When subbing out a general expression, looking to extract just the numbers

Question

Here is the sample text:

initiated to address the deviation to SOP-020583v11.0 Section SOP-016248v2.0 john doe, john doe SOP-020583 fake text, this is all fake

Ideally, the text would look like:

initiated to address the deviation to 020583 Section 016248 john doe, john doe 020583 fake text, this is all fake

Here is the code I have so far:

def dashrepl(matchobj):
    print (type(matchobj))
    return re.findall('[0-9]',matchobj)

re.sub(SOP, dashrepl, long_desc_text[22])

But I'm getting the following error:

TypeError: expected string or buffer

Edit updated content:

long_desc_text[22]

SOP-020583v11.0 Section 8.4.On 17Jan2016 at ATO Site, SOP-016248v2.0 was due for periodic review but the periodic SOP-016248 revision is not tied to any change control records. SOP-020583 tied to a change control record" and notified ID63718 notifiedID22359 of the event. SOP-020583v11.0, fake text fake text

I think your second argument to `findall` is wrong. It should be a string, can you confirm? — Sohaib Farooqi, Dec 13 '17 at 01:59
yes, you're right but now I get this error: `TypeError: sequence item 1: expected string, list found` — madsthaks, Dec 13 '17 at 02:00
`matchobj ` should be a string, as the error message says you are passing a list. — Sohaib Farooqi, Dec 13 '17 at 02:01
I updated my question with the `long_desc_text[22]` and im not sure why a list is being inputted — madsthaks, Dec 13 '17 at 02:06
You should post complete example that I can test, otherwise I can't tell whats wrong. — Sohaib Farooqi, Dec 13 '17 at 02:10
That is the full example. I've included exactly what I'm inputting — madsthaks, Dec 13 '17 at 02:12
`dashrepl` will return a list, you need to iterate over the list and call `re.sub` on each list item — Sohaib Farooqi, Dec 13 '17 at 02:25
ok, so maybe dashrepl is what is incorrect. I'm trying to sub out all of the SOP # with just the number. What would be the best approach? — madsthaks, Dec 13 '17 at 02:28

score 1 · Accepted Answer · answered Dec 13 '17 at 05:22

So, here's my code:

import re

test = "initiated to address the deviation to SOP-020583v11.0 Section SOP-016248v2.0 john doe, john doe SOP-020583 fake text, this is all fake"

regexp = r"SOP-(\d+)(?:v\d+\.\d)?"

test = re.subn(regexp, r"\1", test)

print test[1]

It produces:
"initiated to address the deviation to 020583 Section 016248 john doe, john doe 020583 fake text, this is all fake"

Using the python re function "subn" which finds and replaces all examples of a pattern with the specified string - in this case the first capture group. The "r" in front of the string designates it as a regex object.

For reference I also found this link

Hope this helps.

So I know this isn't what was initially asked but what would I do I wanted to encode the number in a character format. For example, a = 0, b = 1, etc — madsthaks, Dec 13 '17 at 05:46

When subbing out a general expression, looking to extract just the numbers

1 Answers1