Small tool for RPQ generation

Author: gsvgit

tools/gen_RPQ/RDF_edge_stat.py

@@ -0,0 +1,29 @@
+import rdflib, sys
+
+def get_labels_count (rdf):
+    d = {}
+    for s,p,o in rdf:     
+       d[p] = d.get(p,0) + 1
+    
+    sorted_d = [(k, v) for k, v in sorted(d.items(), key=lambda item: item[1])] 	
+    sorted_d.reverse()
+
+    return sorted_d
+
+def print_config(lst, path_to_config):
+    i = 0
+    with open(path_to_config,'w') as config:
+       for x in lst:
+           config.write(x[0] + ' ' + str(i) + '\n')
+           i = i + 1
+
+g=rdflib.Graph()
+
+g.load(sys.argv[1])
+
+r = get_labels_count (g)
+
+for x in r: 
+   print(x[0], ': ', x[1])
+
+print_config(r, sys.argv[2])

tools/gen_RPQ/README.md

@@ -0,0 +1,3 @@
+ First step is edges investigation
+ * Run  ```python3 RDF_edge_stat.py my_rdf.rdf my_config``` . Mapping from URIs to integers will be saved to my_config. You can use tid config for RDf to triples convertion using RDF_to_triple tool. Numbers of edges labelled with URIs will be printed in concole. 
+ * Run ```python3 gen.py my_config n_URIs q_for_each_tpl```, where ```my_config``` is a config generated at the previous step, first ```n_URIs``` labels from config will be used to generate queryes, ```q_for_each_tpl``` queryes will be generated for each template.   

tools/gen_RPQ/gen.py

@@ -0,0 +1,24 @@
+import numpy,sys 
+
+templates = [(1,'%s*'), (2,'%s %s*'), (3,'%s %s* %s*'), (2,'(%s | %s)*'), (3,'(%s | %s | %s)*'), (4,'(%s | %s | %s | %s)*'), 
+             (5,'(%s | %s | %s | %s | %s)*'), (3,'%s %s* %s'), (2,'%s* %s*'), (3,'%s %s %s*'), (2,'%s? %s*'),
+             (2,'(%s | %s)+'), (3,'(%s | %s | %s)+'), (4,'(%s | %s | %s | %s)+'), (5,'(%s | %s | %s | %s | %s)+'),
+             (3,'(%s | %s) %s*'), (4,'(%s | %s | %s) %s*'), (5,'(%s | %s | %s | %s) %s*'), (6,'(%s | %s | %s | %s | %s) %s*'),
+             (2,'%s %s'), (3,'%s %s %s'), (4,'%s %s %s %s'), (5,'%s %s %s %s %s')]
+
+def gen (tpl, n, lst, k):
+    res = set()
+    while (len(res) < n):
+        perm = numpy.random.permutation(lst)
+        res.add(tpl % tuple(perm[0:k]))
+    return res
+
+def gen_from_config(config, num_of_lalbels, num_of_queries):
+    lbls = [ l.split(' ')[1].rstrip() for l in open(config,'r').readlines()]
+    return [gen (tpl[1], num_of_queries, lbls[0:num_of_lalbels], tpl[0]) for tpl in templates]
+
+r = gen_from_config(sys.argv[1],int(sys.argv[2]),int(sys.argv[3]))
+
+for s in r:
+	for q in s:
+		print(q)