From: Peter Rabbitson <ribasushi@cpan.org>
Date: Wed, 20 Feb 2013 10:12:35 +0000 (+0100)
Subject: Optimize the HRI-direct collapser even more
X-Git-Tag: v0.08241~3
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=7596ddca6b7b39d56f7d5526be65855d80c81c4b;hp=bdbd2ae8a0625e196ae820f85427d13629322c96;p=dbsrgits%2FDBIx-Class.git

Optimize the HRI-direct collapser even more

Since we know that we will never output null-branches in the case of
HRI, we can wrap the related assemblers in a definedness condition to
never fire on a null branch, and skip entirely the calculation of
non-NULL %cur_row_ids.
---

diff --git a/Changes b/Changes
index 783f789..3915b79 100644
--- a/Changes
+++ b/Changes
@@ -4,6 +4,8 @@ Revision history for DBIx::Class
         - Revert to passing the original (pre-0.08240) arguments to
           inflate_result() and remove the warning about ResultClass
           inheritance.
+        - Optimize the generated rowparsers even more - no user-visible
+          changes.
 
 0.08240-TRIAL (EXPERIMENTAL BETA RELEASE) 2013-02-14 05:56 (UTC)
     * New Features / Changes
diff --git a/lib/DBIx/Class/ResultSource/RowParser/Util.pm b/lib/DBIx/Class/ResultSource/RowParser/Util.pm
index 48d2b08..4d833d3 100644
--- a/lib/DBIx/Class/ResultSource/RowParser/Util.pm
+++ b/lib/DBIx/Class/ResultSource/RowParser/Util.pm
@@ -127,11 +127,12 @@ sub assemble_collapsing_parser {
 
     my $virtual_column_idx = (scalar keys %{$args->{val_index}} ) + 1;
 
-    $top_node_key_assembler = sprintf '$cur_row_ids{%d} = (%s);',
-      $virtual_column_idx,
-      "\n" . join( "\n  or\n", @path_parts, qq{"\0\$rows_pos\0"} );
+    $top_node_key = "{'\xFF__IDVALPOS__${virtual_column_idx}__\xFF'}";
 
-    $top_node_key = sprintf '{$cur_row_ids{%d}}', $virtual_column_idx;
+    $top_node_key_assembler = sprintf "'\xFF__IDVALPOS__%d__\xFF' = (%s);",
+      $virtual_column_idx,
+      "\n" . join( "\n  or\n", @path_parts, qq{"\0\$rows_pos\0"} )
+    ;
 
     $args->{collapse_map} = {
       %{$args->{collapse_map}},
@@ -145,11 +146,18 @@ sub assemble_collapsing_parser {
 
   my ($data_assemblers, $stats) = __visit_infmap_collapse ($args);
 
-  my $list_of_idcols = join(', ', sort { $a <=> $b } keys %{ $stats->{idcols_seen} } );
+  my @idcol_args = $args->{hri_style} ? ('', '') : (
+    '%cur_row_ids, ', # only declare the variable if we'll use it
+
+    sprintf( <<'EOS', join ', ', sort { $a <=> $b } keys %{ $stats->{idcols_seen} } ),
+  $cur_row_ids{$_} = defined($cur_row_data->[$_]) ? $cur_row_data->[$_] : "\0NULL\xFF$rows_pos\xFF$_\0"
+    for (%s);
+EOS
+  );
 
-  my $parser_src = sprintf (<<'EOS', $list_of_idcols, $top_node_key, $top_node_key_assembler||'', join( "\n", @{$data_assemblers||[]} ) );
+  my $parser_src = sprintf (<<'EOS', @idcol_args, $top_node_key_assembler||'', $top_node_key, join( "\n", @{$data_assemblers||[]} ) );
 ### BEGIN LITERAL STRING EVAL
-  my ($rows_pos, $result_pos, $cur_row_data, %%cur_row_ids, @collapse_idx, $is_new_res) = (0,0);
+  my ($rows_pos, $result_pos, $cur_row_data,%1$s @collapse_idx, $is_new_res) = (0,0);
   # this loop is a bit arcane - the rationale is that the passed in
   # $_[0] will either have only one row (->next) or will have all
   # rows already pulled in (->all and/or unordered). Given that the
@@ -161,24 +169,25 @@ sub assemble_collapsing_parser {
       ||
     ($_[1] and $_[1]->())
   ) {
+    # this code exists only when we are *not* assembling direct to HRI
+    #
     # due to left joins some of the ids may be NULL/undef, and
     # won't play well when used as hash lookups
     # we also need to differentiate NULLs on per-row/per-col basis
     # (otherwise folding of optional 1:1s will be greatly confused
-    $cur_row_ids{$_} = defined $cur_row_data->[$_] ? $cur_row_data->[$_] : "\0NULL\xFF$rows_pos\xFF$_\0"
-      for (%1$s);
+    %2$s
 
-    # maybe(!) cache the top node id calculation
+    # in the case of an underdefined root - calculate the virtual id (otherwise no code at all)
     %3$s
 
-    $is_new_res = ! $collapse_idx[0]%2$s and (
+    $is_new_res = ! $collapse_idx[0]%4$s and (
       $_[1] and $result_pos and (unshift @{$_[2]}, $cur_row_data) and last
     );
 
     # the rel assemblers
-%4$s
+%5$s
 
-    $_[0][$result_pos++] = $collapse_idx[0]%2$s
+    $_[0][$result_pos++] = $collapse_idx[0]%4$s
       if $is_new_res;
   }
 
@@ -189,7 +198,7 @@ EOS
   # !!! note - different var than the one above
   # change the quoted placeholders to unquoted alias-references
   $parser_src =~ s/ \' \xFF__VALPOS__(\d+)__\xFF \' /"\$cur_row_data->[$1]"/gex;
-  $parser_src =~ s/ \' \xFF__IDVALPOS__(\d+)__\xFF \' /"\$cur_row_ids{$1}"/gex;
+  $parser_src =~ s/ \' \xFF__IDVALPOS__(\d+)__\xFF \' /$args->{hri_style} ? "\$cur_row_data->[$1]" : "\$cur_row_ids{$1}" /gex;
 
   $parser_src = "  { use strict; use warnings FATAL => 'all';\n$parser_src\n  }";
 }
@@ -292,16 +301,16 @@ sub __visit_infmap_collapse {
 
       if ($args->{hri_style}) {
 
-        $src[$rel_src_pos] = sprintf( '%s and %s',
-          "( defined '\xFF__VALPOS__${first_distinct_child_idcol}__\xFF' )",
-          $src[$rel_src_pos],
-        );
-
-        splice @src, $rel_src_pos + 1, 0, sprintf ( '%s{%s} ||= %s;',
+        # start of wrap of the entire chain in a conditional
+        splice @src, $rel_src_pos, 0, sprintf "( ! defined %s )\n  ? %s{%s} = %s\n  : do {",
+          "'\xFF__VALPOS__${first_distinct_child_idcol}__\xFF'",
           $node_idx_slot,
           perlstring($rel),
-          $relinfo->{-is_single} ? 'undef' : '[]',
-        );
+          $relinfo->{-is_single} ? 'undef' : '[]'
+        ;
+
+        # end of wrap
+        push @src, '};'
       }
       else {
 
diff --git a/t/resultset/rowparser_internals.t b/t/resultset/rowparser_internals.t
index 1317bb3..5bdebac 100644
--- a/t/resultset/rowparser_internals.t
+++ b/t/resultset/rowparser_internals.t
@@ -268,7 +268,7 @@ is_same_src (
     collapse => 1,
     hri_style => 1,
   }),
-  ' my($rows_pos, $result_pos, $cur_row_data, %cur_row_ids, @collapse_idx, $is_new_res) = (0, 0);
+  ' my($rows_pos, $result_pos, $cur_row_data, @collapse_idx, $is_new_res) = (0, 0);
 
     while ($cur_row_data = (
       ( $rows_pos >= 0 and $_[0][$rows_pos++] ) or do { $rows_pos = -1; undef } )
@@ -276,48 +276,45 @@ is_same_src (
       ( $_[1] and $_[1]->() )
     ) {
 
-      $cur_row_ids{$_} = defined $cur_row_data->[$_] ? $cur_row_data->[$_] : "\0NULL\xFF$rows_pos\xFF$_\0"
-        for (0, 1, 3, 4, 5);
-
       # a present cref in $_[1] implies lazy prefetch, implies a supplied stash in $_[2]
       $_[1] and $result_pos and unshift(@{$_[2]}, $cur_row_data) and last
-        if ( $is_new_res = ! $collapse_idx[0]{$cur_row_ids{4}}{$cur_row_ids{5}} );
+        if ( $is_new_res = ! $collapse_idx[0]{$cur_row_data->[4]}{$cur_row_data->[5]} );
 
       # the rowdata itself for root node
-      $collapse_idx[0]{$cur_row_ids{4}}{$cur_row_ids{5}} ||= { artist => $cur_row_data->[5], title => $cur_row_data->[4], year => $cur_row_data->[2] };
+      $collapse_idx[0]{$cur_row_data->[4]}{$cur_row_data->[5]} ||= { artist => $cur_row_data->[5], title => $cur_row_data->[4], year => $cur_row_data->[2] };
 
       # prefetch data of single_track (placed in root)
-      $collapse_idx[0]{$cur_row_ids{4}}{$cur_row_ids{5}}{single_track} ||= $collapse_idx[1]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}}
-        if defined $cur_row_data->[1];
-      $collapse_idx[0]{$cur_row_ids{4}}{$cur_row_ids{5}}{single_track} ||= undef;
+      (! defined($cur_row_data->[1]) ) ? $collapse_idx[0]{$cur_row_data->[4]}{$cur_row_data->[5]}{single_track} = undef : do {
+        $collapse_idx[0]{$cur_row_data->[4]}{$cur_row_data->[5]}{single_track} ||= $collapse_idx[1]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]};
 
-      # prefetch data of cd (placed in single_track)
-      $collapse_idx[1]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}}{cd} ||= $collapse_idx[2]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}};
+        # prefetch data of cd (placed in single_track)
+        $collapse_idx[1]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]}{cd} ||= $collapse_idx[2]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]};
 
-      # prefetch data of artist ( placed in single_track->cd)
-      $collapse_idx[2]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}}{artist} ||= $collapse_idx[3]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}} ||= { artistid => $cur_row_data->[1] };
+        # prefetch data of artist ( placed in single_track->cd)
+        $collapse_idx[2]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]}{artist} ||= $collapse_idx[3]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]} ||= { artistid => $cur_row_data->[1] };
 
-      # prefetch data of cds (if available)
-      ( defined $cur_row_data->[3] )
-        and
-      (! $collapse_idx[4]{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}} )
-        and
-      push @{$collapse_idx[3]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}}{cds}}, (
-        $collapse_idx[4]{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}} = { cdid => $cur_row_data->[3] }
-      );
-      $collapse_idx[3]{$cur_row_ids{1}}{$cur_row_ids{4}}{$cur_row_ids{5}}{cds} ||= [];
+        # prefetch data of cds (if available)
+        (! defined $cur_row_data->[3] ) ? $collapse_idx[3]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]}{cds} = [] : do {
 
-      # prefetch data of tracks (if available)
-      ( defined $cur_row_data->[0] )
-        and
-      (! $collapse_idx[5]{$cur_row_ids{0}}{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}} )
-        and
-      push @{$collapse_idx[4]{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}}{tracks}}, (
-        $collapse_idx[5]{$cur_row_ids{0}}{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}} = { title => $cur_row_data->[0] }
-      );
-      $collapse_idx[4]{$cur_row_ids{1}}{$cur_row_ids{3}}{$cur_row_ids{4}}{$cur_row_ids{5}}{tracks} ||= [];
+          (! $collapse_idx[4]{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]} )
+            and
+          push @{$collapse_idx[3]{$cur_row_data->[1]}{$cur_row_data->[4]}{$cur_row_data->[5]}{cds}}, (
+            $collapse_idx[4]{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]} = { cdid => $cur_row_data->[3] }
+          );
 
-      $_[0][$result_pos++] = $collapse_idx[0]{$cur_row_ids{4}}{$cur_row_ids{5}}
+          # prefetch data of tracks (if available)
+          ( ! defined $cur_row_data->[0] ) ? $collapse_idx[4]{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]}{tracks} = [] : do {
+
+            (! $collapse_idx[5]{$cur_row_data->[0]}{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]} )
+              and
+            push @{$collapse_idx[4]{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]}{tracks}}, (
+              $collapse_idx[5]{$cur_row_data->[0]}{$cur_row_data->[1]}{$cur_row_data->[3]}{$cur_row_data->[4]}{$cur_row_data->[5]} = { title => $cur_row_data->[0] }
+            );
+          };
+        };
+      };
+
+      $_[0][$result_pos++] = $collapse_idx[0]{$cur_row_data->[4]}{$cur_row_data->[5]}
         if $is_new_res;
     }
     splice @{$_[0]}, $result_pos;
@@ -638,7 +635,7 @@ is_same_src (
     collapse => 1,
     hri_style => 1,
   }),
-  ' my($rows_pos, $result_pos, $cur_row_data, %cur_row_ids, @collapse_idx, $is_new_res) = (0, 0);
+  ' my($rows_pos, $result_pos, $cur_row_data, @collapse_idx, $is_new_res) = (0, 0);
 
     while ($cur_row_data = (
       ( $rows_pos >= 0 and $_[0][$rows_pos++] ) or do { $rows_pos = -1; undef } )
@@ -646,11 +643,8 @@ is_same_src (
       ( $_[1] and $_[1]->() )
     ) {
 
-      $cur_row_ids{$_} = defined $$cur_row_data[$_] ? $$cur_row_data[$_] : "\0NULL\xFF$rows_pos\xFF$_\0"
-        for (0, 2, 3, 4, 8);
-
       # cache expensive set of ops in a non-existent rowid slot
-      $cur_row_ids{10} = (
+      $cur_row_data->[10] = (
         ( ( defined $cur_row_data->[0] ) && (join "\xFF", q{}, $cur_row_data->[0], q{} ))
           or
         ( ( defined $cur_row_data->[2] ) && (join "\xFF", q{}, $cur_row_data->[2], q{} ))
@@ -660,47 +654,46 @@ is_same_src (
 
       # a present cref in $_[1] implies lazy prefetch, implies a supplied stash in $_[2]
       $_[1] and $result_pos and unshift(@{$_[2]}, $cur_row_data) and last
-        if ( $is_new_res = ! $collapse_idx[0]{$cur_row_ids{10}} );
+        if ( $is_new_res = ! $collapse_idx[0]{$cur_row_data->[10]} );
 
-      $collapse_idx[0]{$cur_row_ids{10}} ||= { year => $$cur_row_data[1] };
+      $collapse_idx[0]{$cur_row_data->[10]} ||= { year => $$cur_row_data[1] };
 
-      (defined $cur_row_data->[0])
-        and
-      $collapse_idx[0]{$cur_row_ids{10}}{single_track} ||= ($collapse_idx[1]{$cur_row_ids{0}} ||= { trackid => $$cur_row_data[0] });
-      $collapse_idx[0]{$cur_row_ids{10}}{single_track} ||= undef;
+      (! defined $cur_row_data->[0] ) ? $collapse_idx[0]{$cur_row_data->[10]}{single_track} = undef : do {
 
-      $collapse_idx[1]{$cur_row_ids{0}}{cd} ||= $collapse_idx[2]{$cur_row_ids{0}};
+        $collapse_idx[0]{$cur_row_data->[10]}{single_track} ||= ($collapse_idx[1]{$cur_row_data->[0]} ||= { trackid => $$cur_row_data[0] });
 
-      $collapse_idx[2]{$cur_row_ids{0}}{artist} ||= ($collapse_idx[3]{$cur_row_ids{0}} ||= { artistid => $$cur_row_data[6] });
+        $collapse_idx[1]{$cur_row_data->[0]}{cd} ||= $collapse_idx[2]{$cur_row_data->[0]};
 
-      (defined $cur_row_data->[4])
-        and
-      (! $collapse_idx[4]{$cur_row_ids{0}}{$cur_row_ids{4}} )
-        and
-      push @{$collapse_idx[3]{$cur_row_ids{0}}{cds}}, (
-          $collapse_idx[4]{$cur_row_ids{0}}{$cur_row_ids{4}} = { cdid => $$cur_row_data[4], genreid => $$cur_row_data[7], year => $$cur_row_data[5] }
-      );
-      $collapse_idx[3]{$cur_row_ids{0}}{cds} ||= [];
+        $collapse_idx[2]{$cur_row_data->[0]}{artist} ||= ($collapse_idx[3]{$cur_row_data->[0]} ||= { artistid => $$cur_row_data[6] });
 
-      (defined $cur_row_data->[8])
-        and
-      (! $collapse_idx[5]{$cur_row_ids{0}}{$cur_row_ids{4}}{$cur_row_ids{8}} )
-        and
-      push @{$collapse_idx[4]{$cur_row_ids{0}}{$cur_row_ids{4}}{tracks}}, (
-          $collapse_idx[5]{$cur_row_ids{0}}{$cur_row_ids{4}}{$cur_row_ids{8}} = { title => $$cur_row_data[8] }
-      );
-      $collapse_idx[4]{$cur_row_ids{0}}{$cur_row_ids{4}}{tracks} ||= [];
+        (! defined $cur_row_data->[4] ) ? $collapse_idx[3]{$cur_row_data->[0]}{cds} = [] : do {
 
-      (defined $cur_row_data->[2])
-        and
-      (! $collapse_idx[6]{$cur_row_ids{2}}{$cur_row_ids{3}} )
-        and
-      push @{$collapse_idx[0]{$cur_row_ids{10}}{tracks}}, (
-          $collapse_idx[6]{$cur_row_ids{2}}{$cur_row_ids{3}} = { cd => $$cur_row_data[2], title => $$cur_row_data[3] }
-      );
-      $collapse_idx[0]{$cur_row_ids{10}}{tracks} ||= [];
+          (! $collapse_idx[4]{$cur_row_data->[0]}{$cur_row_data->[4]} )
+            and
+          push @{$collapse_idx[3]{$cur_row_data->[0]}{cds}}, (
+              $collapse_idx[4]{$cur_row_data->[0]}{$cur_row_data->[4]} = { cdid => $$cur_row_data[4], genreid => $$cur_row_data[7], year => $$cur_row_data[5] }
+          );
 
-      $_[0][$result_pos++] = $collapse_idx[0]{$cur_row_ids{10}}
+          (! defined $cur_row_data->[8] ) ? $collapse_idx[4]{$cur_row_data->[0]}{$cur_row_data->[4]}{tracks} = [] : do {
+
+            (! $collapse_idx[5]{$cur_row_data->[0]}{$cur_row_data->[4]}{$cur_row_data->[8]} )
+              and
+            push @{$collapse_idx[4]{$cur_row_data->[0]}{$cur_row_data->[4]}{tracks}}, (
+                $collapse_idx[5]{$cur_row_data->[0]}{$cur_row_data->[4]}{$cur_row_data->[8]} = { title => $$cur_row_data[8] }
+            );
+          };
+        };
+      };
+
+      (! defined $cur_row_data->[2] ) ? $collapse_idx[0]{$cur_row_data->[10]}{tracks} = [] : do {
+        (! $collapse_idx[6]{$cur_row_data->[2]}{$cur_row_data->[3]} )
+          and
+        push @{$collapse_idx[0]{$cur_row_data->[10]}{tracks}}, (
+            $collapse_idx[6]{$cur_row_data->[2]}{$cur_row_data->[3]} = { cd => $$cur_row_data[2], title => $$cur_row_data[3] }
+        );
+      };
+
+      $_[0][$result_pos++] = $collapse_idx[0]{$cur_row_data->[10]}
         if $is_new_res;
     }