openbis_object import OpenBisObject
from .definitions import fetch_option
# import the various openBIS entities
from .space import Space
from .project import Project
from .experiment import Experiment
from .sample import Sample
from .dataset import DataSet
from .person import Person
from .group import Group
from .role_assignment import RoleAssignment
from .tag import Tag
from .semantic_annotation import SemanticAnnotation
Swen Vermeul
committed
import pandas as pd
Chandrasekhar Ramakrishnan
committed
from datetime import datetime
LOG_NONE = 0
LOG_SEVERE = 1
LOG_ERROR = 2
LOG_WARNING = 3
LOG_INFO = 4
LOG_ENTRY = 5
LOG_PARM = 6
LOG_DEBUG = 7
DEBUG_LEVEL = LOG_NONE
def get_search_type_for_entity(entity, operator=None):
""" Returns a dictionary containing the correct search criteria type
for a given entity.
get_search_type_for_entity('space')
# returns:
{'@type': 'as.dto.space.search.SpaceSearchCriteria'}
"""
search_criteria = {
"space": "as.dto.space.search.SpaceSearchCriteria",
"userId": "as.dto.person.search.UserIdSearchCriteria",
"email": "as.dto.person.search.EmailSearchCriteria",
"firstName": "as.dto.person.search.FirstNameSearchCriteria",
"lastName": "as.dto.person.search.LastNameSearchCriteria",
"project": "as.dto.project.search.ProjectSearchCriteria",
"experiment": "as.dto.experiment.search.ExperimentSearchCriteria",
"experiment_type": "as.dto.experiment.search.ExperimentTypeSearchCriteria",
"sample": "as.dto.sample.search.SampleSearchCriteria",
"sample_type": "as.dto.sample.search.SampleTypeSearchCriteria",
"dataset": "as.dto.dataset.search.DataSetSearchCriteria",
"dataset_type": "as.dto.dataset.search.DataSetTypeSearchCriteria",
"external_dms": "as.dto.externaldms.search.ExternalDmsSearchCriteria",
"material": "as.dto.material.search.MaterialSearchCriteria",
"material_type": "as.dto.material.search.MaterialTypeSearchCriteria",
"vocabulary_term": "as.dto.vocabulary.search.VocabularyTermSearchCriteria",
"tag": "as.dto.tag.search.TagSearchCriteria",
"authorizationGroup": "as.dto.authorizationgroup.search.AuthorizationGroupSearchCriteria",
"roleAssignment": "as.dto.roleassignment.search.RoleAssignmentSearchCriteria",
"person": "as.dto.person.search.PersonSearchCriteria",
"code": "as.dto.common.search.CodeSearchCriteria",
"sample_type": "as.dto.sample.search.SampleTypeSearchCriteria",
"global": "as.dto.global.GlobalSearchObject",
}
sc = { "@type": search_criteria[entity] }
if operator is not None:
sc["operator"] = operator
return sc
def get_attrs_for_entity(entity):
""" For a given entity this method returns an iterator for all searchable
attributes.
"""
search_args = {
"person": ['firstName','lastName','email','userId']
}
for search_arg in search_args[entity]:
yield search_arg
def search_request_for_identifier(ident, entity):
search_request = {}
Swen Vermeul
committed
if is_identifier(ident):
search_request = {
"identifier": ident.upper(),
"@type": "as.dto.{}.id.{}Identifier".format(entity.lower(), entity.capitalize())
}
else:
search_request = {
"permId": ident,
"@type": "as.dto.{}.id.{}PermId".format(entity.lower(), entity.capitalize())
}
return search_request
def get_search_criteria(entity, **search_args):
search_criteria = get_search_type_for_entity(entity)
criteria = []
for attr in get_attrs_for_entity(entity):
if attr in search_args:
sub_crit = get_search_type_for_entity(attr)
sub_crit['fieldValue'] = get_field_value_search(attr, search_args[attr])
criteria.append(sub_crit)
search_criteria['criteria'] = criteria
search_criteria['operator'] = "AND"
return search_criteria
return '' if obj is None else str(obj)
return '' if obj['code'] is None else obj['code']
def extract_deletion(obj):
del_objs = []
for deleted_object in obj['deletedObjects']:
del_objs.append({
"reason": obj['reason'],
"permId": deleted_object["id"]["permId"],
"type": deleted_object["id"]["@type"]
})
return del_objs
def extract_identifier(ident):
if not isinstance(ident, dict):
return str(ident)
return ident['identifier']
def extract_nested_identifier(ident):
if not isinstance(ident, dict):
return str(ident)
return ident['identifier']['identifier']
def extract_permid(permid):
if not isinstance(permid, dict):
return str(permid)
return permid['permId']
def extract_nested_permid(permid):
if not isinstance(permid, dict):
return '' if permid is None else str(permid)
return '' if permid['permId']['permId'] is None else permid['permId']['permId']
def extract_property_assignments(pas):
pa_strings = []
for pa in pas:
if not isinstance(pa['propertyType'], dict):
pa_strings.append(pa['propertyType'])
else:
pa_strings.append(pa['propertyType']['label'])
return pa_strings
def extract_role_assignments(ras):
ra_strings = []
for ra in ras:
ra_strings.append({
"role": ra['role'],
"roleLevel": ra['roleLevel'],
"space": ra['space']['code'] if ra['space'] else None
})
return ra_strings
def extract_person(person):
def extract_person_details(person):
if not isinstance(person, dict):
return str(person)
return "{} {} <{}>".format(
person['firstName'],
person['lastName'],
person['email']
)
def extract_id(id):
if not isinstance(id, dict):
return str(id)
else:
return id['techId']
def extract_userId(user):
if isinstance(user, list):
return ", ".join([
u['userId'] for u in user
])
elif isinstance(user, dict):
return user['userId']
else:
return str(user)
"""since Python3 the zlib module returns unsigned integers (2.7: signed int)
"""
for eachLine in open(fileName, "rb"):
prev = zlib.crc32(eachLine, prev)
# return as hex
return "%x" % (prev & 0xFFFFFFFF)
def _create_tagIds(tags=None):
if tags is None:
return None
if not isinstance(tags, list):
tags = [tags]
Swen Vermeul
committed
tagIds.append({
"code": tag,
"@type": "as.dto.tag.id.TagCode"
})
return tagIds
Swen Vermeul
committed
def _tagIds_for_tags(tags=None, action='Add'):
Swen Vermeul
committed
"""creates an action item to add or remove tags.
Action is either 'Add', 'Remove' or 'Set'
"""
if tags is None:
return
if not isinstance(tags, list):
tags = [tags]
items = []
for tag in tags:
items.append({
"code": tag,
"@type": "as.dto.tag.id.TagCode"
})
tagIds = {
"actions": [
{
"items": items,
"@type": "as.dto.common.update.ListUpdateAction{}".format(action.capitalize())
}
],
"@type": "as.dto.common.update.IdListUpdateValue"
}
def _list_update(ids=None, entity=None, action='Add'):
"""creates an action item to add, set or remove ids.
"""
if ids is None:
return
if not isinstance(ids, list):
ids = [ids]
items = []
for ids in ids:
items.append({
"code": ids,
"@type": "as.dto.{}.id.{}Code".format(entity.lower(), entity)
})
list_update = {
"actions": [
{
"items": items,
"@type": "as.dto.common.update.ListUpdateAction{}".format(action.capitalize())
}
],
"@type": "as.dto.common.update.IdListUpdateValue"
}
return list_update
def _create_typeId(type):
return {
"permId": type.upper(),
"@type": "as.dto.entitytype.id.EntityTypePermId"
}
def _create_projectId(ident):
match = re.match('/', ident)
if match:
return {
"identifier": ident,
"@type": "as.dto.project.id.ProjectIdentifier"
}
else:
return {
"permId": ident,
"@type": "as.dto.project.id.ProjectPermId"
}
def _create_experimentId(ident):
return {
"identifier": ident,
"@type": "as.dto.experiment.id.ExperimentIdentifier"
}
def get_field_value_search(field, value, comparison="StringEqualToValue"):
return {
"value": value,
"@type": "as.dto.common.search.{}".format(comparison)
}
def _common_search(search_type, value, comparison="StringEqualToValue"):
sreq = {
"@type": search_type,
"fieldValue": {
"value": value,
"@type": "as.dto.common.search.{}".format(comparison)
}
}
return sreq
def _criteria_for_code(code):
return {
"fieldValue": {
"value": code.upper(),
"@type": "as.dto.common.search.StringEqualToValue"
},
"@type": "as.dto.common.search.CodeSearchCriteria"
}
def _subcriteria_for_userId(userId):
return {
"criteria": [
{
"fieldName": "userId",
"fieldType": "ATTRIBUTE",
"fieldValue": {
"value": userId,
"@type": "as.dto.common.search.StringEqualToValue"
},
"@type": "as.dto.person.search.UserIdSearchCriteria"
}
],
"@type": "as.dto.person.search.PersonSearchCriteria",
"operator": "AND"
}
def _subcriteria_for_type(code, entity):
return {
"@type": "as.dto.{}.search.{}TypeSearchCriteria".format(entity.lower(), entity),
"criteria": [
{
"@type": "as.dto.common.search.CodeSearchCriteria",
"fieldValue": {
"value": code.upper(),
"@type": "as.dto.common.search.StringEqualToValue"
}
}
}
def _subcriteria_for_status(status_value):
status_value = status_value.upper()
valid_status = "AVAILABLE LOCKED ARCHIVED UNARCHIVE_PENDING ARCHIVE_PENDING BACKUP_PENDING".split()
if not status_value in valid_status:
raise ValueError("status must be one of the following: " + ", ".join(valid_status))
return {
"@type": "as.dto.dataset.search.PhysicalDataSearchCriteria",
"operator": "AND",
"criteria": [{
"@type":
"as.dto.dataset.search.StatusSearchCriteria",
"fieldName": "status",
"fieldValue": status_value
sreq = {}
for key, val in req.items():
if key == "criteria":
items = []
for item in req['criteria']:
items.append(_gen_search_criteria(item))
sreq['criteria'] = items
elif key == "code":
sreq["criteria"] = [_common_search(
"as.dto.common.search.CodeSearchCriteria", val.upper()
)]
elif key == "identifier":
if is_identifier(val):
# if we have an identifier, we need to search in Space and Code separately
si = split_identifier(val)
sreq["criteria"] = []
if "space" in si:
sreq["criteria"].append(
_gen_search_criteria({"space": "Space", "code": si["space"]})
)
if "experiment" in si:
pass
if "code" in si:
sreq["criteria"].append(
_common_search(
"as.dto.common.search.CodeSearchCriteria", si["code"].upper()
)
elif is_permid(val):
sreq["criteria"] = [_common_search(
"as.dto.common.search.PermIdSearchCriteria", val
)]
else:
# we assume we just got a code
sreq["criteria"] = [_common_search(
"as.dto.common.search.CodeSearchCriteria", val.upper()
)]
elif key == "operator":
sreq["operator"] = val.upper()
else:
sreq["@type"] = "as.dto.{}.search.{}SearchCriteria".format(key, val)
return sreq
def _subcriteria_for_tags(tags):
if not isinstance(tags, list):
tags = [tags]
criterias = []
for tag in tags:
criterias.append({
"fieldName": "code",
"fieldType": "ATTRIBUTE",
"fieldValue": {
"value": tag,
"@type": "as.dto.common.search.StringEqualToValue"
},
"@type": "as.dto.common.search.CodeSearchCriteria"
})
return {
"@type": "as.dto.tag.search.TagSearchCriteria",
"operator": "AND",
"criteria": criterias
}
def _subcriteria_for_is_finished(is_finished):
return {
"@type": "as.dto.common.search.StringPropertySearchCriteria",
"fieldName": "FINISHED_FLAG",
"fieldType": "PROPERTY",
"fieldValue": {
"value": is_finished,
"@type": "as.dto.common.search.StringEqualToValue"
}
}
def _subcriteria_for_properties(prop, val):
return {
"@type": "as.dto.common.search.StringPropertySearchCriteria",
"fieldName": prop.upper(),
"fieldType": "PROPERTY",
"fieldValue": {
"value": val,
"@type": "as.dto.common.search.StringEqualToValue"
}
}
def _subcriteria_for_permid(permids, entity, parents_or_children='', operator='AND'):
if not isinstance(permids, list):
permids = [permids]
criterias = []
for permid in permids:
criterias.append({
"@type": "as.dto.common.search.PermIdSearchCriteria",
"fieldValue": {
"value": permid,
"@type": "as.dto.common.search.StringEqualToValue"
},
"fieldType": "ATTRIBUTE",
"fieldName": "code"
criteria = {
"criteria": criterias,
"@type": "as.dto.{}.search.{}{}SearchCriteria".format(
entity.lower(), entity, parents_or_children
),
}
return criteria
def _subcriteria_for_code(code, object_type):
""" Creates the often used search criteria for code values. Returns a dictionary.
Example::
_subcriteria_for_code("username", "space")
{
"criteria": [
{
"fieldType": "ATTRIBUTE",
"@type": "as.dto.common.search.CodeSearchCriteria",
"fieldName": "code",
"fieldValue": {
"@type": "as.dto.common.search.StringEqualToValue",
"value": "USERNAME"
}
}
],
"operator": "AND",
"@type": "as.dto.space.search.SpaceSearchCriteria"
}
"""
if code is not None:
if is_permid(code):
fieldname = "permId"
fieldtype = "as.dto.common.search.PermIdSearchCriteria"
else:
fieldname = "code"
fieldtype = "as.dto.common.search.CodeSearchCriteria"
search_criteria = get_search_type_for_entity(object_type.lower())
search_criteria['criteria'] = [{
"fieldName": fieldname,
"fieldType": "ATTRIBUTE",
"fieldValue": {
"value": code.upper(),
"@type": "as.dto.common.search.StringEqualToValue"
},
"@type": fieldtype
}]
search_criteria["operator"] = "AND"
return search_criteria
return get_search_type_for_entity(object_type.lower())
class Openbis:
Swen Vermeul
committed
"""Interface for communicating with openBIS.
A recent version of openBIS is required (minimum 16.05.2).
For creation of datasets, dataset-uploader-api needs to be installed.
def __init__(self, url=None, verify_certificates=True, token=None):
"""Initialize a new connection to an openBIS server.
"""
if url is None:
try:
url = os.environ["OPENBIS_URL"]
token = os.environ["OPENBIS_TOKEN"] if "OPENBIS_TOKEN" in os.environ else None
except KeyError:
raise ValueError("please provide a URL you want to connect to.")
if url_obj.netloc is None:
raise ValueError("please provide the url in this format: https://openbis.host.ch:8443")
Swen Vermeul
committed
if url_obj.hostname is None:
raise ValueError("hostname is missing")
self.url = url_obj.geturl()
self.port = url_obj.port
self.as_v3 = '/openbis/openbis/rmi-application-server-v3.json'
self.as_v1 = '/openbis/openbis/rmi-general-information-v1.json'
self.reg_v1 = '/openbis/openbis/rmi-query-v1.json'
Chandrasekhar Ramakrishnan
committed
self.verify_certificates = verify_certificates
Swen Vermeul
committed
self.token = token
self.dataset_types = None
self.sample_types = None
Swen Vermeul
committed
#self.files_in_wsp = []
Swen Vermeul
committed
self.token_path = None
# use an existing token, if available
if self.token is None:
elif self.is_token_valid(token):
pass
else:
print("Session is no longer valid. Please log in again.")
def __dir__(self):
return [
'url', 'port', 'hostname',
'login()', 'logout()', 'is_session_active()', 'token', 'is_token_valid("")',
"get_dataset('permId')",
"get_datasets()",
"get_dataset_type('raw_data')",
"get_dataset_types()",
"get_datastores()",
"get_deletions()",
"get_experiment('permId', withAttachments=False)",
"get_experiments()",
"get_experiment_type('type')",
"get_experiment_types()",
"get_external_data_management_system(permId)",
"get_material_type('type')",
"get_material_types()",
"get_project('project')",
"get_projects(space=None, code=None)",
"get_sample('id')",
"get_object('id')", # "get_sample('id')" alias
"get_samples()",
"get_objects()", # "get_samples()" alias
"get_sample_type(type))",
"get_object_type(type))", # "get_sample_type(type))" alias
"get_sample_types()",
"get_object_types()", # "get_sample_types()" alias
"get_semantic_annotations()",
"get_semantic_annotation(permId, only_data = False)",
"get_space(code)",
"get_spaces()",
"get_tags()",
"get_terms()",
"new_person(userId, space)",
"get_person(userId)",
"get_groups()",
"get_group(code)",
Swen Vermeul
committed
"get_role_assignments()",
"get_role_assignment(techId)",
"new_group(code, description, userIds)",
'new_space(name, description)',
'new_project(space, code, description, attachments)',
'new_experiment(type, code, project, props={})',
'new_sample(type, space, project, experiment, parents)',
'new_object(type, space, project, experiment, parents)', # 'new_sample(type, space, project, experiment)' alias
'new_dataset(type, parent, experiment, sample, files=[], folder, props={})',
'new_semantic_annotation(entityType, propertyType)',
'update_sample(sampleId, space, project, experiment, parents, children, components, properties, tagIds, attachments)',
'update_object(sampleId, space, project, experiment, parents, children, components, properties, tagIds, attachments)', # 'update_sample(sampleId, space, project, experiment, parents, children, components, properties, tagIds, attachments)' alias
]
@property
def spaces(self):
return self.get_spaces()
@property
def projects(self):
return self.get_projects()
Swen Vermeul
committed
"""Read the token from the cache, and set the token ivar to it, if there, otherwise None.
If the token is not valid anymore, delete it.
"""
token_path = self.gen_token_path()
Chandrasekhar Ramakrishnan
committed
if not os.path.exists(token_path):
Swen Vermeul
committed
return None
Chandrasekhar Ramakrishnan
committed
try:
with open(token_path) as f:
Swen Vermeul
committed
token = f.read()
if token == "":
return None
Swen Vermeul
committed
if not self.is_token_valid(token):
Chandrasekhar Ramakrishnan
committed
os.remove(token_path)
Swen Vermeul
committed
return None
else:
return token
except FileNotFoundError:
Swen Vermeul
committed
return None
Swen Vermeul
committed
def gen_token_path(self, parent_folder=None):
"""generates a path to the token file.
The token is usually saved in a file called
~/.pybis/hostname.token
"""
Chandrasekhar Ramakrishnan
committed
if parent_folder is None:
Swen Vermeul
committed
# save token under ~/.pybis folder
parent_folder = os.path.join(
os.path.expanduser("~"),
'.pybis'
)
path = os.path.join(parent_folder, self.hostname + '.token')
Chandrasekhar Ramakrishnan
committed
return path
Swen Vermeul
committed
def save_token(self, token=None, parent_folder=None):
""" saves the session token to the disk, usually here: ~/.pybis/hostname.token. When a new Openbis instance is created, it tries to read this saved token by default.
Swen Vermeul
committed
if token is None:
token = self.token
token_path = None;
if parent_folder is None:
token_path = self.gen_token_path()
else:
token_path = self.gen_token_path(parent_folder)
# create the necessary directories, if they don't exist yet
Chandrasekhar Ramakrishnan
committed
os.makedirs(os.path.dirname(token_path), exist_ok=True)
with open(token_path, 'w') as f:
Swen Vermeul
committed
f.write(token)
self.token_path = token_path
def delete_token(self, token_path=None):
Swen Vermeul
committed
if token_path is None:
token_path = self.token_path
os.remove(token_path)
""" internal method, used to handle all post requests and serializing / deserializing
data
"""
return self._post_request_full_url(urljoin(self.url,resource), request)
def _post_request_full_url(self, full_url, request):
""" internal method, used to handle all post requests and serializing / deserializing
data
"""
if "jsonrpc" not in request:
request["jsonrpc"] = "2.0"
if request["params"][0] is None:
raise ValueError("Your session expired, please log in again")
if DEBUG_LEVEL >=LOG_DEBUG: print(json.dumps(request))
full_url,
json.dumps(request),
Swen Vermeul
committed
if DEBUG_LEVEL >= LOG_ERROR: print(json.dumps(request))
raise ValueError(resp['error']['message'])
else:
raise ValueError('request did not return either result nor error')
else:
raise ValueError('general error while performing post request')
""" Log out of openBIS. After logout, the session token is no longer valid.
if self.token is None:
return
"method": "logout",
"params": [self.token],
resp = self._post_request(self.as_v3, logout_request)
Swen Vermeul
committed
self.token = None
self.token_path = None
Swen Vermeul
committed
def login(self, username=None, password=None, save_token=False):
"""Log into openBIS.
Expects a username and a password and updates the token (session-ID).
The token is then used for every request.
Chandrasekhar Ramakrishnan
committed
Clients may want to store the credentials object in a credentials store after successful login.
Throw a ValueError with the error message if login failed.
"""
if password is None:
import getpass
password = getpass.getpass()
"method": "login",
"params": [username, password],
result = self._post_request(self.as_v3, login_request)
if result is None:
raise ValueError("login to openBIS failed")
else:
self.token = result
Swen Vermeul
committed
if save_token:
self.save_token()
return self.token
def create_permId(self):
Chandrasekhar Ramakrishnan
committed
"""Have the server generate a new permId"""
Chandrasekhar Ramakrishnan
committed
# Request just 1 permId
request = {
"method": "createPermIdStrings",
"params": [self.token, 1],
}
resp = self._post_request(self.as_v3, request)
if resp is not None:
return resp[0]
else:
raise ValueError("Could not create permId")
Chandrasekhar Ramakrishnan
committed
def get_datastores(self):
""" Get a list of all available datastores. Usually there is only one, but in some cases
there might be multiple servers. If you upload a file, you need to specifiy the datastore you want
request = {
"method": "listDataStores",
"params": [self.token],
}
resp = self._post_request(self.as_v1, request)
if resp is not None:
return DataFrame(resp)[['code', 'downloadUrl', 'hostUrl']]
raise ValueError("No datastore found!")
def new_person(self, userId, space=None):
""" creates an openBIS person
"""
try:
person = self.get_person(userId=userId)
except Exception:
return Person(self, userId=userId, space=space)
raise ValueError(
"There already exists a user with userId={}".format(userId)
)
def new_group(self, code, description=None, userIds=None):
""" creates an openBIS person
"""
return Group(self, code=code, description=description, userIds=userIds)
def get_group(self, code, only_data=False):
""" Get an openBIS AuthorizationGroup. Returns a Group object.
ids = [{
"@type": "as.dto.authorizationgroup.id.AuthorizationGroupPermId",
}]
fetchopts = {}
for option in ['roleAssignments', 'users', 'registrator']:
fetchopts[option] = fetch_option[option]
fetchopts['users']['space'] = fetch_option['space']
request = {
"method": "getAuthorizationGroups",
"params": [
self.token,
ids,
fetchopts
]
}
resp = self._post_request(self.as_v3, request)
if len(resp) == 0:
raise ValueError("No group found!")
for permid in resp:
group = resp[permid]
parse_jackson(group)
if only_data:
return group
else:
return Group(self, data=group)
Swen Vermeul
committed
def get_role_assignments(self, **search_args):
""" Get the assigned roles for a given group, person or space
"""
search_criteria = get_search_type_for_entity('roleAssignment', 'AND')
Swen Vermeul
committed
allowed_search_attrs = ['role', 'roleLevel', 'user', 'group', 'person', 'space']
sub_crit = []
for attr in search_args:
if attr in allowed_search_attrs:
if attr == 'space':
sub_crit.append(
_subcriteria_for_code(search_args[attr], 'space')
)
Swen Vermeul
committed
elif attr == 'person':
userId = ''
if isinstance(search_args[attr], str):
userId = search_args[attr]
else:
userId = search_args[attr].userId
sub_crit.append(
_subcriteria_for_userId(userId)
)
elif attr == 'group':
groupId = ''
if isinstance(search_args[attr], str):
groupId = search_args[attr]
else:
groupId = search_args[attr].code
_subcriteria_for_permid(groupId, 'AuthorizationGroup')
)
elif attr == 'role':
# TODO
raise ValueError("not yet implemented")
elif attr == 'roleLevel':
# TODO
raise ValueError("not yet implemented")
else:
pass
else:
Swen Vermeul
committed
raise ValueError("unknown search argument {}".format(attr))
search_criteria['criteria'] = sub_crit
fetchopts = {}
for option in ['roleAssignments', 'space', 'project', 'user', 'authorizationGroup','registrator']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchRoleAssignments",
"params": [
self.token,
search_criteria,
fetchopts
]
}
Swen Vermeul
committed
attrs=['techId', 'role', 'roleLevel', 'user', 'group', 'space', 'project']
resp = self._post_request(self.as_v3, request)
if len(resp['objects']) == 0:
Swen Vermeul
committed
roles = DataFrame(columns=attrs)
Swen Vermeul
committed
else:
objects = resp['objects']
parse_jackson(objects)
roles = DataFrame(objects)
roles['techId'] = roles['id'].map(extract_id)
roles['user'] = roles['user'].map(extract_userId)
roles['group'] = roles['authorizationGroup'].map(extract_code)
roles['space'] = roles['space'].map(extract_code)
roles['project'] = roles['project'].map(extract_code)
Swen Vermeul
committed
self, entity='role_assignment',
Swen Vermeul
committed
df=roles[attrs],
Swen Vermeul
committed
identifier_name='techId'
Swen Vermeul
committed
def get_role_assignment(self, techId, only_data=False):
""" Fetches one assigned role by its techId.
"""
fetchopts = {}
for option in ['roleAssignments', 'space', 'project', 'user', 'authorizationGroup','registrator']:
fetchopts[option] = fetch_option[option]
request = {
"method": "getRoleAssignments",
"params": [
self.token,
[{
"@type": "as.dto.roleassignment.id.RoleAssignmentTechId"
}],
fetchopts
]
}
resp = self._post_request(self.as_v3, request)
if len(resp) == 0:
raise ValueError("No assigned role found for techId={}".format(techId))
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
for id in resp:
data = resp[id]
parse_jackson(data)
if only_data:
return data
else:
return RoleAssignment(self, data=data)
def assign_role(self, role, **args):
""" general method to assign a role to either
- a person
- a group
The scope is either
- the whole instance
- a space
- a project
"""
userId = None
groupId = None
spaceId = None
projectId = None
for arg in args:
if arg in ['person', 'group', 'space', 'project']:
permId = args[arg] if isinstance(args[arg],str) else args[arg].permId
if arg == 'person':
userId = {
"permId": permId,
"@type": "as.dto.person.id.PersonPermId"
}
elif arg == 'group':
groupId = {
"permId": permId,
"@type": "as.dto.authorizationgroup.id.AuthorizationGroupPermId"
}
elif arg == 'space':
spaceId = {
"permId": permId,
"@type": "as.dto.space.id.SpacePermId"
}
elif arg == 'project':
projectId = {
"permId": permId,
"@type": "as.dto.project.id.ProjectPermId"
}
request = {
"method": "createRoleAssignments",
"params": [
self.token,
[
{
"role": role,
"userId": userId,
"authorizationGroupId": groupId,
"spaceId": spaceId,
"projectId": projectId,
"@type": "as.dto.roleassignment.create.RoleAssignmentCreation",
}
]
]
}
resp = self._post_request(self.as_v3, request)
return
def get_groups(self, **search_args):
""" Get openBIS AuthorizationGroups. Returns a «Things» object.
Usage::
groups = e.get.groups()
groups[0] # select first group
groups['GROUP_NAME'] # select group with this code
for group in groups:
... # a Group object
groups.df # get a DataFrame object of the group list
print(groups) # print a nice ASCII table (eg. in IPython)
groups # HTML table (in a Jupyter notebook)
"""
criteria = []
for search_arg in ['code']:
# unfortunately, there aren't many search possibilities yet...
if search_arg in search_args:
if search_arg == 'code':
criteria.append(_criteria_for_code(search_args[search_arg]))
search_criteria = get_search_type_for_entity('authorizationGroup')
search_criteria['criteria'] = criteria
fetchopts = fetch_option['authorizationGroup']
for option in ['roleAssignments', 'registrator', 'users']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchAuthorizationGroups",
"params": [
self.token,
fetchopts
],
}
resp = self._post_request(self.as_v3, request)
attrs = ['permId', 'code', 'description', 'users', 'registrator', 'registrationDate', 'modificationDate']
if len(resp['objects']) == 0:
groups = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
groups = DataFrame(objects)
groups['permId'] = groups['permId'].map(extract_permid)
groups['registrator'] = groups['registrator'].map(extract_person)
groups['users'] = groups['users'].map(extract_userId)
groups['registrationDate'] = groups['registrationDate'].map(format_timestamp)
groups['modificationDate'] = groups['modificationDate'].map(format_timestamp)
return Things(self, entity='group', df=groups[attrs], identifier_name='permId')
def get_persons(self, **search_args):
""" Get openBIS users
"""
search_criteria = get_search_criteria('person', **search_args)
fetchopts = {}
for option in ['space']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchPersons",
"params": [
self.token,
search_criteria,
fetchopts
],
}
resp = self._post_request(self.as_v3, request)
attrs = ['permId', 'userId', 'firstName', 'lastName', 'email', 'space', 'registrationDate', 'active']
if len(resp['objects']) == 0:
persons = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
persons = DataFrame(resp['objects'])
persons['permId'] = persons['permId'].map(extract_permid)
persons['registrationDate'] = persons['registrationDate'].map(format_timestamp)
persons['space'] = persons['space'].map(extract_nested_permid)
return Things(
self, entity='person', df=persons[attrs], identifier_name='permId'
get_users = get_persons # Alias
def get_person(self, userId, only_data=False):
""" Get a person (user)
"""
ids = [{
"@type": "as.dto.person.id.PersonPermId",
"permId": userId
}]
fetchopts = {}
Swen Vermeul
committed
for option in ['space', 'project']:
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
fetchopts[option] = fetch_option[option]
request = {
"method": "getPersons",
"params": [
self.token,
ids,
fetchopts,
],
}
resp = self._post_request(self.as_v3, request)
if len(resp) == 0:
raise ValueError("No person found!")
for permid in resp:
person = resp[permid]
parse_jackson(person)
if only_data:
return person
else:
return Person(self, data=person)
get_user = get_person # Alias
def get_spaces(self, code=None):
""" Get a list of all available spaces (DataFrame object). To create a sample or a
dataset, you need to specify in which space it should live.
"""
search_criteria = _subcriteria_for_code(code, 'space')
fetchopts = {}
"params": [self.token,
search_criteria,
fetchopts,
}
resp = self._post_request(self.as_v3, request)
attrs = ['code', 'description', 'registrationDate', 'modificationDate']
if len(resp['objects']) == 0:
spaces = DataFrame(columns=attrs)
else:
spaces['registrationDate'] = spaces['registrationDate'].map(format_timestamp)
spaces['modificationDate'] = spaces['modificationDate'].map(format_timestamp)
return Things(self, 'space', spaces[attrs])
def get_space(self, code, only_data=False):
""" Returns a Space object for a given identifier.
code = str(code).upper()
fetchopts = {"@type": "as.dto.space.fetchoptions.SpaceFetchOptions"}
for option in ['registrator']:
fetchopts[option] = fetch_option[option]
"method": "getSpaces",
"params": [
self.token,
[{
"permId": code,
"@type": "as.dto.space.id.SpacePermId"
}],
fetchopts
resp = self._post_request(self.as_v3, request)
raise ValueError("No such space: %s" % code)
for permid in resp:
if only_data:
return resp[permid]
else:
return Space(self, data=resp[permid])
Chandrasekhar Ramakrishnan
committed
Swen Vermeul
committed
def get_samples(self, code=None, permId=None, space=None, project=None, experiment=None, type=None,
Swen Vermeul
committed
withParents=None, withChildren=None, tags=None, props=None, **properties):
""" Get a list of all samples for a given space/project/experiment (or any combination)
"""
sub_criteria = []
if space:
sub_criteria.append(_gen_search_criteria({
"space": "Space",
"operator": "AND",
"code": space
)
exp_crit = _subcriteria_for_code(experiment, 'experiment')
proj_crit = _subcriteria_for_code(project, 'project')
exp_crit['criteria'] = []
exp_crit['criteria'].append(proj_crit)
sub_criteria.append(exp_crit)
sub_criteria.append(_subcriteria_for_code(experiment, 'experiment'))
if properties is not None:
for prop in properties:
sub_criteria.append(_subcriteria_for_properties(prop, properties[prop]))
if type:
sub_criteria.append(_subcriteria_for_code(type, 'sample_type'))
if tags:
sub_criteria.append(_subcriteria_for_tags(tags))
if code:
sub_criteria.append(_criteria_for_code(code))
if permId:
sub_criteria.append(_common_search("as.dto.common.search.PermIdSearchCriteria", permId))
if withParents:
if not isinstance(withParents, list):
withParents = [withParents]
for parent in withParents:
sub_criteria.append(
_gen_search_criteria({
"sample": "SampleParents",
"identifier": parent
})
)
if withChildren:
if not isinstance(withChildren, list):
withChildren = [withChildren]
for child in withChildren:
sub_criteria.append(
_gen_search_criteria({
"sample": "SampleChildren",
"identifier": child
})
)
criteria = {
"criteria": sub_criteria,
"@type": "as.dto.sample.search.SampleSearchCriteria",
"operator": "AND"
}
# build the various fetch options
fetchopts = fetch_option['sample']
for option in ['tags', 'properties', 'registrator', 'modifier', 'experiment']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchSamples",
"params": [self.token,
criteria,
fetchopts,
],
Swen Vermeul
committed
resp = self._post_request(self.as_v3, request)
attrs = ['identifier', 'permId', 'experiment', 'sample_type',
'registrator', 'registrationDate', 'modifier', 'modificationDate']
Swen Vermeul
committed
if len(resp['objects']) == 0:
samples = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
samples = DataFrame(objects)
samples['registrationDate'] = samples['registrationDate'].map(format_timestamp)
samples['modificationDate'] = samples['modificationDate'].map(format_timestamp)
samples['registrator'] = samples['registrator'].map(extract_person)
samples['modifier'] = samples['modifier'].map(extract_person)
samples['identifier'] = samples['identifier'].map(extract_identifier)
samples['permId'] = samples['permId'].map(extract_permid)
samples['experiment'] = samples['experiment'].map(extract_nested_identifier)
samples['sample_type'] = samples['type'].map(extract_nested_permid)
Swen Vermeul
committed
Swen Vermeul
committed
for prop in props:
samples[prop.upper()] = samples['properties'].map(lambda x: x.get(prop.upper(), ''))
Swen Vermeul
committed
attrs.append(prop.upper())
return Things(self, 'sample', samples[attrs], 'identifier')
Swen Vermeul
committed
get_objects = get_samples # Alias
Swen Vermeul
committed
def get_experiments(self, code=None, type=None, space=None, project=None, tags=None, is_finished=None, props=None, **properties):
""" Searches for all experiment which match the search criteria. Returns a
«Things» object which can be used in many different situations.
Usage::
experiments = get_experiments(project='PROJECT_NAME', props=['NAME','FINISHED_FLAG'])
experiments[0] # returns first experiment
experiments['/MATERIALS/REAGENTS/ANTIBODY_COLLECTION']
for experiment in experiment:
# handle every experiment
...
experiments.df # returns DataFrame object of the experiment list
print(experiments) # prints a nice ASCII table
"""
sub_criteria = []
if space:
sub_criteria.append(_subcriteria_for_code(space, 'space'))
sub_criteria.append(_subcriteria_for_code(project, 'project'))
sub_criteria.append(_criteria_for_code(code))
if type:
sub_criteria.append(_subcriteria_for_type(type, 'Experiment'))
if tags:
sub_criteria.append(_subcriteria_for_tags(tags))
if is_finished is not None:
sub_criteria.append(_subcriteria_for_is_finished(is_finished))
if properties is not None:
for prop in properties:
sub_criteria.append(_subcriteria_for_properties(prop, properties[prop]))
search_criteria = get_search_type_for_entity('experiment')
search_criteria['criteria'] = sub_criteria
search_criteria['operator'] = 'AND'
fetchopts = fetch_option['experiment']
for option in ['tags', 'properties', 'registrator', 'modifier', 'project']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchExperiments",
"params": [
self.token,
search_criteria,
fetchopts,
],
}
resp = self._post_request(self.as_v3, request)
attrs = ['identifier', 'permId', 'project', 'type',
Swen Vermeul
committed
'registrator', 'registrationDate', 'modifier', 'modificationDate']
if len(resp['objects']) == 0:
experiments = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
experiments = DataFrame(objects)
experiments['registrationDate'] = experiments['registrationDate'].map(format_timestamp)
experiments['modificationDate'] = experiments['modificationDate'].map(format_timestamp)
experiments['project'] = experiments['project'].map(extract_code)
experiments['registrator'] = experiments['registrator'].map(extract_person)
experiments['modifier'] = experiments['modifier'].map(extract_person)
experiments['identifier'] = experiments['identifier'].map(extract_identifier)
experiments['permId'] = experiments['permId'].map(extract_permid)
experiments['type'] = experiments['type'].map(extract_code)
Swen Vermeul
committed
if props is not None:
for prop in props:
experiments[prop.upper()] = experiments['properties'].map(lambda x: x.get(prop.upper(), ''))
Swen Vermeul
committed
attrs.append(prop.upper())
return Things(self, 'experiment', experiments[attrs], 'identifier')
def get_datasets(self,
code=None, type=None, withParents=None, withChildren=None, status=None,
sample=None, experiment=None, project=None, tags=None, props=None, **properties
sub_criteria = []
if code:
sub_criteria.append(_criteria_for_code(code))
if type:
sub_criteria.append(_subcriteria_for_type(type, 'DataSet'))
if withParents:
sub_criteria.append(_subcriteria_for_permid(withParents, 'DataSet', 'Parents'))
if withChildren:
sub_criteria.append(_subcriteria_for_permid(withChildren, 'DataSet', 'Children'))
if sample:
sub_criteria.append(_subcriteria_for_code(sample, 'Sample'))
if experiment:
sub_criteria.append(_subcriteria_for_code(experiment, 'Experiment'))
if project:
exp_crit = _subcriteria_for_code(experiment, 'Experiment')
proj_crit = _subcriteria_for_code(project, 'Project')
exp_crit['criteria'] = []
exp_crit['criteria'].append(proj_crit)
sub_criteria.append(exp_crit)
if tags:
sub_criteria.append(_subcriteria_for_tags(tags))
if status:
sub_criteria.append(_subcriteria_for_status(status))
if properties is not None:
for prop in properties:
sub_criteria.append(_subcriteria_for_properties(prop, properties[prop]))
Swen Vermeul
committed
search_criteria = get_search_type_for_entity('dataset')
search_criteria['criteria'] = sub_criteria
search_criteria['operator'] = 'AND'
fetchopts = {
"containers": {"@type": "as.dto.dataset.fetchoptions.DataSetFetchOptions"},
"type": {"@type": "as.dto.dataset.fetchoptions.DataSetTypeFetchOptions"}
}
for option in ['tags', 'properties', 'sample', 'experiment', 'physicalData']:
fetchopts[option] = fetch_option[option]
request = {
"method": "searchDataSets",
"params": [self.token,
fetchopts,
],
}
resp = self._post_request(self.as_v3, request)
attrs = ['permId', 'properties', 'type', 'experiment', 'sample', 'registrationDate', 'modificationDate', 'location']
datasets = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
datasets = DataFrame(objects)
datasets['registrationDate'] = datasets['registrationDate'].map(format_timestamp)
datasets['modificationDate'] = datasets['modificationDate'].map(format_timestamp)
datasets['experiment'] = datasets['experiment'].map(extract_nested_identifier)
datasets['sample'] = datasets['sample'].map(extract_nested_identifier)
datasets['type'] = datasets['type'].map(extract_code)
datasets['permId'] = datasets['code']
datasets['location'] = datasets['physicalData'].map(lambda x: x.get('location') if x else '')
if props is not None:
for prop in props:
datasets[prop.upper()] = datasets['properties'].map(lambda x: x.get(prop.upper(), ''))
attrs.append(prop.upper())
return Things(self, 'dataset', datasets[attrs], 'permId')
def get_experiment(self, expId, withAttachments=False, only_data=False):
""" Returns an experiment object for a given identifier (expId).
"""
fetchopts = {
Swen Vermeul
committed
"@type": "as.dto.experiment.fetchoptions.ExperimentFetchOptions",
"type": {
"@type": "as.dto.experiment.fetchoptions.ExperimentTypeFetchOptions",
},
search_request = search_request_for_identifier(expId, 'experiment')
Swen Vermeul
committed
for option in ['tags', 'properties', 'attachments', 'project', 'samples']:
fetchopts[option] = fetch_option[option]
Swen Vermeul
committed
if withAttachments:
fetchopts['attachments'] = fetch_option['attachmentsWithContent']
"method": "getExperiments",
"params": [
self.token,
[search_request],
fetchopts
resp = self._post_request(self.as_v3, request)
if len(resp) == 0:
raise ValueError("No such experiment: %s" % expId)
for id in resp:
if only_data:
return resp[id]
else:
return Experiment(
openbis_obj = self,
type = self.get_experiment_type(resp[expId]["type"]["code"]),
data = resp[id]
)
def new_experiment(self, type, code, project, props=None, **kwargs):
""" Creates a new experiment of a given experiment type.
"""
return Experiment(
openbis_obj = self,
type = self.get_experiment_type(type),
data = None,
props = props,
code = code,
**kwargs
)
def update_experiment(self, experimentId, properties=None, tagIds=None, attachments=None):
params = {
"experimentId": {
"permId": experimentId,
"@type": "as.dto.experiment.id.ExperimentPermId"
},
"@type": "as.dto.experiment.update.ExperimentUpdate"
}
if properties is not None:
params["properties"] = properties
if tagIds is not None:
params["tagIds"] = tagIds
if attachments is not None:
params["attachments"] = attachments
request = {
"method": "updateExperiments",
"params": [
self.token,
[params]
]
}
self._post_request(self.as_v3, request)
def create_sample(self, space_ident, code, type,
project_ident=None, experiment_ident=None, properties=None, attachments=None, tags=None):
tagIds = _create_tagIds(tags)
typeId = _create_typeId(type)
projectId = _create_projectId(project_ident)
experimentId = _create_experimentId(experiment_ident)
if properties is None:
properties = {}
request = {
"method": "createSamples",
"params": [
self.token,
[
{
"properties": properties,
"code": code,
"typeId": typeId,
"projectId": projectId,
"experimentId": experimentId,
"tagIds": tagIds,
"attachments": attachments,
"@type": "as.dto.sample.create.SampleCreation",
}
]
],
}
resp = self._post_request(self.as_v3, request)
return self.get_sample(resp[0]['permId'])
create_object = create_sample # Alias
Chandrasekhar Ramakrishnan
committed
def create_external_data_management_system(self, code, label, address, address_type='FILE_SYSTEM'):
Chandrasekhar Ramakrishnan
committed
"""Create an external DMS.
:param code: An openBIS code for the external DMS.
:param label: A human-readable label.
:param address: The address for accessing the external DMS. E.g., a URL.
Chandrasekhar Ramakrishnan
committed
:param address_type: One of OPENBIS, URL, or FILE_SYSTEM
Chandrasekhar Ramakrishnan
committed
:return:
"""
request = {
"method": "createExternalDataManagementSystems",
"params": [
self.token,
[
{
"code": code,
"label": label,
"addressType": address_type,
"address": address,
"@type": "as.dto.externaldms.create.ExternalDmsCreation",
}
]
],
}
resp = self._post_request(self.as_v3, request)
return self.get_external_data_management_system(resp[0]['permId'])
def update_sample(self, sampleId, space=None, project=None, experiment=None,
parents=None, children=None, components=None, properties=None, tagIds=None, attachments=None):
params = {
"sampleId": {
"permId": sampleId,
"@type": "as.dto.sample.id.SamplePermId"
},
"@type": "as.dto.sample.update.SampleUpdate"
}
if space is not None:
params['spaceId'] = space
if project is not None:
params['projectId'] = project
if properties is not None:
params["properties"] = properties
if tagIds is not None:
params["tagIds"] = tagIds
if attachments is not None:
params["attachments"] = attachments
request = {
"method": "updateSamples",
"params": [
self.token,
[params]
]
}
self._post_request(self.as_v3, request)
update_object = update_sample # Alias
def delete_entity(self, entity, id, reason, id_name='permId'):
"""Deletes Spaces, Projects, Experiments, Samples and DataSets
"""
entity_type = "as.dto.{}.id.{}{}{}".format(
entity.lower(), entity,
id_name[0].upper(), id_name[1:]
)
request = {
"method": "delete{}s".format(entity),
"params": [
self.token,
[
{
id_name: id,
"@type": entity_type
}
],
{
"reason": reason,
"@type": "as.dto.{}.delete.{}DeletionOptions".format(
entity.lower(), entity)
}
]
}
self._post_request(self.as_v3, request)
def get_deletions(self):
request = {
"method": "searchDeletions",
"params": [
self.token,
{},
{
"deletedObjects": {
"@type": "as.dto.deletion.fetchoptions.DeletedObjectFetchOptions"
}
}
]
}
resp = self._post_request(self.as_v3, request)
objects = resp['objects']
parse_jackson(objects)
new_objs = []
for value in objects:
del_objs = extract_deletion(value)
if len(del_objs) > 0:
new_objs.append(*del_objs)
return DataFrame(new_objs)
def new_project(self, space, code, description=None, **kwargs):
return Project(self, None, space=space, code=code, description=description, **kwargs)
def _gen_fetchoptions(self, options):
fo = {}
for option in options:
fo[option] = fetch_option[option]
return fo
def get_project(self, projectId, only_data=False):
options = ['space', 'registrator', 'modifier', 'attachments']
if is_identifier(projectId) or is_permid(projectId):
request = self._create_get_request(
'getProjects', 'project', projectId, options
)
resp = self._post_request(self.as_v3, request)
if only_data:
return resp[projectId]
return Project(self, resp[projectId])
else:
search_criteria = _gen_search_criteria({
'project': 'Project',
'operator': 'AND',
'code': projectId
})
fo = self._gen_fetchoptions(options)
request = {
"method": "searchProjects",
"params": [self.token, search_criteria, fo]
}
resp = self._post_request(self.as_v3, request)
Swen Vermeul
committed
if len(resp['objects']) == 0:
raise ValueError("No such project: %s" % projectId)
if only_data:
return resp['objects'][0]
return Project(self, resp['objects'][0])
def get_projects(self, space=None, code=None):
""" Get a list of all available projects (DataFrame object).
"""
sub_criteria = []
if space:
sub_criteria.append(_subcriteria_for_code(space, 'space'))
if code:
sub_criteria.append(_criteria_for_code(code))
criteria = {
"criteria": sub_criteria,
"@type": "as.dto.project.search.ProjectSearchCriteria",
"operator": "AND"
}
fetchopts = {"@type": "as.dto.project.fetchoptions.ProjectFetchOptions"}
for option in ['registrator', 'modifier', 'leader']:
request = {
"method": "searchProjects",
"params": [self.token,
criteria,
fetchopts,
],
}
resp = self._post_request(self.as_v3, request)
attrs = ['identifier', 'permId', 'leader', 'registrator', 'registrationDate', 'modifier', 'modificationDate']
if len(resp['objects']) == 0:
projects = DataFrame(columns=attrs)
else:
objects = resp['objects']
parse_jackson(objects)
projects = DataFrame(objects)
projects['registrationDate'] = projects['registrationDate'].map(format_timestamp)
projects['modificationDate'] = projects['modificationDate'].map(format_timestamp)
projects['leader'] = projects['leader'].map(extract_person)
projects['registrator'] = projects['registrator'].map(extract_person)
projects['modifier'] = projects['modifier'].map(extract_person)
projects['permId'] = projects['permId'].map(extract_permid)
projects['identifier'] = projects['identifier'].map(extract_identifier)
return Things(self, 'project', projects[attrs], 'identifier')
def _create_get_request(self, method_name, entity, permids, options):
if not isinstance(permids, list):
permids = [permids]
type = "as.dto.{}.id.{}".format(entity.lower(), entity.capitalize())
search_params = []
for permid in permids:
# decide if we got a permId or an identifier
match = re.match('/', permid)
if match:
search_params.append(
{"identifier": permid, "@type": type + 'Identifier'}
else:
{"permId": permid, "@type": type + 'PermId'}
)
fo = {}
for option in options:
fo[option] = fetch_option[option]
request = {
"method": method_name,
"params": [
self.token,
search_params,
fo
],
}
return request
def get_terms(self, vocabulary=None):
""" Returns information about vocabulary, including its controlled vocabulary
search_request = {}
if vocabulary is not None:
search_request = _gen_search_criteria({
"vocabulary": "VocabularyTerm",
"criteria": [{
"vocabulary": "Vocabulary",
"code": vocabulary
}]
})
fetch_options = {
"vocabulary": {"@type": "as.dto.vocabulary.fetchoptions.VocabularyFetchOptions"},
"@type": "as.dto.vocabulary.fetchoptions.VocabularyTermFetchOptions"
}
request = {
"method": "searchVocabularyTerms",
"params": [self.token, search_request, fetch_options]
}
resp = self._post_request(self.as_v3, request)
parse_jackson(resp)
return Vocabulary(resp)
def get_tags(self):
Swen Vermeul
committed
""" Returns a DataFrame of all tags
"""
request = {
"method": "searchTags",
"params": [self.token, {}, {}]
}
resp = self._post_request(self.as_v3, request)
parse_jackson(resp)
objects = DataFrame(resp['objects'])
objects['registrationDate'] = objects['registrationDate'].map(format_timestamp)
return objects[['code', 'registrationDate']]
Swen Vermeul
committed
def get_tag(self, permId):
""" Returns a specific tag
"""
fetchopts = {}
request = {
"method": "getTags",
"params": [
self.token,
[{
"permId": permId,
"@type": "as.dto.tag.id.TagPermId"
}],
fetchopts
],
}
def _search_semantic_annotations(self, criteria):
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
fetch_options = {
"@type": "as.dto.semanticannotation.fetchoptions.SemanticAnnotationFetchOptions",
"entityType": {"@type": "as.dto.entitytype.fetchoptions.EntityTypeFetchOptions"},
"propertyType": {"@type": "as.dto.property.fetchoptions.PropertyTypeFetchOptions"},
"propertyAssignment": {
"@type": "as.dto.property.fetchoptions.PropertyAssignmentFetchOptions",
"entityType" : {
"@type" : "as.dto.entitytype.fetchoptions.EntityTypeFetchOptions"
},
"propertyType" : {
"@type" : "as.dto.property.fetchoptions.PropertyTypeFetchOptions"
}
}
}
request = {
"method": "searchSemanticAnnotations",
"params": [self.token, criteria, fetch_options]
}
resp = self._post_request(self.as_v3, request)
if resp is not None:
objects = resp['objects']
if len(objects) is 0:
raise ValueError("No semantic annotations found!")
parse_jackson(objects)
for object in objects:
object['permId'] = object['permId']['permId']
if object.get('entityType') is not None:
object['entityType'] = object['entityType']['code']
elif object.get('propertyType') is not None:
object['propertyType'] = object['propertyType']['code']
elif object.get('propertyAssignment') is not None:
object['entityType'] = object['propertyAssignment']['entityType']['code']
object['propertyType'] = object['propertyAssignment']['propertyType']['code']
object['creationDate'] = format_timestamp(object['creationDate'])
return objects
else:
raise ValueError("No semantic annotations found!")
def get_semantic_annotations(self):
""" Get a list of all available semantic annotations (DataFrame object).
"""
objects = self._search_semantic_annotations({})
attrs = ['permId', 'entityType', 'propertyType', 'predicateOntologyId', 'predicateOntologyVersion', 'predicateAccessionId', 'descriptorOntologyId', 'descriptorOntologyVersion', 'descriptorAccessionId', 'creationDate']
annotations = DataFrame(objects)
return Things(self, 'semantic_annotation', annotations[attrs], 'permId')
def get_semantic_annotation(self, permId, only_data = False):
criteria = {
"@type" : "as.dto.semanticannotation.search.SemanticAnnotationSearchCriteria",
"criteria" : [{
"@type" : "as.dto.common.search.PermIdSearchCriteria",
"fieldValue" : {
"@type" : "as.dto.common.search.StringEqualToValue",
"value" : permId
}
}]
}
objects = self._search_semantic_annotations(criteria)
object = objects[0]
if only_data:
return object
else:
return SemanticAnnotation(self, isNew=False, **object)
Swen Vermeul
committed
def get_sample_types(self, type=None):
""" Returns a list of all available sample types
"""
return self._get_types_of(
"searchSampleTypes",
"Sample",
type,
Swen Vermeul
committed
["generatedCodePrefix"]
)
get_object_types = get_sample_types # Alias
Swen Vermeul
committed
def get_sample_type(self, type):
try:
return self._get_types_of(
"searchSampleTypes",
Swen Vermeul
committed
"Sample",
type,
["generatedCodePrefix"]
)
except Exception:
raise ValueError("no such sample type: {}".format(type))
get_object_type = get_sample_type # Alias
Swen Vermeul
committed
def get_experiment_types(self, type=None):
""" Returns a list of all available experiment types
"""
return self._get_types_of(
"searchExperimentTypes",
"Experiment",
Swen Vermeul
committed
type
)
def get_experiment_type(self, type):
Swen Vermeul
committed
return self._get_types_of(
"searchExperimentTypes",
"Experiment",
Swen Vermeul
committed
type
)
except Exception:
raise ValueError("No such experiment type: {}".format(type))
def get_material_types(self, type=None):
""" Returns a list of all available material types
"""
return self._get_types_of("searchMaterialTypes", "Material", type)
def get_material_type(self, type):
try:
return self._get_types_of("searchMaterialTypes", "Material", type)
except Exception:
raise ValueError("No such material type: {}".format(type))
def get_dataset_types(self, type=None):
""" Returns a list (DataFrame object) of all currently available dataset types
"""
Yves Noirjean
committed
return self._get_types_of("searchDataSetTypes", "DataSet", type, optional_attributes=['kind'])
Swen Vermeul
committed
def get_dataset_type(self, type):
try:
Yves Noirjean
committed
return self._get_types_of("searchDataSetTypes", "DataSet", type, optional_attributes=['kind'])
Swen Vermeul
committed
except Exception:
raise ValueError("No such dataSet type: {}".format(type))
Yves Noirjean
committed
def _get_types_of(self, method_name, entity, type_name=None, additional_attributes=[], optional_attributes=[]):
""" Returns a list of all available types of an entity.
If the name of the entity-type is given, it returns a PropertyAssignments object
search_request = {}
fetch_options = {}
if type_name is not None:
search_request = _gen_search_criteria({
"operator": "AND",
})
fetch_options = {
Swen Vermeul
committed
"@type": "as.dto.{}.fetchoptions.{}TypeFetchOptions".format(
Swen Vermeul
committed
)
}
Swen Vermeul
committed
fetch_options['propertyAssignments'] = fetch_option['propertyAssignments']
"params": [self.token, search_request, fetch_options],
}
resp = self._post_request(self.as_v3, request)
parse_jackson(resp)
if type_name is not None and len(resp['objects']) == 1:
return PropertyAssignments(self, resp['objects'][0])
if len(resp['objects']) >= 1:
types = DataFrame(resp['objects'])
Swen Vermeul
committed
types['modificationDate'] = types['modificationDate'].map(format_timestamp)
Yves Noirjean
committed
attributes = self._get_attributes(type_name, types, additional_attributes, optional_attributes)
return Things(self, entity.lower() + '_type', types[attributes])
Yves Noirjean
committed
def _get_attributes(self, type_name, types, additional_attributes, optional_attributes):
attributes = ['code', 'description'] + additional_attributes
attributes += [attribute for attribute in optional_attributes if attribute in types]
attributes += ['modificationDate']
if type_name is not None:
attributes += ['propertyAssignments']
return attributes
Swen Vermeul
committed
def is_session_active(self):
""" checks whether a session is still active. Returns true or false.
"""
Swen Vermeul
committed
return self.is_token_valid(self.token)
def is_token_valid(self, token=None):
Chandrasekhar Ramakrishnan
committed
"""Check if the connection to openBIS is valid.
This method is useful to check if a token is still valid or if it has timed out,
requiring the user to login again.
Chandrasekhar Ramakrishnan
committed
:return: Return True if the token is valid, False if it is not valid.
"""
if token is None:
token = self.token
if token is None:
return False
request = {
"method": "isSessionActive",
"params": [token],
resp = self._post_request(self.as_v1, request)
def get_dataset(self, permid, only_data=False):
"""fetch a dataset and some metadata attached to it:
- properties
- sample
- parents
- children
- containers
- dataStore
- physicalData
- linkedData
:return: a DataSet object
"""
criteria = [{
"permId": permid,
"@type": "as.dto.dataset.id.DataSetPermId"
}]
fetchopts = {
"parents": {"@type": "as.dto.dataset.fetchoptions.DataSetFetchOptions"},
"children": {"@type": "as.dto.dataset.fetchoptions.DataSetFetchOptions"},
"containers": {"@type": "as.dto.dataset.fetchoptions.DataSetFetchOptions"},
"type": {"@type": "as.dto.dataset.fetchoptions.DataSetTypeFetchOptions"},
}
for option in ['tags', 'properties', 'dataStore', 'physicalData', 'linkedData',
'experiment', 'sample']:
fetchopts[option] = fetch_option[option]
request = {
"params": [
self.token,
criteria,
fetchopts,
],
resp = self._post_request(self.as_v3, request)
raise ValueError('no such dataset found: ' + permid)
parse_jackson(resp)
for permid in resp:
if only_data:
return resp[permid]
else:
return DataSet(
self,
type=self.get_dataset_type(resp[permid]["type"]["code"]),
data=resp[permid]
)
Swen Vermeul
committed
def get_sample(self, sample_ident, only_data=False, withAttachments=False):
Chandrasekhar Ramakrishnan
committed
"""Retrieve metadata for the sample.
Get metadata for the sample and any directly connected parents of the sample to allow access
to the same information visible in the ELN UI. The metadata will be on the file system.
:param sample_identifiers: A list of sample identifiers to retrieve.
"""
search_request = search_request_for_identifier(sample_ident, 'sample')
Swen Vermeul
committed
fetchopts = {"type": {"@type": "as.dto.sample.fetchoptions.SampleTypeFetchOptions"}}
Swen Vermeul
committed
for option in ['tags', 'properties', 'attachments', 'space', 'experiment', 'registrator', 'dataSets']:
fetchopts[option] = fetch_option[option]
if withAttachments:
fetchopts['attachments'] = fetch_option['attachmentsWithContent']
for key in ['parents','children','container','components']:
fetchopts[key] = {"@type": "as.dto.sample.fetchoptions.SampleFetchOptions"}
sample_request = {
"method": "getSamples",
"params": [
self.token,
[search_request],
Swen Vermeul
committed
fetchopts
resp = self._post_request(self.as_v3, sample_request)
parse_jackson(resp)
raise ValueError('no such sample found: ' + sample_ident)
for sample_ident in resp:
if only_data:
return resp[sample_ident]
else:
return Sample(self, self.get_sample_type(resp[sample_ident]["type"]["code"]), resp[sample_ident])
get_object = get_sample # Alias
def get_external_data_management_system(self, permId, only_data=False):
Chandrasekhar Ramakrishnan
committed
"""Retrieve metadata for the external data management system.
:param permId: A permId for an external DMS.
Chandrasekhar Ramakrishnan
committed
:param only_data: Return the result data as a hash-map, not an object.
"""
request = {
"method": "getExternalDataManagementSystems",
"params": [
self.token,
[{
"@type": "as.dto.externaldms.id.ExternalDmsPermId",
"permId": permId
Chandrasekhar Ramakrishnan
committed
}],
{},
],
}
resp = self._post_request(self.as_v3, request)
parse_jackson(resp)
if resp is None or len(resp) == 0:
raise ValueError('no such external DMS found: ' + permId)
Chandrasekhar Ramakrishnan
committed
else:
for ident in resp:
if only_data:
return resp[ident]
else:
return ExternalDMS(self, resp[ident])
def new_space(self, **kwargs):
return Space(self, None, **kwargs)
def new_git_data_set(self, data_set_type, path, commit_id, repository_id, dms, sample=None, experiment=None, properties={},
dss_code=None, parents=None, data_set_code=None, contents=[]):
""" Create a link data set.
:param data_set_type: The type of the data set
Chandrasekhar Ramakrishnan
committed
:param data_set_type: The type of the data set
:param path: The path to the git repository
:param commit_id: The git commit id
:param repository_id: The git repository id - same for copies
:param dms: An external data managment system object or external_dms_id
:param sample: A sample object or sample id.
Chandrasekhar Ramakrishnan
committed
:param dss_code: Code for the DSS -- defaults to the first dss if none is supplied.
:param properties: Properties for the data set.
:param parents: Parents for the data set.
Chandrasekhar Ramakrishnan
committed
:param data_set_code: A data set code -- used if provided, otherwise generated on the server
:param contents: A list of dicts that describe the contents:
{'file_length': [file length],
'crc32': [crc32 checksum],
'directory': [is path a directory?]
'path': [the relative path string]}
:return: A DataSet object
"""
return pbds.GitDataSetCreation(self, data_set_type, path, commit_id, repository_id, dms, sample, experiment,
properties, dss_code, parents, data_set_code, contents).new_git_data_set()
def new_content_copy(self, path, commit_id, repository_id, edms_id, data_set_id):
"""
Create a content copy in an existing link data set.
:param path: path of the new content copy
"param commit_id: commit id of the new content copy
"param repository_id: repository id of the content copy
"param edms_id: Id of the external data managment system of the content copy
"param data_set_id: Id of the data set to which the new content copy belongs
"""
return pbds.GitDataSetUpdate(self, path, commit_id, repository_id, edms_id, data_set_id).new_content_copy()
Chandrasekhar Ramakrishnan
committed
@staticmethod
def sample_to_sample_id(sample):
"""Take sample which may be a string or object and return an identifier for it."""
return Openbis._object_to_object_id(sample, "as.dto.sample.id.SampleIdentifier", "as.dto.sample.id.SamplePermId");
@staticmethod
def experiment_to_experiment_id(experiment):
"""Take experiment which may be a string or object and return an identifier for it."""
return Openbis._object_to_object_id(experiment, "as.dto.experiment.id.ExperimentIdentifier", "as.dto.experiment.id.SamplePermId");
@staticmethod
def _object_to_object_id(obj, identifierType, permIdType):
object_id = None
if isinstance(obj, str):
if (is_identifier(obj)):
object_id = {
"identifier": obj,
"@type": identifierType
}
else:
object_id = {
"permId": obj,
"@type": permIdType
}
else:
object_id = {
"identifier": obj.identifier,
"@type": identifierType
return object_id
Chandrasekhar Ramakrishnan
committed
@staticmethod
def data_set_to_data_set_id(data_set):
if isinstance(data_set, str):
code = data_set
else:
code = data_set.permId
return {
"permId": code,
"@type": "as.dto.dataset.id.DataSetPermId"
}
def external_data_managment_system_to_dms_id(self, dms):
if isinstance(dms, str):
dms_id = {
"permId": dms,
"@type": "as.dto.externaldms.id.ExternalDmsPermId"
}
else:
dms_id = {
"identifier": dms.code,
"@type": "as.dto.sample.id.SampleIdentifier"
}
return dms_id
Swen Vermeul
committed
def new_sample(self, type, props=None, **kwargs):
""" Creates a new sample of a given sample type.
Swen Vermeul
committed
return Sample(self, self.get_sample_type(type), None, props, **kwargs)
new_object = new_sample # Alias
def new_dataset(self, type=None, files=None, props=None, folder=None, **kwargs):
""" Creates a new dataset of a given sample type.
"""
if files is None:
raise ValueError('please provide at least one file')
elif isinstance(files, str):
files = [files]
type_obj = self.get_dataset_type(type.upper())
return DataSet(self, type=type_obj, files=files, folder=folder, props=props, **kwargs)
def new_semantic_annotation(self, entityType=None, propertyType=None, **kwargs):
return SemanticAnnotation(
openbis_obj=self, isNew=True,
entityType=entityType, propertyType=propertyType, **kwargs
)
def _get_dss_url(self, dss_code=None):
""" internal method to get the downloadURL of a datastore.
"""
Swen Vermeul
committed
dss = self.get_datastores()
if dss_code is None:
return dss['downloadUrl'][0]
else:
Swen Vermeul
committed
return dss[dss['code'] == dss_code]['downloadUrl'][0]
class LinkedData():
def __init__(self, data=None):
self.data = data if data is not None else []
self.attrs = ['externalCode', 'contentCopies']
def __dir__(self):
return self.attrs
def __getattr__(self, name):
if name in self.attrs:
if name in self.data:
return self.data[name]
else:
return ''
Swen Vermeul
committed
class PhysicalData():
def __init__(self, data=None):
if data is None:
data = []
self.data = data
self.attrs = ['speedHint', 'complete', 'shareId', 'size',
'fileFormatType', 'storageFormat', 'location', 'presentInArchive',
'storageConfirmation', 'locatorType', 'status']
Swen Vermeul
committed
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
def __dir__(self):
return self.attrs
def __getattr__(self, name):
if name in self.attrs:
if name in self.data:
return self.data[name]
else:
return ''
def _repr_html_(self):
html = """
<table border="1" class="dataframe">
<thead>
<tr style="text-align: right;">
<th>attribute</th>
<th>value</th>
</tr>
</thead>
<tbody>
"""
for attr in self.attrs:
html += "<tr> <td>{}</td> <td>{}</td> </tr>".format(
attr, getattr(self, attr, '')
Swen Vermeul
committed
)
html += """
</tbody>
</table>
"""
return html
def __repr__(self):
headers = ['attribute', 'value']
lines = []
for attr in self.attrs:
lines.append([
attr,
getattr(self, attr, '')
])
return tabulate(lines, headers=headers)
Swen Vermeul
committed
class ExternalDMS():
""" managing openBIS external data management systems
Swen Vermeul
committed
def __init__(self, openbis_obj, data=None, **kwargs):
self.__dict__['openbis'] = openbis_obj
Swen Vermeul
committed
Swen Vermeul
committed
self.__dict__['data'] = data
Swen Vermeul
committed
if kwargs is not None:
for key in kwargs:
setattr(self, key, kwargs[key])
Swen Vermeul
committed
def __getattr__(self, name):
return self.__dict__['data'].get(name)
Swen Vermeul
committed
"""all the available methods and attributes that should be displayed
when using the autocompletion feature (TAB) in Jupyter
"""
return ['code', 'label', 'urlTemplate', 'address', 'addressType', 'openbis']
Swen Vermeul
committed
def __str__(self):
return self.data.get('code', None)
Swen Vermeul
committed
class Things():
"""An object that contains a DataFrame object about an entity available in openBIS.
"""
Swen Vermeul
committed
def __init__(self, openbis_obj, entity, df, identifier_name='code'):
self.openbis = openbis_obj
self.entity = entity
self.df = df
self.identifier_name = identifier_name
Swen Vermeul
committed
def __repr__(self):
return tabulate(self.df, headers=list(self.df))
Swen Vermeul
committed
def __len__(self):
return len(self.df)
Swen Vermeul
committed
Swen Vermeul
committed
def _repr_html_(self):
return self.df._repr_html_()
Swen Vermeul
committed
def get_parents(self, **kwargs):
if self.entity not in ['sample', 'dataset']:
raise ValueError("{}s do not have parents".format(self.entity))
Swen Vermeul
committed
if self.df is not None and len(self.df) > 0:
dfs = []
for ident in self.df[self.identifier_name]:
# get all objects that have this object as a child == parent
try:
parents = getattr(self.openbis, 'get_' + self.entity.lower() + 's')(withChildren=ident, **kwargs)
dfs.append(parents.df)
except ValueError:
pass
Swen Vermeul
committed
if len(dfs) > 0:
return Things(self.openbis, self.entity, pd.concat(dfs), self.identifier_name)
else:
return Things(self.openbis, self.entity, DataFrame(), self.identifier_name)
Swen Vermeul
committed
def get_children(self, **kwargs):
if self.entity not in ['sample', 'dataset']:
raise ValueError("{}s do not have children".format(self.entity))
Swen Vermeul
committed
if self.df is not None and len(self.df) > 0:
dfs = []
for ident in self.df[self.identifier_name]:
# get all objects that have this object as a child == parent
try:
parents = getattr(self.openbis, 'get_' + self.entity.lower() + 's')(withParent=ident, **kwargs)
dfs.append(parents.df)
except ValueError:
pass
Swen Vermeul
committed
Swen Vermeul
committed
if len(dfs) > 0:
return Things(self.openbis, self.entity, pd.concat(dfs), self.identifier_name)
else:
return Things(self.openbis, self.entity, DataFrame(), self.identifier_name)
Swen Vermeul
committed
Swen Vermeul
committed
if self.entity not in ['space', 'project', 'experiment']:
raise ValueError("{}s do not have samples".format(self.entity))
if self.df is not None and len(self.df) > 0:
dfs = []
for ident in self.df[self.identifier_name]:
args = {}
args[self.entity.lower()] = ident
try:
samples = self.openbis.get_samples(**args, **kwargs)
Swen Vermeul
committed
dfs.append(samples.df)
except ValueError:
pass
if len(dfs) > 0:
return Things(self.openbis, 'sample', pd.concat(dfs), 'identifier')
else:
return Things(self.openbis, 'sample', DataFrame(), 'identifier')
get_objects = get_samples # Alias
if self.entity not in ['sample', 'experiment']:
Swen Vermeul
committed
raise ValueError("{}s do not have datasets".format(self.entity))
if self.df is not None and len(self.df) > 0:
dfs = []
for ident in self.df[self.identifier_name]:
args = {}
args[self.entity.lower()] = ident
try:
datasets = self.openbis.get_datasets(**args, **kwargs)
Swen Vermeul
committed
dfs.append(datasets.df)
except ValueError:
pass
if len(dfs) > 0:
return Things(self.openbis, 'dataset', pd.concat(dfs), 'permId')
else:
return Things(self.openbis, 'dataset', DataFrame(), 'permId')
def __getitem__(self, key):
if self.df is not None and len(self.df) > 0:
row = None
if isinstance(key, int):
# get thing by rowid
row = self.df.loc[[key]]
elif isinstance(key, list):
# treat it as a normal dataframe
return self.df[key]
row = self.df[self.df[self.identifier_name] == key.upper()]
# invoke the openbis.get_<entity>() method
return getattr(self.openbis, 'get_' + self.entity)(row[self.identifier_name].values[0])
Swen Vermeul
committed
def __iter__(self):
for item in self.df[[self.identifier_name]][self.identifier_name].iteritems():
yield getattr(self.openbis, 'get_' + self.entity)(item[1])
Swen Vermeul
committed
# return self.df[[self.identifier_name]].to_dict()[self.identifier_name]