Hi there,
I am looking for a shell script which recursively searches all the files under all the directories for a pattern specified in the script. For e.g., i am looking forward to search for the file names which contains numbers of the form 001*****.

Thanks in advance,
Naik

Try this bit of code:
Not thinking too clearly though. Other guys probably will give you a better solution.

I'd probably do:

find <dir_you_want_search_> -type f -exec grep -lE '^001\d{5}$' {} \;

saves on some typing. =)

Okay I like the diff command.

diff arlog arlog.NWS | tr -d '\t ' | awk -F \| ' { print $1$2 } ' >> NEWERRORS

Next:
I would like to add: NEW TIME STAMP AT: `date` in the new file. Where can I put this portion of text. Let me explain. Since I will be searching the NEWERRORS log and it will look something like:

<NoLBAHLingotitdirectory
<****Removingarmonfilein/export/home/fininter/load
</loader/findev/log/FriFeb1716:30:21EST2006
<****Creatingarmonfilein/export/home/fininter/load
<*********LOCKBOX-GETBANKFILEFROMendeavour
<lbhap_get_bank_fileY
<CheckforLBHAPfileinendeavourdirectory
<NoLBHAPingotitdirectory
<lbahl_get_bank_fileY
<CheckforLBAHLfileinendeavourdirectory
<NoLBAHLingotitdirectory
<****Removingarmonfilein/export/home/fininter/load
</loader/findev/log/FriFeb1716:32:21EST2006
<****Creatingarmonfilein/export/home/fininter/load
<*********LOCKBOX-GETBANKFILEFROMendeavour
<lbhap_get_bank_fileY
<CheckforLBHAPfileinendeavourdirectory
<NoLBHAPingotitdirectory
<lbahl_get_bank_fileY
<CheckforLBAHLfileinendeavourdirectory
<NoLBAHLingotitdirectory
<****Removingarmonfilein/export/home/fininter/load
---

>NEWTIMESTAMPAT:WedFeb1515:25:12EST2006

>/appl/finance/findev/ar/bin/arlog.NWS:NEWTIMESTAMPAT:WedFeb1515:25:12EST2006
>NEWTIMESTAMPAT:WedFeb1515:48:16EST2006

<data> 0403-031 The fork function failed. There is not enough memory available.
<data>
<data>ftp: connect: Connection timed out.
<data>
<data>
>NEWTIMESTAMPAT:WedFeb1519:04:07EST2006
<data>
<data>ORA-06508: PL/SQL: could not find program unit being called.
<data>
<data>
>NEWTIMESTAMPAT:WedFeb1520:05:38EST2006findev:/appl/finance/findev/ar/bin: F \| ' { print $1$2 } ' >> errors <
findev:/appl/finance/findev/ar/bin: vi errors
"errors" 548 lines, 19849 characters
"errors" 548 lines, 19849 characters
1,541c1,5


Of course it doesn't look like this as of yet but I would like it to get there.

I am trying to put this script together and I am having grave problems doing so, this is the structure:

#!/bin/sh
# Every 6 hours

# Get all lines from the arlog based on the last PREVIOUS_TIME date and time.

PREVIOUS_TIME='DATE AND TIME: `date`' in the arlog

NEW_TIME='TODAYS DATE AND TIME: `date` ' in the arlog

#LIST OF ERRORS THAT WILL BE SEARCHED.

ERRMSG='CHECKING FOR ERRORS IN ARLOG: ERROR 9999'
ERRMSG1='0403-031 The fork function failed. There is not enough memory available.'
ERRMSG2='Login failed.'
ERRMSG3='ftp: connect: Connection timed out.'
ERRMSG4='ftp: connect: Connection refused.'
ERRMSG5='endeavour: unknown host or invalid literal address.'
ERRMSG6='ftp: Unknown host diamprodf0'
ERRMSG7='ORA-06508: PL/SQL: could not find program unit being called.'
ERRMSG8='ORA-01034: ORACLE not available.'
ERRMSG9='ORA-27101: shared memory realm does not exist.'
ERRMSG10='ORA-01005: null password given; logon denied.'
ERRMSG11='SP2-0157: unable to CONNECT to ORACLE after 3 attempts, exiting SQL.'
ERRMSG12='ORA-04068: existing state of packages has been discarded.'
ERRMSG13='ORA-04063: package body "FININTER.HAP_CUSTOMERS_INTERFACE" has errors.'
ERRMSG14='ORA-06512: IN LINE ERROR.'
ERRMSG15='ORA-01005: null password given.'

grep 'DATE STAMP: $PREVIOUS_TIME ' arlog >> TEMP_ARLOG.log

if $NEW_TIME in the arlog NOT EQUAL $PREVIOUS_TIME in arlog
then
egrep "(${ERRMSG} | ${ERRMSG1} | ${ERRMSG2} |........) from arlog.
if NO ERRORS EXIST
then
put $NEW_TIME 'No Errors To Date and time" >> arlog
else
mailx -s '$ERRORS `date` to [FMT] with message describing error or errors.
grep 'LAST TIME TRANSMISSION:' $PREVIOUS_TIME >> arlog
fi
fi

Can anyone help me? 12green