Untitled

mail@pastecode.io avatar
unknown
plain_text
a year ago
1.4 kB
1
Indexable
Never
class ConvertLogFileToDataFrameDoFn(beam.DoFn):
    def process(self, element):
        file_path = element
        print(file_path)
        list_of_files.append(file_path)
        print(list_of_files)
        # read the text file and split each line by comma
        with open(file_path, "r") as f:
            lines = f.readlines()

        # create an empty DataFrame with the specified columns
        columns=['timestamp','hostname','process_name','process_id','log_text']
        df = pd.DataFrame(columns=columns)

        # use regular expressions to extract the values for each column

        data = []
        for line in lines:
            #match = re.search(r'Time: (\S+).*Computer: (\S+).*Event Id: (\d+),Message: (.*)Level: (\d+),Channel: (\S+)', line)
            match = re.search(r'Time: (\S+),\s*Computer: (\S+),\s*Event Id: (\d+),\s*Message: ([^,]*).*Level: (\d+),\s*Channel: (\S+)', line)
            if match:
                timestamp, hostname, event_id, log_text, level, channel = match.groups()
                hostname_prefix = hostname.split(".")[0]
                data.append((timestamp, hostname_prefix , channel, event_id, log_text.strip()))

        # convert the list of tuples to a DataFrame
        df = pd.DataFrame(data, columns=columns)
        #print(df.head(6))

        return [df]