Here's an example of my logs in a txt file (trans.txt):
22 July 2021 09:35:54 Withdrawn: RM500
22 July 2021 09:35:54 Withdrawn: RM500
22 August 2021 09:35:54 Withdrawn: RM500
22 August 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 October 2021 09:35:54 Withdrawn: RM500
22 October 2021 09:35:54 Withdrawn: RM500
22 November 2021 09:35:54 Withdrawn: RM500
22 November 2021 09:35:54 Withdrawn: RM500
22 December 2021 09:35:54 Withdrawn: RM500
22 December 2021 09:35:54 Withdrawn: RM500
how to print a specific range of logs based on months? Imagine if i wanna print logs quarterly or half-yearly, and my pc local time is November.
I'm expecting python to print out all logs from September to November, since i want to print logs quarterly based on my local time.
EDIT:
Below are my attempt, but still can't achieve what i intended
# ↓Pulls out local time's from user pc
local_timeMonth = time.strftime("%B", obj)
# ↓Opens user's transaction logs and put them in a list
hand1 = open("trans.txt", "r")
list1 = hand1.read().splitlines()
hand1.close()
# ↓Creates a another file to store all logs with the month that is
# intended to be printed and excludes months that are not relevant,
# but all it does is store logs from November back until January
#it excludes December though (Pc local time is November)
for i in range(0, len(list1)):
if local_timeMonth in list1[i]:
test = "\n".join(list1[i::-1])
hand = open("tempLogs.txt", "w")
hand.write(test)
hand.close()
# ↓Place logs only from 3 months into list
f = open("tempLogs.txt", "r")
line_numbers = [0, 1, 2]
lines = []
# ↓Puts specific month's of log in to another list
for i, line in enumerate(f):
if i in line_numbers:
lines.append(line.strip())
elif i > 2:
break
# ↓Print list out into readable format
for i in lines:
print(i)
f.close()
CodePudding user response:
Here is a naive way to process your logs.
Let's import data you provided in your MCVE:
import io
import pandas as pd
text = io.StringIO("""22 July 2021 09:35:54 Withdrawn: RM500
22 July 2021 09:35:54 Withdrawn: RM500
22 August 2021 09:35:54 Withdrawn: RM500
22 August 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 September 2021 09:35:54 Withdrawn: RM500
22 October 2021 09:35:54 Withdrawn: RM500
22 October 2021 09:35:54 Withdrawn: RM500
22 November 2021 09:35:54 Withdrawn: RM500
22 November 2021 09:35:54 Withdrawn: RM500
22 December 2021 09:35:54 Withdrawn: RM500
22 December 2021 09:35:54 Withdrawn: RM500""")
frame = pd.read_csv(text, header=None, names=["raw"])
If adding a separator between timestamp and message or formatting date in a fixed length format such ISO-8601 is not an option then you need to cope with an extra challenge: your data is not a Fixed With Format nor a CSV file format.
Let's parse raw log lines naively (probably not efficient when scaling):
raw = frame.pop("raw")
frame["timestamp"] = raw.apply(lambda x: pd.to_datetime(" ".join(x.split(" ")[:4])))
frame["type"] = raw.apply(lambda x: x.split(" ")[4].replace(":", ""))
frame["message"] = raw.apply(lambda x: " ".join(x.split(" ")[5:]))
frame = frame.set_index("timestamp")
Once your frame is setup, indexing quarterly is as simple as:
t0 = pd.Timestamp.now().round("1D")
q1 = t0 - pd.offsets.QuarterBegin(n=1)
q2 = t0 pd.offsets.QuarterEnd(n=0)
frame.loc[q1:q2,:]
Which returns expected lines:
type message
timestamp
2021-09-22 09:35:54 Withdrawn RM500
2021-09-22 09:35:54 Withdrawn RM500
2021-09-22 09:35:54 Withdrawn RM500
2021-10-22 09:35:54 Withdrawn RM500
2021-10-22 09:35:54 Withdrawn RM500
2021-11-22 09:35:54 Withdrawn RM500
2021-11-22 09:35:54 Withdrawn RM500
2021-12-22 09:35:54 Withdrawn RM500
2021-12-22 09:35:54 Withdrawn RM500
If you have to parse high volume of logs then you probably will need to improve performance of this naive solution. A good start would be in anyway change log format into a well known format either CSV or FWF.