Regex: How do I match...

[PaulM]

...all but the first x chars in a string of arbitrary length?

Apologies if this is the wrong forum; I wasn't sure the best place to post about Regex.

Background: I am new to Regex for pattern matching.
I have started using Yahoo Pipes to manipulate RSS feeds. Yahoo Pipes includes a regex gadget to do string manipulation, but no "stateful" programming.

My goal is to truncate a string of arbitrary length to a fixed length (of, say, the first 15 chars), to produce something like:
"Man bites dog..." from a starting string that looks like "Man bites dog in dark alley."
Counting from the end of the string and doing it from that direction is a problem, because each string will be of a different length.
The Yahoo Pipes regex parser only allows the replacement of static text (including deletions) for matched text.

So I can find the first 10 chars as ^.{10} but I'm struggling with negation and other tricks in order to match all but the first x chars!

Is this an easy problem for a resident regex expert?
Thanks in advance to anyone who can crack this for me!

Paul M.

[bartonc]

...all but the first x chars in a string of arbitrary length?

Apologies if this is the wrong forum; I wasn't sure the best place to post about Regex.

Background: I am new to Regex for pattern matching.
I have started using Yahoo Pipes to manipulate RSS feeds. Yahoo Pipes includes a regex gadget to do string manipulation, but no "stateful" programming.

My goal is to truncate a string of arbitrary length to a fixed length (of, say, the first 15 chars), to produce something like:
"Man bites dog..." from a starting string that looks like "Man bites dog in dark alley."
Counting from the end of the string and doing it from that direction is a problem, because each string will be of a different length.
The Yahoo Pipes regex parser only allows the replacement of static text (including deletions) for matched text.

So I can find the first 10 chars as ^.{10} but I'm struggling with negation and other tricks in order to match all but the first x chars!

Is this an easy problem for a resident regex expert?
Thanks in advance to anyone who can crack this for me!

Paul M.

I just go a great book: Mastering Regular Expressions. I'll look and see if you can match the end of a string that has no newline at the end.

PS Sorry your not a Pythoneer.

[ghostdog74]

can you provide samples of input strings and your desired output.? I have problem understanding your requirement. At first i thought you could simply get the first x chars off , like

>>> s = "Man bites dog in dark alley"
>>> s[15:]
'n dark alley'
 

then use whatever regexp you need on this result. but i suspect its not so simple , right?

[PaulM]

can you provide samples of input strings and your desired output.? I have problem understanding your requirement. At first i thought you could simply get the first x chars off , like

>>> s = "Man bites dog in dark alley"
>>> s[15:]
'n dark alley'
 

then use whatever regexp you need on this result. but i suspect its not so simple , right?

OK, real examples include the following RSS feed titles from boardgamenews.com:

Test Your Gaming Knowledge
More About Chicago Poker from Bruno Faidutti
Gone Cardboard News: BattleLore Call to Arms Rules Available for Downloading

The desired output (assuming 16 chars kept) would be:

Test Your Gaming...
More About Chica...
Gone Cardboard N...

(BTW, don't worry about the '...'; this can be easily appended to the resulting string once everything but the first 16 chars is deleted.)

The trouble is that the length of each string is arbitrary, and the regex widget does not allow me to count the chars and then work from this; ie., it is effectively stateless.
So no obvious scripting is possible, such as s=right(s,16,(length(s)-16)), or s=substr(s,16,$) etc. But basically, I want to achieve the same effect in a single regex statement.

[ghostdog74]

well, you don't need regular expression in this case as its overkill. If you have those sample data in a file, called file, you can do this:

for line in open("file"):
    print line[:16]
 

output:

# ./test.py
Test Your Gaming
More About Chica
Gone Cardboard N
 

If you are inclined to regular expression,

import re
for line in open("file"):
    print re.findall("^.{16}",line)[0]
 

output:

# ./test.py
Test Your Gaming
More About Chica
Gone Cardboard N
 

[PaulM]

Thanks GhostDog, I appreciate the reply, although that doesn't seem solve the specific issue I've had.
The challenge is that Yahoo Pipes does not provide the depth of functionality that any other scripting language might (which means I can't actually carry out the simple function you describe).

However, in the meantime I have discovered more about the regex match-replace widget that is not (yet) documented, and this has in fact allowed me to solve my problem.

In particular, it seems one can parenthesise tokens in the match query and replace them with $1 in the replace field.
So match (^.{16}).*$ replace $1...
for string: "Gone Cardboard News: BattleLore Call to Arms Rules Available for Downloading"
returns:
"Gone Cardboard N..."
which is exactly the result I was after, and perhaps even does the same thing internally that you describe (it was just non-intuitive and undocumented).

Furthermore, match (^.{25}).*(\([^(]*\)$) replace $1... $2
for string: "Gone Cardboard News: BattleLore Call to Arms Rules Available for Downloading (Boardgamenews, 2007-4-22)"
returns:
"Gone Cardboard News: Batt... (Boardgamenews, 2007-4-22)"

...which is even better.

Thanks again,
Paul M.

[randomdeanna]

Hey Paul!

I'm a Yahoo Pipes person as well, found your posts while googling for regex answers on the same problem you had. Couldn't get your string to work for me; any chance you can post your Pipe link here, and I could poke around the source?

thanks much in advance...!
d