Regex Help

[DoubleMcLovin]

I am looking for a regex pattern that will match "bobby.01.yaddy.yaddy.0101010.avi" but not match "example.101.02.The.example.Returns.Part.1.2.Moonsong.avi" by looking at the digits. It needs to match exactly 2 characters and keep the surrounding text using a pattern similar to

(?i)(.+)\.(\d{2})\.?(.*)(\..+)\b

It needs to be able to "see" that there are exactly two digits inside a decimal and they are not directly followed by a possible decimal (\.?) and then more digits up to 4 (\d{1,4}) but I cannot get the negative lookaround to match without excluding all matches. Sorry if this is a bit confusing.

[Melba23]

DoubleMcLovin,

exactly two digits inside a decimal and they are not directly followed by a possible decimal and then more digits up to 4

Try this - it works on your 2 examples:

(\.\d{2}\.)(?!\d{1,4})

(\.\d{2}\.) - look for "point - 2 digits - point"

(?!\d{4}) - not followed by "1 to 4 digits"

Any good?

M23

[GEOSoft]

It is a bit confusing but this might be what you are looking for

#include<array.au3>;; For displaying results only
$sStr = "bobby.01.yaddy.yaddy.0101010.avi" & @CRLF
$sStr &= "example.101.02.The.example.Returns.Part.1.2.Moonsong.avi"

$aMatches = StringRegExp($sStr, "([a-z]+\.\d{2}[.\D].+)", 3)
If NOT @Error Then
    _ArrayDisplay($aMatches)
Else
    MsgBox(4096, "Error", "No matches to the pattern")
EndIf

When posting for an SRE response please put your examples on separate lines in the post. It makes it much easier to see what you are doing that way. In this case, if I read it correctly, it should have been like this for readability

Match

bobby.01.yaddy.yaddy.0101010.avi

Don't Match

example.101.02.The.example.Returns.Part.1.2.Moonsong.avi

[DoubleMcLovin]

Sorry, I should have been more specific about those dots. This is to help sort my personal media files, the format of each file is different but in general follows the scheme:

"showName.season#.ep#.episodeTitle.ext"

The regex is broken into parts, this part is to catch when there is no 'season#' present. The trouble is, sometimes (assume randomly) 'episodeTitle' does not exist either. So the string could be either "showName.ep#.episodeTitle.ext" or "showName.ep#.ext". We need this StringRegExp() to capture into an array

[1]showName

[2]ep#

[3](possible)episodeTitle

[4].ext

So my idea using your example is:

(?i)(.+)\.(\d{2})\.?(?!(\d{1,4}))\.?(.*)(\..+)\b

This doesn't work when I run it through the regex tester since it captures all the sample strings =\

[GEOSoft]

@M23

You are getting much faster at replying to SRE topics. Good to see.

[Melba23]

George,

Only way to learn is to try!

M23

[jchd]

M23,

Try and try and yet try again ... and hence to loose hair faster (look at George's head...)

[GEOSoft]

Now I'm getting French insults too when the Aussie ones were enough. I pulled all my hair out just from being on these forums for too long.

[jchd]

Yeah, you're right: responsabilities are more probably 80% forum 20% PCRE.

BTW I didn't drop work on PCRE but keep being distracting by last year' emergencies.

[GEOSoft]

I hope you don't drop that project. It can be important. I had a couple of things I was going to ask about in a PM but I forgot what they were. Next we have Valik coming along with "bald AND senile".

[DoubleMcLovin]

Here is the string I came up with to match this the way I need to. It looks ugly, but gets the job done.

(?i)(?=(?:\b\D*\d{2}\D{2}.*\b))(.+)\.(\d{2}(?!(?:\d+|.*\d+\D+\d+)))\.?(.*)(\..+)\b

[GEOSoft]

Whatever works. You don't need the (?i) since you are not using litteral characters in the expression.