produce ascii histograms with scripts/matplotlib_histograms.py

Add new options to matplotlib_histograms.py:
	-i - open interactive matplotlib plots
	-n - no ascii histogram

chattyparser.plot_hist_ascii handles histograms with non int keys

2 files changed, 34 insertions, 19 deletions

diff --git a/scripts/matplotlib_histograms.py b/scripts/matplotlib_histograms.py
index 4d3af45..7fae59a 100755
--- a/scripts/matplotlib_histograms.py
+++ b/scripts/matplotlib_histograms.py
@@ -11,37 +11,44 @@ import sys
 
 import matplotlib.pyplot as plt
 
+currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
+parentdir = os.path.dirname(currentdir)
+sys.path.insert(0, parentdir)
+
+import src.chattyparser
 
 def main():
     parser = argparse.ArgumentParser(description="Plot histograms using a malt or chattymalloc output file")
     parser.add_argument("input_file", help="path to malt or chattymalloc output file", type=str)
     parser.add_argument("-e", "--export", help="export to csv", action="store_true")
+    parser.add_argument("-n", "--no-ascii", help="don't output a ascii histogram", action="store_true")
+    parser.add_argument("-i", "--interactive", help="open interactive matplotlib histogram plots",
+                        action="store_true")
     args = parser.parse_args()
 
-    fname, fext = os.path.splitext(args.input_file)
-    fname = os.path.basename(fname)
+    fpath, fext = os.path.splitext(args.input_file)
+    fname = os.path.basename(fpath)
     # chattymalloc
     if fname.startswith("chatty") and fext == ".txt":
-        # import chattyparser
-        currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
-        parentdir = os.path.dirname(currentdir)
-        sys.path.insert(0, parentdir)
-
-        chattyparser = importlib.import_module("src.chattyparser")
-
-        hist, _, _ = chattyparser.parse(args.input_file, coll_size=False)
+        hist, calls, _ = src.chattyparser.parse(args.input_file, coll_size=False)
     # malt
     else:
         with open(args.input_file, "r") as json_file:
-            hist = json.load(json_file)["memStats"]["sizeMap"]
+            malt_res =  json.load(json_file)
+
+        hist = malt_res["memStats"]["sizeMap"]
+        calls = {}
+        for thread in malt_res["threads"]:
+            for func, data in thread["stats"].items():
+                calls[func] = calls.get(func, 0) + data["count"]
 
     if args.export:
-        with open(f"{fname}.csv", "w") as csv_file:
+        with open(f"{fpath}.csv", "w") as csv_file:
             print("Size", "Amount", file=csv_file)
             for size, amount in hist.items():
                 print(size, amount, file=csv_file)
 
-    else:
+    if args.interactive:
         sizes = []
         sizes_smaller_4K = []
         sizes_smaller_32K = []
@@ -74,5 +81,8 @@ def main():
         plt.title("All sizes")
         plt.show()
 
+    if not args.no_ascii:
+        src.chattyparser.plot_hist_ascii(f"{fpath}.hist.txt", hist, calls)
+
 if __name__ == "__main__":
     main()
diff --git a/src/chattyparser.py b/src/chattyparser.py
index 4e4acae..87a5c05 100755
--- a/src/chattyparser.py
+++ b/src/chattyparser.py
@@ -177,17 +177,21 @@ def plot_profile(total_sizes, trace_path, plot_path, sizes):
 
 def plot_hist_ascii(path, hist, calls):
     """Plot an ascii histogram"""
+    # make sure all sizes are stored as ints
+    for nonint_key in [key for key in hist.keys() if type(key) is not int]:
+        hist[int(nonint_key)] = hist[nonint_key]
+        del hist[nonint_key]
+
     bins = {}
     for size in sorted(hist):
-        size_class = int(size / 16)
+        size_class = size / 16
         bins[size_class] = bins.get(size_class, 0) + hist[size]
 
 
-    total = sum(calls.values()) - calls["free"]
     with open(path, "w") as hist_file:
-        print("Total function calls:", total, file=hist_file)
-        for func in TRACE_REGEX:
-            print(func, calls[func], file=hist_file)
+        print("Total function calls:", sum(calls.values()), file=hist_file)
+        for func, func_calls in calls.items():
+            print(func, func_calls, file=hist_file)
 
         print(file=hist_file)
 
@@ -201,6 +205,7 @@ def plot_hist_ascii(path, hist, calls):
         print(file=hist_file)
 
         print("Histogram of sizes:", file=hist_file)
+        total = sum(hist.values())
         sbins = sorted(bins)
         binmaxlength = str(len(str(sbins[-1])) + 1)
         amountmaxlength = str(len(str(sorted(bins.values())[-1])))
@@ -220,7 +225,7 @@ if __name__ == "__main__":
         print("License GPLv3: GNU GPL version 3 <http://gnu.org/licenses/gpl.html>")
         exit(0)
 
-    if len(sys.argv) != 2:
+    if len(sys.argv) != 2 or sys.argv[1] in ["-h", "--help"]:
         print("chattyparser: parse chattymalloc output and",
               "create size histogram and memory profile", file=sys.stderr)
         print(f"Usage: {sys.argv[0]} chattymalloc-file", file=sys.stderr)