Prepare getdates for running average calculation

Include days with 0 tweets in getdates output

getdates.py

@@ -4,74 +4,95 @@ import tools
 
 import sys, datetime
 
-def checkDate(date):
+def getDate(date):
  try:
-  datetime.datetime.strptime(date, '%Y-%m-%d')
-  return True
+  return datetime.datetime.strptime(date, '%Y-%m-%d')
  except ValueError:
-  return False
+  raise ValueError("Dates must be in YYYY-MM-DD format.")
 
-def dateArgs(argv = sys.argv):
- fr = ""
- to = ""
+def dateArgs(argv = sys.argv[1:]):
  strings = []
+ fr = None
+ to = None
+ av = 0
 
  mode = 0
 
- for arg in argv[1:]:
-  if mode == 0:
+ for arg in argv:
+  if mode == 0 or (mode == 3 and not isinstance(arg,int)):
    if arg == "-f":
     mode = 1
    elif arg == "-t":
     mode = 2
+   elif arg == "-a":
+    mode = 3
    else:
     strings += [arg]
-  elif checkDate(arg):
-   if mode == 1:
-    fr = arg
     mode = 0
-   else:
-    to = arg
+  elif mode == 3:
+   av = arg
+   mode = 0
+  elif mode == 1:
+    fr = getDate(arg)
     mode = 0
   else:
-   raise ValueError("Dates must be in YYYY-MM-DD format.")
+    to = getDate(arg)
+    mode = 0
 
- if not mode == 0:
+ if mode in (1, 2):
   raise ValueError("Date missing.")
- return strings, fr, to
+
+ if to != None and fr != None and to < fr:
+  raise ValueError("From date must be before To date.")
+ 
+ return strings, fr, to, av
     
 
-def queryBuilder(strings = [], fr = "", to = ""):
- query = "SELECT SUBSTR(t.timestamp,0,11) AS 'Date', (SELECT COUNT(*) FROM tweets e WHERE SUBSTR(e.timestamp,0,11) = SUBSTR(t.timestamp,0,11)) AS 'Tweets'"
-
- for string in strings:
-  query += ", (SELECT COUNT(*) FROM tweets e WHERE SUBSTR(e.timestamp,0,11) = SUBSTR(t.timestamp,0,11) AND LOWER(e.text) LIKE '%" + string.lower() + "%') AS '" + string + "'"
-
- query += "FROM tweets t "
- status = 0
-
- if not fr == "":
-  status += 1
-
- if not to == "":
-  status += 2
-
- if status > 0:
-  query += "WHERE "
-  if status in (1,3):
-   query += "SUBSTR(t.timestamp,0,11) >= '" + fr + "' "
-   if status == 3:
-    query += "AND "
-  if status > 1:
-   query += "SUBSTR(t.timestamp,0,11) <= '" + to + "' "
-    
- return query + "GROUP BY SUBSTR(t.timestamp,0,11)"
+def queryBuilder(date, string = ""):
+ return "SELECT COUNT(*) FROM tweets WHERE SUBSTR(timestamp,0,11) = '%s' AND LOWER(text) LIKE '%%%s%%'" % (date, string)
 
 
-def getTweetsByDate(strings = [], path = tools.config.dbpath, fr = "", to = ""):
+def getFLDate(db, val = 0):
+ if val == 0:
+  mode = "MIN"
+ else:
+  mode = "MAX"
+
+ return getDate(str(db.executeQuery("SELECT %s(SUBSTR(timestamp,0,11)) FROM tweets" % mode)[0][0]))
+
+
+def dateList(fr, to):
+ return [[(fr+datetime.timedelta(days=i)).strftime('%Y-%m-%d')] for i in range((to+datetime.timedelta(days=1)-fr).days)]
+
+
+def fillList(db, string, cur, av):
+ for day in cur:
+  day += db.executeQuery(queryBuilder(day[0], string))[0]
+
+ return cur
+
+
+def getHeaders(strings, av):
+ return [["Date", "Tweets"] + [string for string in strings]]
+
+
+def getTweetsByDate(strings = [], fr = None, to = None, av = 0, path = tools.config.dbpath, headers = False):
  db = tools.dbHelper(path)
- return list(db.executeQuery(queryBuilder(strings,fr,to)))
+
+ if fr == None:
+  fr = getFLDate(db)
+ if to == None:
+  to = getFLDate(db, 1)
+
+ cur = dateList(fr, to)
+
+ for string in [""] + strings:
+  cur = fillList(db, string, cur, av)
+
+ if headers:
+  cur = getHeaders(strings, av) + cur
+
+ return cur
 
 if __name__ == "__main__":
- strings, fr, to = dateArgs()
- tools.printCSV([["Date", "Tweets"] + [s for s in strings]] + getTweetsByDate(strings = strings, fr = fr, to = to))
+ tools.printCSV(getTweetsByDate(*dateArgs(), headers = True))