D-Net

from elasticsearch import *

from elasticsearch_dsl import *

from os import path

import os

from elasticsearch_dsl.response import Response

def get_property():

    f = open(path.join(os.path.dirname(os.path.realpath(__file__)), '../../../../api.properties'))

    p = {}

    for line in f:

        if not line.startswith("#"):

            data = line.strip().split("=")

            p[data[0].strip()] = data[1].strip()

    return p

class DLIESConnector(object):

    def __init__(self):

        props = get_property()

        index_name = props['api.index']

        index_host = [x for x in props['es_index'].split(',')]

        self.client = Elasticsearch(hosts=index_host)

        self.index_name = index_name

    def create_pidType_query(self, value, start):

        args = {start + '.identifier.schema': value}

        return Q('nested', path=start + '.identifier', query=Q('bool', must=[Q('match', **args)]))

    def create_pid_query(self, value, start):

        args = {start + '.identifier.identifier': value.lower()}

        return Q('nested', path=start + '.identifier', query=Q('bool', must=[Q('match', **args)]))

    def create_typology_query(self, value):

        args = {'target.objectType': value}

        return Q('nested', path='target', query=Q('bool', must=[Q('match', **args)]))

    def create_dataSource_query(self, value):

        args = {'linkprovider.name': value}

        return Q('nested', path='linkprovider', query=Q('bool', must=[Q('match', **args)]))

    def create_publisher_query(self, value, start):

        args = {start + '.publisher.name': value}

        q = Q('nested', path=start + '.publisher', query=Q('bool', must=[Q('match', **args)]))

        return Q('nested', path=start, query=q)

    def list_datasources(self, ds_name=None):

        search_object = Search(using=self.client, index=self.index_name).doc_type('scholix')

        if ds_name:

            search_object = search_object.query(self.create_dataSource_query(ds_name))

        else:

            search_object = search_object.query()

        search_object.aggs.bucket('all_datasources', 'nested', path='linkprovider').bucket('all_names', 'terms',

                                                                                           field='linkprovider.name',

                                                                                           size=100)

        response = search_object.execute()

        if ds_name:

            return [dict(name=item.key, totalRelationships=item.doc_count) for item in

                    response.aggs.all_datasources.all_names.buckets if item.key == ds_name]

        return [dict(name=item.key, totalRelationships=item.doc_count) for item in

                response.aggs.all_datasources.all_names.buckets]

    def list_publisher(self, start, pub_name=None):

        search_object = Search(using=self.client, index=self.index_name).doc_type('scholix')

        if pub_name:

            search_object = search_object.query(self.create_publisher_query(pub_name, start))

        else:

            search_object = search_object.query()

        search_object.aggs.bucket('all_targets', 'nested', path=start).bucket('all_t_pubs', 'nested',

                                                                              path=start + '.publisher').bucket(

            'all_pubs', 'terms',

            field=start + '.publisher.name',

            size=1000000)

        response = search_object.execute()

        if pub_name:

            return [dict(name=item.key, totalRelationships=item.doc_count) for item in

                    response.aggs.all_targets.all_t_pubs.all_pubs.buckets if item.key == pub_name]

        return [dict(name=item.key, totalRelationships=item.doc_count) for item in

                response.aggs.all_targets.all_t_pubs.all_pubs.buckets]

    def links(self, provider=None, s_pid=None, t_pid=None, s_publisher=None, t_publisher=None, s_pid_type=None,

              t_pid_type=None, target_Type=None, page=0):

        queries = []

        if provider:

            queries.append(self.create_dataSource_query(provider))

        if s_pid:

            queries.append(self.create_pid_query(s_pid, 'source'))

        if t_pid:

            queries.append(self.create_pid_query(t_pid, 'target'))

        if s_publisher:

            queries.append(self.create_publisher_query(s_publisher, 'source'))

        if t_publisher:

            queries.append(self.create_publisher_query(t_publisher, 'target'))

        if s_pid_type:

            queries.append(self.create_pidType_query(s_pid_type, 'source'))

        if t_pid_type:

            queries.append(self.create_pidType_query(s_pid_type, 'target'))

        if target_Type:

            queries.append(self.create_typology_query(target_Type))

        q = None

        for item in queries:

            if not q:

                q = item

            else:

                q = q & item

        search_object = Search(using=self.client, index=self.index_name).doc_type('scholix').query(q)

        return search_object[page:page + 100].execute()

    def realtionToPid(self, pid, pidType=None, datasource=None, typology=None, page=0):

        if pidType:

            query = self.create_pid_pidType_query(pidType.lower(), pid.lower())

        else:

            query = self.create_source_pid_query(pid.lower())

        filters = []

        if datasource and len(datasource):

            filters.append(self.create_dataSource_query(datasource))

        if typology and len(typology):

            filters.append(self.create_typology_query(typology))

        search_object = Search(using=self.client, index=self.index_name).doc_type('scholix').query(query)

        if len(filters):

            search_object = search_object.filter(Q('bool', must=filters))

        return search_object[page:page + 100].execute()

    def realtionToTypology(self, typology, page=0):

        search_object = Search(using=self.client, index=self.index_name).doc_type('scholix').query(

            self.create_typology_query(typology))

        return search_object[page:page + 100].execute()

dli_index = DLIESConnector()