On 06/29/2016 06:03 AM, Eric de Hont wrote:
Op 29-06-16 om 06:35 schreef Danny Wong:
Hi Perl GURUs,
I have a json file that needs parsing.
Here is a typical string I’m searching for. I want to delete
everything but the last 2 character “],”.
],
[
"ansible",
"2.1.0.0-1ppa~trusty",
false
],
Here is what I tried:
I slurp the whole file into a variable.
my $SAVE = $/;
my $WHOLE_JSON_FILE = `cat ${JSON_FILE}`;
$/ = $SAVE;
while($WHOLE_JSON_FILE !~ /.*?(\s+\]\,\s+\[\s+\"ansible\".*?)\]\,?/gs)
{
print "\$1 is $1";
}
The print statement is printing out the “matching string” but how do
I remove that section of string from the slurp $WHOLE_JSON_FILE
variable which contains the entire file content?
Hi Danny,
JSON is a strictly defined data format. Modifying such a file with
regular expressions will sooner or later
break the structure of your config file in a nasty way. The same goes
for XML, HTML en the like.
Therefore the real guru's have created modules to help us keep these
files sane.
sub slurp_file {
my $file = shift;
local $/;
open my $fh, '<', $file or die "Can't open $_: $!\n";
$_ is not set anywhere. you likely meant to use $file
<$fh>;
}
sub write_file {
my $file = shift;
open my $fh, '>', $file or die "Can't open $_: $!\n";
same bug.
print $fh @_ ;
}
since you are correct about modules being already there, why do you
write your own versions of
slurp_file and write_file? the module File::Slurp has those functions
which are stable, fast and debugged.
uri
