YoutubeTemplate.py

Go to the documentation of this file.

import re
import awswrangler as wr
from pandas import to_datetime
 
class YoutubeTemplate:
    def preprocessing(self, filename, features, session, rel_col):
        """Loads file and do some fixes to fit it to our standard and make it able to be processed
 
        Args:
            filename (str): s3 full path of filename
            features (dict): contains delimiter, skip_rows and encoding required for current file
            session (boto3 obj): AWS client connection
            rel_col (str): column name used to identify upc/release_id
        Returns: df (pandas dataframe)
        """
        delimiter = features["delimiter"]
        skip_rows = features["skip_rows"]
        encoding = features["encoding"]
        df = wr.s3.read_csv(filename, sep=delimiter, skiprows=skip_rows, encoding=encoding, dtype={rel_col:str}, keep_default_na=False, low_memory=False, boto3_session=session)
        try:
            df["Month"] = to_datetime(df["Month"], format='%Y%m')
        except:
            df["Day"] = to_datetime(df["Day"], format='%Y%m%d')
        df["gettypeyt"] = df.apply(lambda Row: self.type(Row['Asset Type']), axis=1)
        df["Asset Type"] = df["gettypeyt"]
        df.drop("gettypeyt", axis=1, inplace=True)
        return df
    
    def type(self, asset_type):
        """Maps the sale type with the name specified by the client according their requirements
        Args:
            asset_type (str): sale type shown in the original file
        Returns: (sale type in a single one letter)
        """
        if "Sound Recording" in asset_type:
            return 'Youtube Content ID'
        elif "Art Track" in asset_type or "Music Video" in asset_type:
            return 'Youtube Music'
        # return 'Youtube'
 
    def date(self, filename):
        """Sets date column given the filename (it contains the date)
           For example YouTube_GSDistroPartners_M_20220801_asset_raw_v1-1.csv
        Args:
            filename (str): current filename
        Returns: date_str (str)
        """
        # "YouTube_GSDistroPartners_M_20220801_asset_raw_v1-1.csv"
        date = re.findall(r'M_20\d{6}', filename)[0]
        date = date.replace("M_", "")
        date_str = date[:4]+"-"+date[4:6]
        # df["date_from_file"] = date_str
        return date_str