x404.co.uk :: View topic - SED/regex help

x404.co.uk http://www.x404.co.uk/forum/

SED/regex help http://www.x404.co.uk/forum/viewtopic.php?f=3&t=5672	Page 1 of 1

Author:

forquare1 [ Fri Jan 22, 2010 2:10 pm ]

Post subject:

SED/regex help

Hi all,

I have a number of files with a line like this:


	Quote: <h2 class="title" style="clear: both"><a xmlns="http://www.w3.org/1999/xhtml" id="id2826402"/>Conclusion: From Custom to Customary Law</h2></div></div><p xmlns="http://www.w3.org/1999/xhtml">We have examined the customs which regulate the ownership and control

I have a statement doing this:


	Code: # Mod <p> and <h> tags to preserve them cat $FILE \| sed 's/<p.>/[p]/g' \| sed 's/<\/p>/[\/p]/g' \| sed 's/<h$[1-3]$$.$></[h\1]</g' \| sed 's/<\/h$[1-3]$$.$*></[\/h\1]</g' > $tmp1

Cat the file, change <> tags around p and /p to [] and tags from h* and /h* to []. However, the statement does this:


	Quote: [h2][p]We have examined the customs which regulate the ownership and control

The third SED expression is matching everything until the last '>' (just before the [p] tag). How do I make it so it only matches up to the FIRST '>' it comes to? I.E. I want this:


	Quote: [h]<a xmlns="http://www.w3.org/1999/xhtml" id="id2826402"/>Conclusion: From Custom to Customary Law[/h2]</div></div>[p]We have examined the customs which regulate the ownership and control

Thanks,
Ben

Author:

forquare1 [ Sat Jan 23, 2010 5:36 pm ]

Post subject:

Re: SED/regex help

Solved

I split the statement into two lines in the end when I cleaned up the script, I also played around and got it to do what I wanted it to:


	Code: # Mod <p> and <h> tags to preserve them cat $FILE \| sed 's/<p.>/[p]/g' \| sed 's/<\/p>/[\/p]/g' > $tmp1 cat $tmp1 \| sed 's/<h$[1-3]$$[^>]$>/[h\1]/g' \| sed 's/<\/h$[1-3]$$.$*>/[\/h\1]/g' > $tmp2

Simple after I re-read my book on pattern matching, I had missed it the first few times I scanned through it.

Now after a few more scripts I've got a legal, up-to-date copy of the book "The Cathedral and the Bazaar" by Eric Raymond, it's a good read

Author:

EddArmitage [ Tue Feb 02, 2010 4:27 pm ]

Post subject:

Re: SED/regex help

My turn:

It's been a bit of a long day and my brain's clearly missing something obvious. I have a file containing lines of input. I want to use grep to select those that end in a forward slash (ultimately I want to select everything but them, but that's a simple flag). What regexp do I need? I thought I've tried everything obvious:


	Code: egrep "/$" < input

Edd

Author:	Nick [ Tue Feb 02, 2010 8:34 pm ]
Post subject:	Re: SED/regex help
Argh my head has just imploded. We had to write Sed in C last year. Absolute hell!!!!!!!

Author:

forquare1 [ Wed Feb 03, 2010 11:51 am ]

Post subject:

Re: SED/regex help

EddArmitage wrote:


	Code: egrep "/$" < input

Edd

I'd do something like:


	Code: egrep \/$ < input

Author:

EddArmitage [ Wed Feb 03, 2010 11:57 am ]

Post subject:

Re: SED/regex help

forquare1 wrote:

EddArmitage wrote:


	Code: egrep "/$" < input

I'd do something like:


	Code: egrep \/$ < input

It worked fine in the end as was, when the input was piped straight in from the previous stage. I swear there must be something installed that uses hamsters as line endings on these damn CSC machines!

Page 1 of 1	All times are UTC
Powered by phpBB® Forum Software © phpBB Group https://www.phpbb.com/