Python regex issue

[FreQUENCY]

Hello.

I cant find a way to get the text inside a sentence in python. This is the code 

 

import re
f=open("weather.txt", mode="r", encoding="utf8")
content=f.read()
f.close()

pattern=re.compile("[0-9]") # TO BE CHANGED!
result=pattern.findall(content)

print(result)
print(len(result))

I cant get the proper command for re.compile. For eg. I need to get the program to tell me the items inside the ( ).

 

[C2dan88]

10 minutes ago, FreQUENCY said:

I need to get the program to tell me the items inside the ( )

So in weather.txt you have numbers within () that you want to match ie (123)

Regex would be

\([0-9]+\)

The backslash is used to mean match literal () as they have special meaning in regex.

 

 

[FreQUENCY]

@C2dan88The result i get is [] . Inside the parentheses are items like (24.2 C) or (25.2 C)

[C2dan88]

The regex becomes a bit more complex

\(([0-9]+(\.[0-9]+)?) C\)

This interactive website will help to explain it

https://regex101.com/r/vI6O1w/1

[FreQUENCY]

3 minutes ago, C2dan88 said:

The regex becomes a bit more complex

\(([0-9]+(\.[0-9]+)?) C\)

This interactive website will help to explain it

https://regex101.com/r/vI6O1w/1

for eg  the items inside (-1.7 C) will be output as ('0.7', '.7'), There is something with the decimals.I will look into the link in the meanwhile

[JogerJ]

Just now, FreQUENCY said:

for eg  the items inside (-1.7 C) will be output as ('0.7', '.7'), There is something with the decimals.I will look into the link in the meanwhile

\((-?[0-9]+(?:\.[0-9]+)?) C\)

Use above pattern to include negatives. With this you should get only 1 output. Don't forget these are strings, you'll need to parse them into double (use "float(x)")

[C2dan88]

3 minutes ago, FreQUENCY said:

will be output as ('0.7', '.7')

Ooops, forgot non capture group

Also added optional match for + or -

\(([+-]?[0-9]+(?:\.[0-9]+)?) C\)

[FreQUENCY]

@C2dan88 Works but i need to tweak it to get the C too. The result is

['-1.7', '0.7', '1.4', '1.5', '1.7', '1.9', '24.3', '24.2', '23.8', '23.3', '23.1', '22.8', '22.7', '22.6']
14

Is your way the "correct" way or is it a workaround for this case?

[JogerJ]

4 minutes ago, FreQUENCY said:

@C2dan88 Works but i need to tweak it to get the C too. The result is

['-1.7', '0.7', '1.4', '1.5', '1.7', '1.9', '24.3', '24.2', '23.8', '23.3', '23.1', '22.8', '22.7', '22.6']
14

Is your way the "correct" way or is it a workaround for this case?

Then simply extend the capture group to include the 'C'

Quote

\(([+-]?[0-9]+(?:\.[0-9]+)? C)\)

 

[FreQUENCY]

6 minutes ago, JogerJ said:

Then simply extend the capture group to include the 'C'

 

Works like a charm.I picked a course that teaches regex too and i am so lost.

I need to figure out how to:

• get the word that is previous of the date
• find the words that start with a Capitalised letter and have at least 3 characters in length.

@JogerJCan you please explain me in english how this is formed? So i can understand it ?

Quote

\(([+-]?[0-9]+(?:\.[0-9]+)? C)\)

\(([+-]?[0-9]+(?:\.[0-9]+)? C)\)

 

 

[JogerJ]

Well I don't like doing other people's homework but this should provide a clearer breakdown https://regexr.com/6l8cg
You should at least familiarize yourself with the cheatsheet so you know what you can do with regex, then work your way with it.
I'll provide some hints:

• Getting word previous of date:
  Pattern for the word belongs in first capture group, then follows a non-capturing group with the date
• Capitalised 3+ char words: 
  First letter is uppercase, then followed by at 2 or more lowercase letters (or also uppercase or even numbers, depending on definition of "word")

Give a shot at it and I'll help correct mistakes you might make

[PaulCam]

I just want to drop this here.

 

When learning regular expressions, please, please, please read about "greediness", particular in relation to .* and please, please, please learn to "anchor" your regexps.  This is by far the most common problem with regular expressions in the wild.

 

Example:

Make sure this is our domain:    .*.ourdomain.com

 

Nope.  That's not what that does.  If fact.  These match it.

www.yourdomain.com

lets-hack.ourdomain.com.xackorspz.org

Make sure this is our domain:    ourdomain\.com$

 

Better.  The .* is pointless and the $ forces this now to match only strings which end with the literal ourdomain.com.  It's also simplier.

 

I would recommend, taking any code which uses a regular expression, pulling the regexp bit out into a method and unit testing the sh1t out of it.  Give it a list of things, positive and negative tests.  Send it round the team and make it a challenge.  Can you break it?