def find_substructures(expressions_data):
sub_groups = []
query_expression, candidates_data = expressions_data
if len(candidates_data) > 1:
query = Query("query", query_expression)
# create query tree ....
rank = -1
scores = [-1.0, 0, 0]
for data_idx, candidate_data in enumerate(candidates_data):
candidate_exp = candidate_data[0]
rank = int(candidate_data[1])
query.add_result(0, "", 0, candidate_exp, 0.0)
result = query.results[candidate_exp]
candidate_tree = result.tree
try:
scores, matched_q, matched_c, unified_c = similarity_v04(query.tree, candidate_tree, query.constraints)
except:
print("Error processing: ")
print("Q: " + query_expression, flush=True)
print("C: " + candidate_exp, flush=True)
continue
result.set_unified_elements(unified_c)
result.set_matched_elements(matched_c)
result.new_scores = scores
query.sort_results()
group = query.sorted_results[0]
# for each sub group ...
structures = []
current_structure = 0
for subgroup in group:
# next substructure group in the overall rank...
current_structure += 1
structure_elements = []
for sg_idx, expression in enumerate(subgroup):
structure_elements.append(expression)
structures.append(structure_elements)
else:
# just one expression in rank, no need to re-evaluate score ...
candidate_data = candidates_data[0]
candidate_exp = candidate_data[0]
rank = int(candidate_data[1])
scores = [float(part) for part in candidate_data[2:5]]
# the list of structures only contains one structure with the same structure
structures = [[candidate_exp]]
return (rank, scores, structures)
def find_substructures(expressions_data):
sub_groups = []
query_expression, candidates_data = expressions_data
if len(candidates_data) > 1:
query = Query("query", query_expression)
# create query tree ....
rank = -1
scores = None
prev_scores = None
for data_idx, candidate_data in enumerate(candidates_data):
candidate_exp = candidate_data[0]
rank = int(candidate_data[1])
query.add_result(0, "", 0, candidate_exp, 0.0)
result = query.results[candidate_exp]
try:
sim_res = SIM_FUNCTION(query.tree, result.tree, query.constraints)
scores, matched_q, matched_c, unified_c, wildcard_c, unified = sim_res
except:
print("Error processing: ")
print("Q: " + query_expression, flush=True)
print("C: " + candidate_exp, flush=True)
continue
result.set_unified_elements(unified_c)
result.set_matched_elements(matched_c)
result.set_wildcard_matches(wildcard_c)
result.new_scores = scores
if prev_scores is None:
prev_scores = scores
else:
if prev_scores != scores:
print("Error: Scores changed!")
print(prev_scores)
print(scores)
prev_scores = scores
query.sort_results()
if len(query.sorted_results) > 1:
print("Error: Did not expect More than 1 group")
print("-> " + str(len(query.sorted_results)))
group = query.sorted_results[0]
# for each sub group ...
structures = []
current_structure = 0
for subgroup in group:
# next substructure group in the overall rank...
current_structure += 1
structure_elements = []
for sg_idx, expression in enumerate(subgroup):
structure_elements.append(expression)
structures.append(structure_elements)
else:
# just one expression in rank, no need to re-evaluate score ...
candidate_data = candidates_data[0]
candidate_exp = candidate_data[0]
rank = int(candidate_data[1])
scores = [float(part) for part in candidate_data[2:(2 + N_SCORES)]]
# the list of structures only contains one structure with the same structure
structures = [[candidate_exp]]
return (rank, scores, structures)
