将 Android LogCat 文本文件结构化为结构化 Pandas DF

Question

我想将 LogCat 文本文件的行转换为结构化的 Pandas DF。我似乎无法正确地概念化我将如何做到这一点......这是我的基本伪代码：

dateTime = []
processID = []
threadID = []
priority = []
application = []
tag = []
text = []

logFile = "xxxxxx.log"

for line in logfile:
     split the string according to the basic structure
     dateTime = [0]
     processID = [1]
     threadID = [2]
     priority = [3]
     application = [4]
     tag = [5]
     text = [6]
     append each to the empty list above

write the lists to pandas dataframe & add column names

问题是：我不知道如何用这种结构正确定义分隔符

08-01 14:28:35.947 1320 1320 D wpa_xxxx: wlan1:skip--ssid

Answer 1

import re
import pandas as pd

ROW_PATTERN = re.compile(r"""(\d{2}\-\d{2} \d{2}\:\d{2}\:\d{2}\.\d+) (\d+) (\d+) ([A-Z]) (\S+) (\S+) (\S+)""")

with open(logFile) as f:
    s = pd.Series(f.readlines())

df = s.extract(ROW_PATTERN)
df.columns = ['dateTime', 'processID', 'threadID', 'priority', 'application', 'tag', 'text']

这会将logFile的每一行读入Series中的一行，然后可以通过正则表达式中的每个组将其扩展为DataFrame。这假定08-01 14:28:35.947 是每行中的第一个值，后续值用空格分隔。