In <1174683868.851106.205880@e65g2000hsc.googlegroups. com> "Mike" writes:

> >From a script, I want to print all the lines of a file up to a line
> containing the word "END".

Did you really mean "containing" the word END? From your examples below,
I don't think you did.

> Something like:

> # grep -avw -m 1 "END" -f bla.txt

> or

> # cat bla.txt | while read line; do
> echo $line
> if [ "$line" == "END" ]; then break; fi
> done

> ...except a script that actually works. The second one above does not
> work because the "read line" corrupts the lines by removing whitespace.
> I don't know why the first one doesn't work.

The first one doesn't work for two reasons.

First, you're misusing the -f flag. It is not used to specify the file
to be searched.

Second, by using the -m 1 flag, you're telling grep to stop after finding
one match. But you're also using -v, which reverses all matches. So you're
actually telling grep to stop after finding one line that DOES NOT match
"END" -- which, presumably, would be the very first line in the file.

Anyway, here's a sed script that should do what you want:

sed -n '1,/^END$/p' bla.txt

This script will also print the line "END", so if you don't want that,
add this:

| grep -v '^END$'

--
John Gordon A is for Amy, who fell down the stairs
gordon@panix.com B is for Basil, assaulted by bears
-- Edward Gorey, "The Gashlycrumb Tinies"

Mike wrote:
>>From a script, I want to print all the lines of a file up to a line
> containing the word "END".

awk '1;/END/{exit}' file

or

awk '/END/{exit}1' file

depending on whether or not you want the matching line printed.

Ed.

On 2007-03-23, Mike wrote:
>>From a script, I want to print all the lines of a file up to a line
> containing the word "END".
> Something like:
>
> # grep -avw -m 1 "END" -f bla.txt

That will stop at the first line that does NOT contain "END" (the -v
option reverses the test).

> or
>
> # cat bla.txt | while read line; do
> echo $line
> if [ "$line" == "END" ]; then break; fi
> done
>
> ...except a script that actually works. The second one above does
> not work because the "read line" corrupts the lines by removing
> whitespace.

Unless it's a small file, using a while loop to read it is very
inefficient.

> cat bla.txt | while read line; do
> echo $line
> if [ "$line" == "END" ]; then break; fi
> done

First of all, you don't need cat.

Secondly, you are stripping off leading and trailing spaces
because you did not set IFS, and also because you did not quote
$line when you echoed it.

while IFS= read -r line; do
printf "%s\n" "$line"
if [ "$line" = "END" ]; then break; fi
done < bla.txt

If the line containing END has other characters on it, use case
instead of if:

case $line in
*END*) break ;;
esac

> I don't know why the first one doesn't work.

See above.

--
Chris F.A. Johnson, author |
Shell Scripting Recipes: | My code in this post, if any,
A Problem-Solution Approach | is released under the
2005, Apress | GNU General Public Licence

2007-03-23, 14:04(-07), Mike:
>>From a script, I want to print all the lines of a file up to a line
> containing the word "END".
> Something like:
>
> # grep -avw -m 1 "END" -f bla.txt
>
> or
>
> # cat bla.txt | while read line; do
> echo $line
> if [ "$line" == "END" ]; then break; fi
> done
>
>
> ...except a script that actually works. The second one above does not
> work because
> the "read line" corrupts the lines by removing whitespace. I don't
> know why the
> first one doesn't work.
[...]

sed /END/q < bla.txt

Remember a shell is a command line interpreter, pick up the
command most suited for the job. Using a while read loop is the
most alien way to do shell programming.

--
Stéphane