I am using regex to parse the contents of a file using python. I use re.sub to remove comments before parsing, by removing any text from any line beginning with #.

import re

filetext = """
# This is the contents of a file

var: 800  # A var I defined

# This is a comment
    # Indented comment

# Use "=" to set parameters explicitly

param1 = 123
param2 = 456

# Block of code

block1: {
    # var_name = val
    var1 = 7
    var2 = 9
}

# Block of code 2

block2: {
    # var_name = val
    var1 = 7
    var2 = 9
}
"""

scrubbed_text = re.sub(r'#.*', '', filetext, re.M)

print(scrubbed_text)

Output:

var: 800






param1 = 123
param2 = 456



block1: {

    var1 = 7
    var2 = 9
}



block2: {
    # var_name = val
    var1 = 7
    var2 = 9
}

It removes every single comment from the text except for the one inside “block2” at the end. If I remove re.M from the re.sub call, it removes all the comments which is what I want it to do.

This behavior baffles me. Why does re.M multiline mode cause the one match to be skipped?