[cmd] Rewrite solutions2common_graph..

- to be renamed
- support of CSV, JSON, GraphML exports

command_line/cadbiom_cmd/solution_sort.py

@@ -344,7 +344,7 @@ def get_solution_graphs(sol_steps, transitions):
 
 
 def queries_2_common_graph(output_dir, model_file, path,
-                           make_graphs=True, make_csv=False, make_json=True): # force option
+                           make_graphs=True, make_csv=False, make_json=False):
     """Entry point for solutions_2_common_graph
 
     Create a GraphML formated file containing a unique representation of **all**
@@ -359,29 +359,107 @@ def queries_2_common_graph(output_dir, model_file, path,
     :param output_dir: Output path.
     :param model_file: Filepath of the model.
     :param path: Filepath/directory of a/many complete solutions files.
+    :key make_graphs: (optional) Make a GraphML for each query results in path.
+        default: True
+    :key make_csv: (optional) Make a **global** CSV for all query results in path.
+        default: False
+    :key make_json: (optional) Make a JSON dump of each query results in path.
+        default: False
     :type output_dir: <str>
     :type model_file: <str>
     :type path: <str>
+    :type make_graphs: <boolean>
+    :type make_csv: <boolean>
+    :type make_json: <boolean>
     """
 
+    def write_csv(data):
+        """Write given data in CSV file; then flush the file descriptor
+
+        .. note:: data is modified in place.
+
+        :param data: Data to be serialized in JSON.
+        """
+        # Handle nested dictionaries by flattening them
+        data.update({
+            sub_key: sub_val
+            for key, val in data.items() if isinstance(val, dict)
+            for sub_key, sub_val in val.items()
+        })
+
+        writer.writerow(data)
+        f_d.flush()
+
+
+    def do_magic(solution_file):
+        # Get query string from the name of the solution file
+        query = get_query_from_filename(model_file, solution_file)
+
+        LOGGER.info("Processing %s query...", query)
+
+        # generator of ("Ax Bx", [['h2', 'h00'], ['h3'], ['h0', 'h1'], ['hlast']])
+        solutions = tuple(load_solutions(solution_file))
+        graphs = get_solution_graphs(
+            solutions,
+            model_transitions
+        )
+        # Get common graph
+        graph = merge_graphs(graphs)
+
+        if make_graphs:
+            # Write graph
+            export_graph(output_dir, query, '', graph)
+
+        if make_json or make_csv:
+            # Export to json file (similaire à model_graph_info() ...)
+            # Frontiers
+            frontiers = set(it.chain(*[sol.split() for sol, _ in solutions]))
+            solutions_info = {
+                'modelFile': model_file,
+                'query': query,
+                'solutions': len(solutions),  # nb trajectories/solutions
+                'boundaries': len(frontiers),  # frontier places
+            }
+            # Advanced metrics
+            get_solutions_graph_data(graph, solutions_info, True)
+
+            if make_json:
+                # Save to <solution_filename>_graph_summary.json
+                write_json(output_dir, solution_file, "_graph_summary", solutions_info)
+
+            if make_csv:
+                # Export to csv graphs_summary.csv
+                write_csv(solutions_info)
+
+
     # Check valid input file/directory
     assert os.path.isfile(path) or os.path.isdir(path)
 
     # Get transitions from the model
     model_transitions = get_transitions_from_model_file(model_file)
 
-    if os.path.isfile(path):
-        # The given path is a solution file
-        graphs = get_solution_graphs(
-            load_solutions(path),
-            model_transitions
+
+    if make_csv:
+        # Init the CSV file now
+        import csv
+        f_d = open(output_dir + 'graphs_summary.csv', 'w')
+        writer = csv.DictWriter(
+            f_d,
+            [
+                "query", "solutions", "boundaries", "graph_nodes", "graph_edges",
+                "strongly_connected", "max_degree", "min_degree", "average_degree"
+            ],
+            extrasaction="ignore", # Ignore unknown fieldnames
+            delimiter=str(';'),
         )
-        # Get query string from the name of the solution file
-        query = get_query_from_filename(model_file, path)
-        # Write graph
-        export_graph(output_dir, query, '', merge_graphs(graphs))
+        writer.writeheader()
 
 
+    # File management...
+    if os.path.isfile(path):
+        # The given path is a solution file
+        do_magic(path)
+
     elif os.path.isdir(path):
         # The given path is a directory
         path = path if path[-1] == '/' else path + '/'
@@ -391,24 +469,16 @@ def queries_2_common_graph(output_dir, model_file, path,
         for file_number, solution_file in \
                 enumerate(glob.glob(path + '*mac_complete.txt'), 1):
 
-            # Get query string from the name of the solution file
-            query = get_query_from_filename(model_file, solution_file)
-
-            LOGGER.info("Processing %s query...", query)
-
-            # generator of ("Ax Bx", [['h2', 'h00'], ['h3'], ['h0', 'h1'], ['hlast']])
-            solutions = tuple(load_solutions(solution_file))
-            graphs = get_solution_graphs(
-                solutions,
-                model_transitions
-            )
-
-            # Write graph
-            export_graph(output_dir, query, '', merge_graphs(graphs))
+            do_magic(solution_file)
 
         LOGGER.info("Files processed: %s", file_number)
         assert file_number != 0, "No *mac_complete.txt files found!"
 
+
+    if make_csv:
+        # Close the CSV file descriptor
+        f_d.close()
+
 ## Matrices of occurrences #####################################################
 
 def solutions_2_occcurrences_matrix(output_dir, model_file, path,