Skip to content
Advertisement

How to parse and match with multiple regexes

I have an input data of the form:

[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

I need to parse through this data and the IN: / OUT: /INOUT: depending on three regexes given as:

regex1 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s"
regex2 = r"[2]s*OUT:s*(S+?)s*INOUT:s*(S+?)s"
regex3 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s.*?.s*OUT:s*(S+?)s*INOUT:s*(S+?)s"

My output should be:

IN_r1 2.12 INOUT_r1 3.52
IN_r3 2.12 INOUT1_r3 3.52 OUT_r3 2.42 INOUT2_r3 2.62

OUT_r2 2.42 INOUT_r2 2.62
IN_r3 2.12 INOUT1_r3 3.52 OUT_r3 2.42 INOUT2_r3 2.62

IN_r1 2.12 INOUT_r1 3.52 
OUT_r2 2.42 INOUT_r2 2.62
IN_r3 2.12 INOUT1_r3 3.52 OUT_r3 2.42 INOUT2_r3 2.62

The code I tried:

import re
regex1 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s"
regex2 = r"[2]s*OUT:s*(S+?)s*INOUT:s*(S+?)s"
regex3 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s.*?.s*OUT:s*(S+?)s*INOUT:s*(S+?)s"

data = "
[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)
"

lines = re.split("[2]",data)

for line in lines:


    if re.search(regex1,data):
        tracks = re.findall(regex1,data,re.DOTALL)
        for track in tracks:
            input,inout = (float(z) for z in track)
            with open("checked_ant.txt",'a') as a:
                print("IN_r1",input,"INOUT_r1",inout,file=a)
    elif re.search(regex2,data):
        tracks = re.findall(regex2,data,re.DOTALL)
        for track in tracks:
            output,inout = (float(z) for z in track)
            with open("checked_ant.txt",'a') as a:
                print("OUT_r2",output,"INOUT_r2",inout,file=a)
    elif re.search(regex3,data):
        tracks = re.findall(regex3,data,re.DOTALL)
        for track in tracks:
            input,inout1,output,inout2 = (float(z) for z in track)
            with open("checked_ant.txt",'a') as a:
                print("IN_r3",input,"INOUT1_r3",inout1,"OUT_r3",output,"INOUT2_r3",inout2,file=a)

The problem I face is that it does not parse correctly and it is not getting matched for each subdata beginning with [2]

Advertisement

Answer

Though I find the requirement quite strange(regex is provided and cannot change), I got the expected result. Can you try.

import re

s = '''[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)

[2]   IN: 2.12    INOUT: 3.52  (Input)
[2]   OUT: 2.42   INOUT: 2.62  (Output)
[2]   IN: 2.12    INOUT: 3.52  (Input)
      OUT: 2.42   INOUT: 2.62  (Output)'''

r1 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s"
r2 = r"[2]s*OUT:s*(S+?)s*INOUT:s*(S+?)s"
r3 = r"[2]s*IN:s*(S+?)s*INOUT:s*(S+?)s.*?.s*OUT:s*(S+?)s*INOUT:s*(S+?)s"


def g(reg, s, n):
    return float(re.search(reg, s).group(n))


paras = s.split('nn')
for p in paras:
    if re.search(r1, p):
        print(f'IN_r1 {g(r1, p, 1)} INOUT_r1 {g(r1, p, 2)}')
    if re.search(r2, p):
        print(f'OUT_r2 {g(r2, p, 1)} INOUT_r2  {g(r2, p, 2)}')
    if re.search(r3, p):
        print(
            f'IN_r3 {g(r3, p, 1)} INOUT1_r3 {g(r3, p, 2)} OUT_r3 {g(r3, p, 3)} INOUT2_r3 {g(r3, p, 4)}')

Update

For better performance, you can match only once, and get the groups. Take r1 as example:

gs = re.search(r1, p)
if gs:
    print(f'IN_r1 {gs.group(1)} INOUT_r1 {gs.group(2)}')
User contributions licensed under: CC BY-SA
4 People found this is helpful
Advertisement