Merge branch 'master' of http://git.eseb.ovh:3006/DemiSel/Messenger_1Million_Stats

export_data-v2.py

@@ -6,14 +6,14 @@ from matplotlib import pyplot as plt
 import os
 import json
 import re
-
+import sys, getopt
 
 
 # ------------------------------------------------
-#                   Constants
+#                   Globals
 # ------------------------------------------------
 
-DATA_PATH = 'D:/Files/Data/Messenger/'
+DATA_PATH = './data/'
 
 OTHER_LABEL = 'Les Autres'
 
@@ -25,11 +25,30 @@ CONTENT = 'content'
 TIMESTAMP = 'timestamp_ms'
 SENDER = 'sender_name'
 
+HELP = """General options :
+    -h, --help          Consulter l'aide
+    --path=<path>       Redéfinir le chemin d'accès aux données (par défaut ./data)
+    """
 
 # ------------------------------------------------
 #                   Functions
 # ------------------------------------------------
 
+def handleArguments(argv):
+    try:
+        opts, args = getopt.getopt(argv, 'h',['help','path='])
+    except getopt.GetoptError:
+        print('Usage:\n '+os.path.basename(__file__)+' <command> [option]\n')
+        print(HELP)
+        sys.exit(2)
+
+    for opt, arg in opts:
+        if opt in ('-h', '--help'):
+            print(HELP)
+            sys.exit()
+        elif opt in ('--path'):
+            DATA_PATH = arg
+
 def readBrokenFbJson(datafile_path):
     # ntm facebook
     # https://stackoverflow.com/questions/50008296/facebook-json-badly-encoded
@@ -42,11 +61,12 @@ def readBrokenFbJson(datafile_path):
         return json.loads(repaired.decode('utf8'))
 
 def computeData():
-    # Tous les fichiers du dossier sont traités sans validation
-    datafiles_path = [DATA_PATH + filename for filename in os.listdir(DATA_PATH) if filename [-4:] == "json"]
-    print(datafiles_path)
+    # Tous les fichiers du dossier sont traités sans distinction
+    datafiles_path = [DATA_PATH + filename for filename in os.listdir(DATA_PATH)]
     messages, participants = [], []
 
+    print(datafiles_path)
+    
     for datafile_path in datafiles_path:
         datacontent = readBrokenFbJson(datafile_path)
         if datacontent is None : continue
@@ -81,8 +101,8 @@ def computeParticipation(messages):
     return sorted(result.items(), key = lambda x: x[1])
 
 def mergeSmallParticipation(rawParticipation, threshold = 1):
-    values = [e[1] for e in participation]
-    labels = [e[0] for e in participation]
+    values = [e[1] for e in rawParticipation]
+    labels = [e[0] for e in rawParticipation]
 
     totalValues = sum(values)
 
@@ -107,9 +127,14 @@ def displayParticipation(participation):
 #                   Main Code
 # ------------------------------------------------
 
-participants, messages = computeData()
-messages = filterMessages(messages)
+def main(argv):
+    handleArguments(argv)
+
+    participants, messages = computeData()
+    messages = filterMessages(messages)
 
-participation = computeParticipation(messages)
-displayParticipation(participation)
+    participation = computeParticipation(messages)
+    displayParticipation(participation)
 
+if __name__ == "__main__":
+    main(sys.argv[1:])