Shell Programming and Scripting

Hi,
Im doing a script to find the number of lines included inside a file newly. These lines are in between #ifdef FLAG1 and #else or #endif or #else and #endif.
I tried like this,
awk '/#ifdef Flag1/,/#e/{print}' aa.c | wc -l

awk '/#ifndef Flag1/,/#endif/{print}' aa.c | awk '/#else/,/#endif/{print}' | wc -l

But it considers the first #endif that follows #ifdef. In some cases there are some nested #ifdef.
Can some one help me to solve this?

Thanks,
Priya.

Please Post the sample input and the output your expecting.

Regards

Ravi

Sample Input:

#include <>
...
#ifdef FLAG1
#include <FLAG.h>
#ifdef OS
s=splnet()
#else
s=splx()
#endif
#else
s=splx(s)
#endif
#ifndef FLAG1
#include <>
splx(s)
#else
/* Some code */
#endif

Sample Output:

#ifdef FLAG1
#include <FLAG.h>
#ifdef OS
s=splnet()
#else
/* Some code */

Thanks

panyam - We do not really need much more information then what he provided. Create test data and mess with it.

Priya - You can use sed, which is probably a bit easier. You have to use the -n otherwise it will send output twice. You can see a lot of info on sed at the following pages:
Famous Sed One-Liners Explained, Part I - good coders code, great reuse
Sed - UNIX Stream Editor - Cheat Sheet - good coders code, great reuse
http://www.oracle.com/technology/pub...laney_sed.html

Here is a solution for your issue below (followed by an issue you may have)

Code:

-bash-3.2$ cat test.txt
One
Two
Three
Four
Five
Orange
Red
Blue
Black
Yellow
Silver
-bash-3.2$ sed -n '/Four/,/Black/ p' test.txt
Four
Five
Orange
Red
Blue
Black
-bash-3.2$ sed -n '/Four/,/Black/ p' test.txt | wc -l
6

If you have two sets of matches, for instance, going with the above example, you had another 'Four'/'Black' further down the file. It will return both groups, not between the groups, but just the groups themselves.. for instance:

Code:

-bash-3.2$ cat test.txt
One
Two
Three
Four
Five
Orange
Red
Blue
Black
Yellow
Silver
Four
Hundred
Black
-bash-3.2$ sed -n '/Four/,/Black/ p' test.txt
Four
Five
Orange
Red
Blue
Black
Four
Hundred
Black
-bash-3.2$ sed -n '/Four/,/Black/ p' test.txt | wc -l
9

Which may be confusing, but it is correct. Hope this helps!

Sorry!

As Rhije said,

try the following :

Code:

sed '/#ifdef/,/#endif/ p'  input_file | wc -l

sed search for the pattern till the end and the last matched it will consider.

Rhije : I asked priya to post input as well as output expected , since we will get clear clarity ( than textual explanation).

Thanks

Thanks to all of you,
But, the file which i've, has more than one "#ifdef FLAG1 #endif" block. I've to get the total count from every block. Between these blockes there 'll be some lines. I should not include those lines. The last script treats it as a single block

Thanks,
Sella.