This may not be the best or most efficient way to do it, but it does appear to work (written on multiple lines for readability):

Code:

perl -ne 'if ( $flag ) { $flag++; $flag %= 3; } if ( m/error/ ) 
{ $flag = 1; } print $., q{: }, $_ if ( $flag );' <filename>
[download]

Example output:

$ perl -ne 'if ( $flag ) { $flag++; $flag %= 3; } 
if (m/\$line/) { $flag = 1; }; 
print $., q{: }, $_ if ( $flag );' test.pl
15: while ( my $line = <$df> ) {
16:     chomp $line;
17:     next if ( $line =~ m/^\s*$/ );
18:
19:     if ( substr( $line, 3, 8 ) eq q{ACCOUNTA} ) {
20:         print $/, $line, qq{\n};
21:     }
[download]

I started with a single if (m/.../ or $flag) statement, but in testing found that it missed the following line if the search value appeared in consecutive lines. If you want more following context, change the mod value to N+2, where N is the number of lines you want to see including the matching one (i.e., if you want 1 extra line, set to 3, 2 extra to 4, etc.).

Hope that helps.