[SOLVED] grep unmatched [ or [^

1Volt · 02-14-2013, 11:18 AM

Welcome ;

i got a text file contain images urls like

PHP Code:



<img id="thumb-m6qsZ" class="thumb-title" title="" alt="" data-src="http://domain.com/m6qsZs.jpg" data-index="0" />
<img id="thumb-KwvJK" class="thumb-title" title="" alt="" data-src="http://domain.com/KwvJKs.jpg" data-index="1" />

how can i get the urls that's start with http:// and ends with .jpg and save them in another text file which will be every url in a new line . Like this :

Code:

http://domain.com/m6qsZs.jpg
http://domain.com/KwvJKs.jpg

i tried : grep -i "/http.[^"]*\.jpg/" i.txt > o.txt

but see what i get :

user@user ~
$ grep -i "/http.[^"]*\.jpg/" "i.txt" > "o.txt"
> #when i press enter i get a new line says " > " again
>
>
> " #when i put " i get
grep: Unmatched [ or [^

how can i do it ?

jpollard · 02-14-2013, 11:24 AM

Use apostrophe (') instead of the ".

the problem with the $ grep -i "grep -i "/http.[^"]*\.jpg/" "i.txt" > "o.txt" is the string identification.

Start with the first " and go to the next occurrence... and you see a syntax error.

The command should be:
grep -i '/http.[^"]*\.jpg\"' i.txt > o.txt

Now all " characters are contained within an apostrophe quote (') which will not be interpreted by the shell.

1Volt · 02-14-2013, 11:40 AM

Quote:

Originally Posted by jpollard

Use apostrophe (') instead of the ".

the problem with the $ grep -i "grep -i "/http.[^"]*\.jpg/" "i.txt" > "o.txt" is the string identification.

Start with the first " and go to the next occurrence... and you see a syntax error.

The command should be:
grep -i '/http.[^"]*\.jpg\"' i.txt > o.txt

Now all " characters are contained within an apostrophe quote (') which will not be interpreted by the shell.

Thank You , No errors but the file is empty no urls in
could u help me with this
plus a note: that i just copied the command i don't understand it
so please gimme a full solution based on my question .

druuna · 02-14-2013, 11:43 AM

@1Volt: Your code would still print the complete line when a hit is found.

Have a look at grep's -o option:

Code:

grep -io "http://.*\.jpg" i.txt > o.txt

1Volt · 02-14-2013, 11:49 AM

Quote:

Originally Posted by druuna

@1Volt: Your code would still print the complete line when a hit is found.

Have a look at grep's -o option:

Code:

grep -io "http://.*\.jpg" i.txt > o.txt

Thank You , it works but not every url in a new line

druuna · 02-14-2013, 11:51 AM

Quote:

Originally Posted by 1Volt

Thank You , it works but not every url in a new line

Post a _relevant_ example of your infile.

1Volt · 02-14-2013, 12:00 PM

Quote:

Originally Posted by druuna

Post a _relevant_ example of your infile.

Ops i'm very sorry
cuz i was using Cygwin in windows and i opened the file with notepad
and as u know the text format between win&linux is different
with notepad++ the lines comes correctly thank you very much

druuna · 02-14-2013, 12:02 PM

Quote:

Originally Posted by 1Volt

Ops i'm very sorry
cuz i was using Cygwin in windows and i opened the file with notepad
and as u know the text format between win&linux is different
with notepad++ the lines comes correctly thank you very much

No problem

Can you put up the [SOLVED] tag.
- above the first post -> Please Mark this thread as solved if you feel a solution has been provided.
- -or- -
- first post -> Thread Tools -> Mark this thread as solved