Make SDATA cumulative and improve performance

Each `SDATA` extractor will add to the SDATA hash, instead of having the
"last" collector win.  This shouldn't affect much, but would allow for
logs containing JSON and lazy K/V pairs to expose both via SDATA.

The order of the SDATA collectors determines precedence. The highest
priority is RFC5424 Structured Data, followed by JSON, followed by
"lazy" K/V data.

Also:
* Bump everything to `use v5.16` as that's the minimum Perl version
* Add KV and JSON detection to the benchmark script for flame graphs on those parts
* Fix cpanfile generation
* Fix Github actions
* Add `normalize_test_result()` function for handling elements in the
  results that might change based on the environment
* Remove named captures. Named captures do make things more readable,
  but they are dramatically slower. Removing these from the library has
  increased the parse speed significantly.
* Add `psl_enable_sdata()` function to adjust enable full structured
  data parse mode.
* Don't create the empty hash as the copy takes more time than not
  reallocating it. Adjust tests to remove the empty keys.

.github/workflows/test.yml

@@ -1,12 +1,9 @@
-# This is a basic workflow to help you get started with Actions
-
 name: Test Matrix
 
 # Controls when the workflow will run
 on:
-  # Triggers the workflow on push or pull request events but only for the "master" branch
   push:
-    branches: [ "master" ]
+    branches: [ "master", "develop" ]
   pull_request:
     branches: [ "master" ]
 
@@ -20,7 +17,7 @@ jobs:
     strategy:
       matrix:
         os: ['ubuntu-latest']
-        perl: [ '5.40', '5.38', '5.36', '5.34', '5.32', '5.30', '5.28', '5.26', '5.24', '5.22', '5.20', '5.18', '5.16', '5.14' ]
+        perl: [ '5.40', '5.38', '5.36', '5.34', '5.32', '5.30', '5.28', '5.26', '5.24', '5.22', '5.20', '5.18', '5.16' ]
     name: Perl ${{ matrix.perl }} on ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v3

.gitignore

@@ -12,3 +12,4 @@ Makefile.old
 .cvsignore
 pm_to_blib
 MYMETA.*
+nytprof*

README.mkdn

@@ -4,7 +4,7 @@ Parse::Syslog::Line - Simple syslog line parser
 
 # VERSION
 
-version 6.0
+version 6.1
 
 # SYNOPSIS
 
@@ -251,14 +251,6 @@ The changes to the API and fields returned are as follows:
     - `NormalizeToUTC` is **deprecated**, every log now returns `datetime_utc`
     - `OutputTimeZone` is **deprecated**, use `TimeMomentFormatString`
 - **Field Changes**
-    - `date_raw`
-
-        **Removed** from the fields returned, use `datetime_raw`.
-
-    - `datetime_obj`
-
-        **Removed** from the fields returned as we dropped support for [DateTime](https://metacpan.org/pod/DateTime).
-
     - `datetime_utc`
 
         Present in every document, use this for portability.
@@ -455,6 +447,10 @@ whitespace will be assumed to be a continuation of the previous line.
 
 It is not exported by default.
 
+## psl\_enable\_sdata
+
+Call this to turn on all the Structured Data Parsing Options
+
 ## preamble\_priority
 
 Takes the Integer portion of the syslog messsage and returns

benchmarks/00-defaults.pl

@@ -1,10 +1,11 @@
 #!perl
 
-use v5.14;
+use v5.16;
 use warnings;
 use Dumbbench;
 use Const::Fast;
 use Parse::Syslog::Line;
+psl_enable_sdata();
 
 use FindBin;
 use lib "$FindBin::Bin/../t/lib";
@@ -21,7 +22,7 @@
     my ($test) = @_;
     @copy = @msgs unless @copy and $last ne $test;
     $last=$test;
-    parse_syslog_line(shift @copy);
+    my $doc = parse_syslog_line(shift @copy);
 };
 
 my $bench = Dumbbench->new(

benchmarks/01-parse.pl

@@ -1,6 +1,6 @@
 #!perl
 
-use v5.14;
+use v5.16;
 use warnings;
 use Benchmark qw/timethese cmpthese/;
 use Const::Fast;

benchmarks/02-benchsmarter.pl

@@ -1,6 +1,6 @@
 #!perl
 
-use v5.14;
+use v5.16;
 use warnings;
 use Const::Fast;
 use Dumbbench;

benchmarks/xt-date-parsing-iso.pl

@@ -1,6 +1,6 @@
 #!/usr/bin/perl
 #
-use v5.14;
+use v5.16;
 use warnings;
 
 use Benchmark qw(cmpthese);

benchmarks/xt-date-parsing-legacy.pl

@@ -1,6 +1,6 @@
 #!/usr/bin/perl
 #
-use v5.14;
+use v5.16;
 use warnings;
 
 use Benchmark qw(cmpthese);

bin/parse-syslog-line.pl

@@ -1,14 +1,14 @@
 #!perl
 # PODNAME: parse-syslog-line.pl
 # ABSTRACT: Parse a syslog message and display the structured data
-use v5.14;
+use v5.16;
 use warnings;
 
 use Data::Printer;
 use Getopt::Long::Descriptive;
 use Pod::Usage;
 use JSON::MaybeXS;
-use Parse::Syslog::Line qw( parse_syslog_line );
+use Parse::Syslog::Line;
 use YAML::XS;
 
 my $enc;
@@ -94,9 +94,9 @@ =head1 EXAMPLES
 =cut
 
 # Configure
+psl_enable_sdata() if $opt->sdata;
 $Parse::Syslog::Line::PruneEmpty = !$opt->empty;
 $Parse::Syslog::Line::PruneRaw   = !$opt->raw;
-$Parse::Syslog::Line::AutoDetectJSON = $Parse::Syslog::Line::AutoDetectKeyValues = $opt->sdata;
 $Parse::Syslog::Line::RFC5424StructuredData = !$opt->disable_rfc_sdata;
 $Parse::Syslog::Line::RFC5424StructuredDataStrict = $opt->strict_rfc_sdata;
 

cpanfile

@@ -7,6 +7,7 @@ requires "Data::Printer" => "0";
 requires "English" => "0";
 requires "Exporter" => "0";
 requires "Getopt::Long::Descriptive" => "0";
+requires "Hash::Merge::Simple" => "0";
 requires "JSON::MaybeXS" => "0";
 requires "Module::Load" => "0";
 requires "Module::Loaded" => "0";
@@ -15,7 +16,7 @@ requires "Pod::Usage" => "0";
 requires "Ref::Util" => "0";
 requires "Time::Moment" => "0";
 requires "YAML::XS" => "0";
-requires "perl" => "5.014";
+requires "perl" => "v5.16.0";
 requires "warnings" => "0";
 recommends "Cpanel::JSON::XS" => "0";
 
@@ -34,16 +35,15 @@ on 'test' => sub {
   requires "Test::Deep" => "0";
   requires "Test::MockTime" => "0";
   requires "Test::More" => "0";
-  requires "YAML" => "0";
   requires "bignum" => "0";
   requires "lib" => "0";
-  requires "perl" => "5.014";
+  requires "perl" => "v5.16.0";
   requires "strict" => "0";
 };
 
 on 'configure' => sub {
   requires "ExtUtils::MakeMaker" => "0";
-  requires "perl" => "v5.14.0";
+  requires "perl" => "v5.16.0";
 };
 
 on 'develop' => sub {

dist.ini

@@ -12,6 +12,7 @@ exclude_filename = cpanfile
 exclude_filename = dist.ini
 exclude_filename = weaver.ini
 exclude_filename = weaver-ci.ini
+exclude_filename = cpanfile
 [ExecDir]
 dir = bin
 [PruneCruft]