Shell Programming and Scripting

I need to create a script called 'find_data' that will search a file by passing two parameters to it . I'm pretty good with scripting but haven't really gotten into sed and awk that much (and no perl whatsoever). I'm really at a loss and need to get something working in the next couple days (yes, we still do work on weekends).

The script should work like this.

find_data <parameter 1> <parameter 2>

The first parameter would be the data before the colon (':').
The second parameter would be the data after the colon (':')
If both are found the data between the preceeding BEGINJOB and ENDJOB (inclusive) should be displayed to stdout.

Examples using the data:

1 find_data start 01:15 would find, and display, the first set of data.
2 find_data days tu would find, and display, the last set of data.
3 find_data days all would find, and display, the 2nd, 3rd and 4th sets of data.

I really appreciate your help with this. It's a great site.



Sample data (only five shown but there are several hundred):

BEGINJOB---------------------------
insert_job: cir_cdezp07b job_type: c
command: /apps/usr/circ/bin/cdezp07b.cmd
days_of_week: all
start_times: "01:15"
description: "Mass Zip Code Update"
ENDJOB
BEGINJOB---------------------------
insert_job: cir_ccsds03b job_type: c
command: /apps/usr/circ/bin/ccsds03b.cmd
days_of_week: all
start_times: "02:00"
description: "Deletes Info. from Dispatch_table"
ENDJOB
BEGINJOB---------------------------
insert_job: cir_ccsda78r job_type: c
command: /apps/usr/circ/bin/ccsda78r.cmd REP
days_of_week: all
start_times: "03:00"
description: "Daily Reports"
ENDJOB
BEGINJOB---------------------------
insert_job: cir_ccsda95i job_type: c
command: /apps/usr/circ/bin/ccsda95i.cmd 6 1000 0 25
days_of_week: all
start_times: "04:00"
condition: s (cir_ccbdd01b_daily) & s (cir_ccbdd01b_sat) & s (cir_ccbdd01b_sun)
description: "Archive job, runs M-SU"
ENDJOB
BEGINJOB---------------------------
insert_job: cir_cdtsm01i job_type: c
command: /apps/usr/circ/bin/cdtsm01i.cmd
days_of_week: tu,we,th,fr,sa
start_times: "04:00"
description: "Templated Solo Mail - aoets20b.tx to aoets20b_txt"
ENDJOB

frankly you lost me, I am sorry if I am not able to understand this, but I dont understand the above scenario.
The output is a part of file ? (if yes, is this file same as input file ?)
what is parameter and what is data, where is data, in input file ?
you are finding a file, or finding for something in a file ?

You could split the file into job arrays

Code:

open (FILE,"/path/to/file");
@LINES=<FILE>;
close(FILE);

$count = 0;

foreach $_ (@LINES) {
  $curarray = "array".$count;
  if ($_ =~ /ENDJOB/) {$count++}
  push (@$curarray,$_);
}

then just check each array for what you are looking for and print it out if it matches.

That code will give you:
@array0
@array1
and so on, and each array will be the jobs from the BEGIN to the END lines.

$array2[3] would be
days_of_week: all
for example

Code:

awk 'BEGIN{RS="(BEGIN|END)JOB-*"}$0~(x ".*:.*" y)' x=days y=tu file1

...gives...

Code:

insert_job: cir_cdtsm01i job_type: c
command: /apps/usr/circ/bin/cdtsm01i.cmd
days_of_week: tu,we,th,fr,sa
start_times: "04:00"
description: "Templated Solo Mail - aoets20b.tx to aoets20b_txt"

Code:

awk -v Before="$1" -v After="$2" '

BEGIN {
   FS = ":";
   Before_Pattern = ".*"  Before ".*";
   After_Pattern  = ":.*" After  ".*";
}

/^BEGINJOB/,/^ENDJOB/ {
   Memo[++Line_Count] = $0;
   if ($1 ~  Before_Pattern) debug "Before match";
   if ($1 ~  After_Pattern)  debug "After  match";
   if ($1 ~  Before_Pattern && $0 ~ After_Pattern)
      Print_Datas = 1;
}

/^ENDJOB/ {
   if (Print_Datas) {
      for (i=1; i<=Line_Count; i++)
         print Memo[i];
   }
   Line_Count = 0;
   Print_Datas = 0;
}

' input.txt

Code:

$ find_data days tu
BEGINJOB---------------------------
insert_job: cir_cdtsm01i job_type: c
command: /apps/usr/circ/bin/cdtsm01i.cmd
days_of_week: tu,we,th,fr,sa
start_times: "04:00"
description: "Templated Solo Mail - aoets20b.tx to aoets20b_txt"
ENDJOB

Thanks aigles but I'm getting errors when I execute this. Here's the script. And the errors I'm getting (btw, I'm running this on a Solaris server).

Script:

#! /bin/ksh
# find_script

awk -v Before="$1" -v After="$2" '

BEGIN {
FS = ":";
Before_Pattern = ".*" Before ".*";
After_Pattern = ":.*" After ".*";
}

/^BEGINJOB/,/^ENDJOB/ {
Memo[++Line_Count] = $0;
if ($1 ~ Before_Pattern) debug "Before match";
if ($1 ~ After_Pattern) debug "After match";
if ($1 ~ Before_Pattern && $0 ~ After_Pattern)
Print_Datas = 1;
}

/^ENDJOB/ {
if (Print_Datas) {
for (i=1; i<=Line_Count; i++)
print Memo[i];
}
Line_Count = 0;
Print_Datas = 0;
}

' input.txt

exit 0

Errors:

% find_script days tu
awk: syntax error near line 1
awk: bailing out near line 1
%

If I change this: awk -f circjobs1 -v Before="$1" -v After="$2" '
the errors are:

% find_script days tu
awk: syntax error near line 3
awk: bailing out near line 3
awk: newline in string near line 317
awk: newline in string near line 3844
awk: newline in string near line 4758
awk: newline in string near line 5411
awk: newline in string near line 5816
awk: newline in string near line 9824
awk: newline in string near line 10337
awk: newline in string near line 10517
awk: newline in string near line 11808
%