Match and Extract everything between 2 words in a file

[joeferns79]

I had posed a similar topic some time back but I want some additional information from the input file.
The log file is as shown...

[ERROR] [06 Apr 2009 09:45:56] [Trace] [WebContainer : 48] - App Number: 0
[ERROR] [06 Apr 2009 09:45:56] [Trace] [WebContainer : 48] - Response for AE Completion No such RID Found
[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure:ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
    at sun.reflect.GeneratedMethodAccessor186.invoke(Unknown Source)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java(Compiled Code))
    at java.lang.reflect.Method.invoke(Method.java(Compiled Code))
    at com.ibm.rmi.util.ProxyUtil$2.run(ProxyUtil.java(Compiled Code))
    at java.security.AccessController.doPrivileged1(Native Method)
    at java.security.AccessController.doPrivileged(AccessController.java(Compiled Code))
    at com.ibm.rmi.util.ProxyUtil.invokeWithPrivilege(ProxyUtil.java(Compiled Code))
    at com.ibm.CORBA.iiop.ClientDelegate.invoke(ClientDelegate.java(Compiled Code))
    at $Proxy90.invoke(Unknown Source)
nested exception is:
infrastructure:RUN_ID_RUNTIME: A runtime exception occurred: javax.transaction.RollbackException.
        at sun.reflect.GeneratedMethodAccessor186.invoke(Unknown Source)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java(Compiled Code))
    at java.lang.reflect.Method.invoke(Method.java(Compiled Code))
    at com.ibm.rmi.util.ProxyUtil$2.run(ProxyUtil.java(Compiled Code))
    at java.security.AccessController.doPrivileged1(Native Method)
    at java.security.AccessController.doPrivileged(AccessController.java(Compiled Code))
    at com.ibm.rmi.util.ProxyUtil.invokeWithPrivilege(ProxyUtil.java(Compiled Code))
    at com.ibm.CORBA.iiop.ClientDelegate.invoke(ClientDelegate.java(Compiled Code))
    at $Proxy90.invoke(Unknown Source)
 
[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 49] - infrastructure:ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
    at sun.reflect.GeneratedMethodAccessor186.invoke(Unknown Source)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java(Compiled Code))
    at java.lang.reflect.Method.invoke(Method.java(Compiled Code))
    at com.ibm.rmi.util.ProxyUtil$2.run(ProxyUtil.java(Compiled Code))
    at java.security.AccessController.doPrivileged1(Native Method)
 

I want the Perl script to return everything from the file called "App.log" shown above, that matches the line containing "ID_UNHANDL ED", i.e. the line ...

"[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure:ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator." 
upto the next occurence of the string "[ERROR]" 
i.e. it should include the following ...
 
[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure:ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
....
...
...
nested exception is:
infrastructure:RUN_ID_RUNTIME: A runtime exception occurred: 
...
...
...
at $Proxy90.invoke(Unknown Source)
 

which is the last line of that exception. It shouldn't include the next "[ERROR]" line.

I could come up with the following code but it returned only the lines containing the strings mentioned.

while ($line = <LOGFILE>) {
     # chomp $line;
    if ($line =~ /^\[ERROR\](.*?)infrastructure:ID_UNHANDLED:.*$/) {
         push @message, $line;
         $match = 1;
         next;
        }
    if ($match && ($line =~ /^nested exception is:/)) {
        $line = <LOGFILE>;
        push @message, $line;
        $match = 0;
        next;
       }
    }
 

[KevinADC]

its not clear if that is the entire log file or if there is more and you need to continue matching more records. If that is the entire log file then all you really need to do is use the "last" control to break the loop:

while ($line = <LOGFILE>) {
   if ($line =~ /^\[ERROR\](.*?)infrastructure:ID_UNHANDLED:/) {
      push @message, $line;
   }
   elsif ($line =~ /^nested exception is:/) {
      $line = <LOGFILE>;
      push @message, $line;
      last;
   }
}
 

[joeferns79]

That's not the entire log file. There are many more records like the one I showed that need to be matched. It's just that I want the entire stack to be printed and not just that 1 line.

[Kelicula]

I was just skimming the post I may be mistaken but it seems you are using a non-greedy match. Were you want a greedy match.

if ($line =~ /^\[ERROR\](.*?)infrastructure:ID_UNHANDLED:/) {
 

Take out the question mark after the * character.

if ($line =~ /^\[ERROR\](.*)infrastructure:ID_UNHANDLED:/) {
 

Hope it helps!

[joeferns79]

No, it doesn't. It returns the same result ...

[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure: ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
infrastructure: RUN_ID_RUNTIME: A runtime exception occurred: javax.transacti on.RollbackExce ption.
[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 49] - infrastructure: ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
infrastructure: RUN_ID_RUNTIME: A runtime exception occurred: javax.transacti on.RollbackExce ption.

i.e. it returns just the line that the string matches and not the full stack.

I want it to return the following ...

[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure: ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
at sun.reflect.Gen eratedMethodAcc essor186.invoke (Unknown Source)
at sun.reflect.Del egatingMethodAc cessorImpl.invo ke(DelegatingMe thodAccessorImp l.java(Compiled Code))
at java.lang.refle ct.Method.invok e(Method.java(C ompiled Code))
at com.ibm.rmi.uti l.ProxyUtil$2.r un(ProxyUtil.ja va(Compiled Code))
at java.security.A ccessController .doPrivileged1( Native Method)
at java.security.A ccessController .doPrivileged(A ccessController .java(Compiled Code))
at com.ibm.rmi.uti l.ProxyUtil.inv okeWithPrivileg e(ProxyUtil.jav a(Compiled Code))
at com.ibm.CORBA.i iop.ClientDeleg ate.invoke(Clie ntDelegate.java (Compiled Code))
at $Proxy90.invoke (Unknown Source)
nested exception is:
infrastructure: RUN_ID_RUNTIME: A runtime exception occurred: javax.transacti on.RollbackExce ption.
at sun.reflect.Gen eratedMethodAcc essor186.invoke (Unknown Source)
at sun.reflect.Del egatingMethodAc cessorImpl.invo ke(DelegatingMe thodAccessorImp l.java(Compiled Code))
at java.lang.refle ct.Method.invok e(Method.java(C ompiled Code))
at com.ibm.rmi.uti l.ProxyUtil$2.r un(ProxyUtil.ja va(Compiled Code))
at java.security.A ccessController .doPrivileged1( Native Method)
at java.security.A ccessController .doPrivileged(A ccessController .java(Compiled Code))
at com.ibm.rmi.uti l.ProxyUtil.inv okeWithPrivileg e(ProxyUtil.jav a(Compiled Code))
at com.ibm.CORBA.i iop.ClientDeleg ate.invoke(Clie ntDelegate.java (Compiled Code))
at $Proxy90.invoke (Unknown Source)

[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 49] - infrastructure: ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator.
at sun.reflect.Gen eratedMethodAcc essor186.invoke (Unknown Source)
at sun.reflect.Del egatingMethodAc cessorImpl.invo ke(DelegatingMe thodAccessorImp l.java(Compiled Code))
at java.lang.refle ct.Method.invok e(Method.java(C ompiled Code))
at com.ibm.rmi.uti l.ProxyUtil$2.r un(ProxyUtil.ja va(Compiled Code))
at java.security.A ccessController .doPrivileged1( Native Method

[Kelicula]

Hum....

I think your answer lies here?

[joeferns79]

I changed it to look like this ..

while ($line = <LOGFILE>) {
     # chomp $line;
    if ($line =~ /^\[ERROR\](.*?)infrastructure:ID_UNHANDLED: (.*)\[ERROR\]$/ms) {
         push @message, $line;
         $match = 1;
         next;
        }
      }
 

but now it won't return anything at all. Is the RE incorrect?

[Kelicula]

Please use the code tags, so others can see it better and possibly use it to help you reach a solution. (copy&paste). Just an FYI.

I think one problem may be that "$line" is always filtered in "one line at a time" but you are attempting to capture more than one line at a time, but using push to append an array one line at a time. In other words...

No "single" line matches the criteria. You only check for match each line at a time, replacing the value of $line each iteration. Even if there was a match the multiple lines would have been pushed into message one at a time. Instead of what you want, each message element representing the whole match.

You will have to create a result string or something.
Like:

 
my @message;
 
my $file = <LOGFILE>;
 
while($file =~ /^(\[ERROR\].*?infrastructure:ID_UNHANDLED: .*?)\[ERROR\]$/mgs) {
push(@message, $1);
}
 

Untested though.

But the idea is:
Stream the file in continuously pushing the entire match into @message, one "match" at a time.

I removed the capturing parenthesis around the .'s, and added them around the actual match (without ending [ERROR]) So they would get captured in $1.

You may want to match the actual double \n\n's before the next [ERROR] so the next match is successful.

You DO still want the s and m modifiers. I added a g, for global matches.

I have no Perl on this computer, can't test anything...sorr y.

[KevinADC]

I'm still not clear on what it is you need to store in the array from the file.

[joeferns79]

Kelicula,

I changed it as per your suggestions, but it won't return anything at all...

  sub genreport {
 
   my ($logfile, $reportfile) = @_;
   my ($file, @message);
 
   open(LOGFILE, "<$logfile")
      or die "Can't open $logfile: $!";
 
   open (REPORTFILE, ">$reportfile")
      or die "Can't open $reportfile: $!";
 
 
   while($file = <LOGFILE>) {
     # chomp $line;
    if ($file =~ /^(\[ERROR\].*?infrastructure:ID_UNHANDLED: .*?)\[ERROR\]$/mgs) {
         push @message, $1;
 
        }
 
    }
}
 
 

KevinADC - I want to capture the following contents from the log ...

[ERROR] [06 Apr 2009 09:46:02] [Trace] [WebContainer : 20] - infrastructure:ID_UNHANDLED: An un-handled server exception occurred. Please contact your administrator. 
    at sun.reflect.GeneratedMethodAccessor186.invoke(Unknown Source) 
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java(Compiled Code)) 
    at java.lang.reflect.Method.invoke(Method.java(Compiled Code)) 
    at com.ibm.rmi.util.ProxyUtil$2.run(ProxyUtil.java(Compiled Code)) 
    at java.security.AccessController.doPrivileged1(Native Method) 
    at java.security.AccessController.doPrivileged(AccessController.java(Compiled Code)) 
    at com.ibm.rmi.util.ProxyUtil.invokeWithPrivilege(ProxyUtil.java(Compiled Code)) 
    at com.ibm.CORBA.iiop.ClientDelegate.invoke(ClientDelegate.java(Compiled Code)) 
    at $Proxy90.invoke(Unknown Source) 
nested exception is: 
infrastructure:RUN_ID_RUNTIME: A runtime exception occurred: javax.transaction.RollbackException. 
        at sun.reflect.GeneratedMethodAccessor186.invoke(Unknown Source) 
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java(Compiled Code)) 
    at java.lang.reflect.Method.invoke(Method.java(Compiled Code)) 
    at com.ibm.rmi.util.ProxyUtil$2.run(ProxyUtil.java(Compiled Code)) 
    at java.security.AccessController.doPrivileged1(Native Method) 
    at java.security.AccessController.doPrivileged(AccessController.java(Compiled Code)) 
    at com.ibm.rmi.util.ProxyUtil.invokeWithPrivilege(ProxyUtil.java(Compiled Code)) 
    at com.ibm.CORBA.iiop.ClientDelegate.invoke(ClientDelegate.java(Compiled Code)) 
    at $Proxy90.invoke(Unknown Source) 
 

The above belongs to 1 un-handled exception. But the log file contains many un-handled exceptions, so I want all these returned.