需要用Python解析工具日志文件并将结果保存为Excel或CSV

Question

我需要用Python来解析下面的日志文件：

Log file created: 9/10/2014 4:03:21 PM
----------------------------------------
Ticks = 2408967      <3360> <UpdatePlaybackStatusInfo> <0> Avg Prefetch(ms): 157.739, Avg Render(ms): 25.7375, Avg Display FPS: 27.3688
Ticks = 3371181      <3360> <UpdatePlaybackStatusInfo> <0>  Frames dropped during playback: 0 / 219, Preroll(ms): 812.849
Ticks = 3371181      <3360> <UpdatePlaybackStatusInfo> <0> Avg Prefetch(ms): 17.1389, Avg Render(ms): 33.8339, Avg Display FPS: 29.5562
Ticks = 3465531      <10548> <Assert> <0> Debug Assert failed!
Ticks = 3465531      <10548> <Assert> <0> wglMakeCurrent failed: Error 0: The operation completed successfully.

我想提取<UpdatePlaybackStatusInfo>和<Assert>这两种记录，每种记录都要包含它的各个字段，然后把它们保存成Excel或CSV文件。

有没有人能帮我做到这一点呢？

Answer 1

我不太确定不等号里面的值是什么，所以我用foo和bar代替了它们。像这样应该可以解决问题：

import re
import csv

filtered_messages = ['UpdatePlaybackStatusInfo', 'Assert']
fieldnames = ['ticks', 'foo', 'type', 'bar', 'message']

with open('log.txt') as log:
    with open('output.csv', 'w') as csv_file:
        writer = csv.DictWriter(csv_file, delimiter=',', fieldnames=fieldnames)
        writer.writerow(dict((fn,fn) for fn in fieldnames))
        for line in log:
            match = re.search(r'^Ticks = (?P<ticks>\d+)\s+<(?P<foo>\d+)> <(?P<type>\w+)> <(?P<bar>\d+)>\s+(?P<message>.+)$', line)
            if match is not None and match.group('type') in filtered_messages:
                writer.writerow(match.groupdict())

输出（以CSV格式）：

ticks   foo type    bar message

2408967 3360    UpdatePlaybackStatusInfo    0   Avg Prefetch(ms): 157.739, Avg Render(ms): 25.7375, Avg Display FPS: 27.3688

3371181 3360    UpdatePlaybackStatusInfo    0   Frames dropped during playback: 0 / 219, Preroll(ms): 812.849

3371181 3360    UpdatePlaybackStatusInfo    0   Avg Prefetch(ms): 17.1389, Avg Render(ms): 33.8339, Avg Display FPS: 29.5562

3465531 10548   Assert  0   Debug Assert failed!

3465531 10548   Assert  0   wglMakeCurrent failed: Error 0: The operation completed successfully.