Charles Karpati | Importing Data

class Intake:
 
   # 1. Recursively calls self/getData until something valid is given.
   #    Returns df or False. Calls readInGeometryData. or pulls csv directly.
   # Returns df or False.
   def getData(url, interactive=False):
     escapeQuestionFlags = ["no", '', 'none']
     if ( Intake.isPandas(url) ): return url
     if (str(url).lower() in escapeQuestionFlags ): return False
     if interactive: print('Getting Data From: ', url)
     try:
       if ([ele for ele in ['pgeojson', 'shp', 'geojson'] if(ele in url)]):
         print('gothere', url)
         from dataplay import geoms
         print('gothere1')
         df = geoms.readInGeometryData(url=url, porg=False, geom='geometry', lat=False, lng=False, revgeocode=False,  save=False, in_crs=2248, out_crs=False)
       elif  ('csv' in url): df = pd.read_csv( url )
       return df
     except:
       if interactive: return Intake.getData(input("Error: Try Again?  ( URL/ PATH or  'NO'/  ) " ), interactive)
       return False
 
   # 1ai. A misnomer. Returns Bool.
   def isPandas(df): return isinstance(df, pd.DataFrame) or isinstance(df, gpd.GeoDataFrame) or isinstance(df, tuple)
 
 
   # a1. Used by Merge Lib. Returns valid (df, column) or (df, False) or (False, False).
   def getAndCheck(url, col='geometry', interactive=False):
     df = Intake.getData(url, interactive) # Returns False or df
     if ( not Intake.isPandas(df) ):
       if(interactive): print('No data was retrieved.', df)
       return False, False
     if (isinstance(col, list)):
       for colm in col:
         if not Intake.getAndCheckColumn(df, colm):
           if(interactive): print('Exiting. Error on the column: ', colm)
           return df, False
     newcol = Intake.getAndCheckColumn(df, col, interactive) # Returns False or col
     if (not newcol):
       if(interactive): print('Exiting. Error on the column: ', col)
       return df, col
     return df, newcol
 
   # a2. Returns Bool
   def checkColumn(dataset, column): return {column}.issubset(dataset.columns)
 
   # b1. Used by Merge Lib. Returns Both Datasets and Coerce Status
   def coerce(ds1, ds2, col1, col2, interactive):
     ds1, ldt, lIsNum = Intake.getdTypeAndFillNum(ds1, col1, interactive)
     ds2, rdt, rIsNum  = Intake.getdTypeAndFillNum(ds2, col2, interactive)
 
     ds2 = Intake.coerceDtypes(lIsNum, rdt, ds2, col2, interactive)
     ds1 = Intake.coerceDtypes(rIsNum, ldt, ds1, col1, interactive)
 
     # Return the data and the coerce status
     return ds1, ds2, (ds1[col1].dtype == ds2[col2].dtype)
 
    # b2. Used by Merge Lib. fills na with crazy number
   def getdTypeAndFillNum(ds, col, interactive):
     dt = ds[col].dtype
     isNum = dt == 'float64' or dt == 'int64'
     if isNum: ds[col] = ds[col].fillna(-1321321321321325)
     return ds, dt, isNum
 
    # b3. Used by Merge Lib.
   def coerceDtypes(isNum, dt, ds, col, interactive):
     if isNum and dt == 'object':
       if(interactive): print('Converting Key from Object to Int' )
       ds[col] = pd.to_numeric(ds[col], errors='coerce')
       if interactive: print('Converting Key from Int to Float' )
       ds[col] = ds[col].astype(float)
     return ds
 
   # a3. Returns False or col. Interactive calls self
   def getAndCheckColumn(df, col, interactive):
     if Intake.checkColumn(df, col) : return col
     if (not interactive): return False
     else:
         print("Invalid column given: ", col);
         print(df.columns);
         print("Please enter a new column fom the list above.");
         col = input("Column Name: " )
         return Intake.getAndCheckColumn(df, col, interactive);