merge verarbeitung to main repo

Merge branch 'main' of https://gitlab.rrz.uni-hamburg.de/bav1758/ci-s-projekt-verarbeitung into main

verarbeitung/construct_new_graph/add_citations_rec.py

+# -*- coding: utf-8 -*-
+"""
+Functions to add citations recursivly for multiple ACS/Nature journals
+
+"""
+
+__authors__ = "Donna Löding, Alina Molkentin, Xinyi Tang, Judith Große, Malte Schokolowski"
+__email__ = "cis-project2021@zbh.uni-hamburg.de"
+__status__ = "Production"
+#__copyright__ = ""
+#__credits__ = ["", "", "", ""]
+#__license__ = ""
+#__version__ = ""
+#__maintainer__ = ""
+
+
+import sys  
+from pathlib import Path
+from os import error
+sys.path.append("../")
+
+from input.publication import Publication
+from verarbeitung.get_pub_from_input import get_pub
+
+def get_cit_type_list(pub, cit_type):
+    '''
+        :param pub:                 Publication which citations will be added
+        :type pub:                  Publication
+
+        :param cit_type:            variable to differenciate citation and reference call
+        :type cit_type:             String
+
+        function to create nodes and edges and call create_graph_structure_citations
+    '''
+    if (cit_type == "Citation"):
+        return(pub.citations)
+    else:
+        return(pub.references)
+
+def create_global_lists_cit(input_nodes, input_edges, pub, search_depth, search_depth_max, cit_type, test_var):
+    '''
+        :param input_nodes:         list of nodes from Processing
+        :type input_nodes:          List[Publication]
+
+        :param input_edges:         list of edges from Processing
+        :type input_edges:          List[String, String]
+
+        :param pub:                 Publication which citations will be added
+        :type pub:                  Publication
+
+        :param search_depth:        current depth to search for citations
+        :type search_depth_max:     int
+
+        :param search_depth_max:    maximum depth to search for citations
+        :type search_depth_max:     int
+
+        :param cit_type:            variable to differenciate citation and reference call
+        :type cit_type:             String
+
+        :param test_var:            variable to differenciate between test and url call
+        :type test_var:             boolean
+
+        function to create nodes and edges and call create_graph_structure_citations
+    '''
+
+    global nodes, edges
+    nodes = input_nodes
+    edges = input_edges
+
+    return create_graph_structure_citations(pub, search_depth, search_depth_max, cit_type, test_var)
+
+
+def create_graph_structure_citations(pub, search_depth, search_depth_max, cit_type, test_var):
+    '''
+        :param pub:                 publication which citations will be added
+        :type pub:                  Publication
+
+        :param search_depth:        current depth to search for citations
+        :type search_depth_max:     int
+
+        :param search_depth_max:    maximum depth to search for citations
+        :type search_depth_max:     int
+
+        :param cit_type:            variable to differenciate citation and reference call
+        :type cit_type:             String
+
+        :param test_var:            variable to differenciate between test and url call
+        :type test_var:             boolean
+
+        adds a node for every citing publication unknown
+        adds edges to added citations
+    '''
+
+    citations_pub_obj_list = []
+    for citation in get_cit_type_list(pub, cit_type):
+        not_in_nodes = True
+        for node in nodes: # checks every citation for duplication 
+            if (citation.doi_url == node.doi_url):
+                not_in_nodes = False
+                break
+        if (not_in_nodes):
+            if (search_depth < search_depth_max): #checks if its a test and chooses input function accordingly
+                citation_pub_obj = get_pub(citation.doi_url, test_var)
+                if (type(citation_pub_obj) != Publication):
+                    print(pub)
+                    continue 
+                
+                if (cit_type == "Citation"):
+                    citation_pub_obj.group = search_depth + 1
+                    edges.append([citation_pub_obj.doi_url,pub.doi_url])
+                else:
+                    citation_pub_obj.group = -(search_depth + 1)
+                    edges.append([pub.doi_url,citation_pub_obj.doi_url])
+                nodes.append(citation_pub_obj)                   
+                citations_pub_obj_list.append(citation_pub_obj)
+
+        # adds just the edge if citation already exists   
+        else:
+            if (cit_type == "Citation"):      
+                if ([citation.doi_url,pub.doi_url] not in edges):
+                    edges.append([citation.doi_url,pub.doi_url])
+            else:
+                if ([pub.doi_url,citation.doi_url] not in edges):
+                    edges.append([pub.doi_url,citation.doi_url])   
+    return citations_pub_obj_list
+
+
+def process_citations_rec(citations_pub_obj_list, search_depth, search_depth_max, cit_type, test_var):  
+    '''
+        :param citations_pub_obj_list:  list of publications which citations will be added
+        :type citations_pub_obj_list:   List[Publication]
+
+        :param search_depth:        current depth to search for citations
+        :type search_depth_max:     int
+
+        :param search_depth_max:    maximum depth to search for citations
+        :type search_depth_max:     int
+
+        :param cit_type:            variable to differenciate citation and reference call
+        :type cit_type:             String
+
+        :param test_var:            variable to differenciate between test and url call
+        :type test_var:             boolean
+
+        recursive function to implement depth-first-search on citations
+    '''
+
+    # adds next level to nodes/edges
+    for pub in citations_pub_obj_list:
+        new_citation_pub_obj_list = create_graph_structure_citations(pub, search_depth, search_depth_max, cit_type, test_var)   
+
+        # If the maximum depth has not yet been reached, calls function recursivly with increased depth 
+        if (search_depth < search_depth_max):
+            process_citations_rec(new_citation_pub_obj_list, search_depth+1, search_depth_max, cit_type, test_var)
+
+
+def add_citations(input_nodes, input_edges, citations_pub_obj_list, search_depth, search_depth_max, cit_type, test_var):
+    '''
+        :param input_nodes:             list of nodes from Processing
+        :type input_nodes:              List[Publication]
+
+        :param input_edges:             list of edges from Processing
+        :type input_edges:              List[String, String]
+
+        :param citations_pub_obj_list:  list of publications which citations will be added
+        :type citations_pub_obj_list:   List[Publication]
+
+        :param search_depth:        current depth to search for citations
+        :type search_depth_max:     int
+
+        :param search_depth_max:    maximum depth to search for citations
+        :type search_depth_max:     int
+
+        :param cit_type:            variable to differenciate citation and reference call
+        :type cit_type:             String
+
+        :param test_var:            variable to differenciate between test and url call
+        :type test_var:             boolean
+
+        function to call recursive depth-first-search of citations
+    '''
+    global nodes, edges
+    nodes = input_nodes
+    edges = input_edges
+
+    process_citations_rec(citations_pub_obj_list, search_depth, search_depth_max, cit_type, test_var)
+    #return(nodes, edges)
\ No newline at end of file