X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=lib%2FHTML%2FZoom%2FFilterBuilder.pm;h=50398b354dc09f4d213da6762f5474d1f812891a;hb=a42917f6295747a1b4b3d081db5e2f1feaca9d00;hp=71b2d5f20a1cc54d95177098feef18ccdd372529;hpb=456a815dd80c8a43c7d5a04d353377426a94048f;p=catagits%2FHTML-Zoom.git

diff --git a/lib/HTML/Zoom/FilterBuilder.pm b/lib/HTML/Zoom/FilterBuilder.pm
index 71b2d5f..50398b3 100644
--- a/lib/HTML/Zoom/FilterBuilder.pm
+++ b/lib/HTML/Zoom/FilterBuilder.pm
@@ -1,43 +1,35 @@
 package HTML::Zoom::FilterBuilder;
 
-use Devel::Dwarn;
-
 use strict;
 use warnings FATAL => 'all';
+use base qw(HTML::Zoom::SubObject);
 use HTML::Zoom::CodeStream;
 
-sub new { bless({}, shift) }
-
 sub _stream_from_code {
-  HTML::Zoom::CodeStream->new({ code => $_[1] })
+  shift->_zconfig->stream_utils->stream_from_code(@_)
 }
 
 sub _stream_from_array {
-  shift; # lose $self
-  HTML::Zoom::CodeStream->from_array(@_)
+  shift->_zconfig->stream_utils->stream_from_array(@_)
+}
+
+sub _stream_from_proto {
+  shift->_zconfig->stream_utils->stream_from_proto(@_)
 }
 
 sub _stream_concat {
-  shift; # lose $self
-  my @streams = @_;
-  my $cur_stream = shift(@streams) or die "No streams passed";
-  HTML::Zoom::CodeStream->new({
-    code => sub {
-      return unless $cur_stream;
-      my $evt;
-      until (($evt) = $cur_stream->next) {
-        return unless $cur_stream = shift(@streams);
-      }
-      return $evt;
-    }
-  });
+  shift->_zconfig->stream_utils->stream_concat(@_)
+}
+
+sub _flatten_stream_of_streams {
+  shift->_zconfig->stream_utils->flatten_stream_of_streams(@_)
 }
 
 sub set_attribute {
-  my ($self, $args) = @_;
-  my ($name, $value) = @{$args}{qw(name value)};
+  my $self = shift;
+  my ($name, $value) = $self->_parse_attribute_args(@_);
   sub {
-    my $a = (my $evt = shift)->{attrs};
+    my $a = (my $evt = $_[0])->{attrs};
     my $e = exists $a->{$name};
     +{ %$evt, raw => undef, raw_attrs => undef,
        attrs => { %$a, $name => $value },
@@ -48,11 +40,23 @@ sub set_attribute {
    };
 }
 
+sub _parse_attribute_args {
+  my $self = shift;
+  # allow ->add_to_attribute(name => 'value')
+  #    or ->add_to_attribute({ name => 'name', value => 'value' })
+  my ($name, $value) = @_ > 1 ? @_ : @{$_[0]}{qw(name value)};
+  return ($name, $self->_zconfig->parser->html_escape($value));
+}
+
 sub add_attribute {
-  my ($self, $args) = @_;
-  my ($name, $value) = @{$args}{qw(name value)};
+    die "renamed to add_to_attribute. killing this entirely for 1.0";
+}
+
+sub add_to_attribute {
+  my $self = shift;
+  my ($name, $value) = $self->_parse_attribute_args(@_);
   sub {
-    my $a = (my $evt = shift)->{attrs};
+    my $a = (my $evt = $_[0])->{attrs};
     my $e = exists $a->{$name};
     +{ %$evt, raw => undef, raw_attrs => undef,
        attrs => {
@@ -68,9 +72,9 @@ sub add_attribute {
 
 sub remove_attribute {
   my ($self, $args) = @_;
-  my $name = $args->{name};
+  my $name = (ref($args) eq 'HASH') ? $args->{name} : $args;
   sub {
-    my $a = (my $evt = shift)->{attrs};
+    my $a = (my $evt = $_[0])->{attrs};
     return $evt unless exists $a->{$name};
     $a = { %$a }; delete $a->{$name};
     +{ %$evt, raw => undef, raw_attrs => undef,
@@ -80,28 +84,85 @@ sub remove_attribute {
   };
 }
 
+sub collect {
+  my ($self, $options) = @_;
+  my ($into, $passthrough, $content, $filter, $flush_before) =
+    @{$options}{qw(into passthrough content filter flush_before)};
+  sub {
+    my ($evt, $stream) = @_;
+    # We wipe the contents of @$into here so that other actions depending
+    # on this (such as a repeater) can be invoked multiple times easily.
+    # I -suspect- it's better for that state reset to be managed here; if it
+    # ever becomes painful the decision should be revisited
+    if ($into) {
+      @$into = $content ? () : ($evt);
+    }
+    if ($evt->{is_in_place_close}) {
+      return $evt if $passthrough || $content;
+      return;
+    }
+    my $name = $evt->{name};
+    my $depth = 1;
+    my $_next = $content ? 'peek' : 'next';
+    $stream = do { local $_ = $stream; $filter->($stream) } if $filter;
+    my $collector = $self->_stream_from_code(sub {
+      return unless $stream;
+      while (my ($evt) = $stream->$_next) {
+        $depth++ if ($evt->{type} eq 'OPEN');
+        $depth-- if ($evt->{type} eq 'CLOSE');
+        unless ($depth) {
+          undef $stream;
+          return if $content;
+          push(@$into, $evt) if $into;
+          return $evt if $passthrough;
+          return;
+        }
+        push(@$into, $evt) if $into;
+        $stream->next if $content;
+        return $evt if $passthrough;
+      }
+      die "Never saw closing </${name}> before end of source";
+    });
+    if ($flush_before) {
+      if ($passthrough||$content) {
+        $evt = { %$evt, flush => 1 };
+      } else {
+        $evt = { type => 'EMPTY', flush => 1 };
+      }
+    }
+    return ($passthrough||$content||$flush_before)
+             ? [ $evt, $collector ]
+             : $collector;
+  };
+}
+
+sub collect_content {
+  my ($self, $options) = @_;
+  $self->collect({ %{$options||{}}, content => 1 })
+}
+
 sub add_before {
   my ($self, $events) = @_;
-  sub { return $self->_stream_from_array(@$events, shift) };
+  sub { return $self->_stream_from_array(@$events, $_[0]) };
 }
 
 sub add_after {
   my ($self, $events) = @_;
+  my $coll_proto = $self->collect({ passthrough => 1 });
   sub {
-    my ($evt, $stream) = @_;
+    my ($evt) = @_;
     my $emit = $self->_stream_from_array(@$events);
-    if ($evt->{is_in_place_close}) {
-      return [ $evt, $emit ];
-    }
-    my ($filtered_evt, $coll) = @{$self->collect(undef, 1)->(@_)};
-    return [ $filtered_evt, $self->_stream_concat($coll, $emit) ];
+    my $coll = &$coll_proto;
+    return ref($coll) eq 'HASH' # single event, no collect
+      ? [ $coll, $emit ]
+      : [ $coll->[0], $self->_stream_concat($coll->[1], $emit) ];
   };
-}  
+}
 
-sub prepend_inside {
+sub prepend_content {
   my ($self, $events) = @_;
   sub {
-    my $evt = shift;
+    my ($evt) = @_;
     if ($evt->{is_in_place_close}) {
       $evt = { %$evt }; delete @{$evt}{qw(raw is_in_place_close)};
       return [ $evt, $self->_stream_from_array(
@@ -112,46 +173,306 @@ sub prepend_inside {
   };
 }
 
-sub replace {
+sub append_content {
   my ($self, $events) = @_;
+  my $coll_proto = $self->collect({ passthrough => 1, content => 1 });
   sub {
-    my ($evt, $stream) = @_;
-    my $emit = $self->_stream_from_array(@$events);
+    my ($evt) = @_;
     if ($evt->{is_in_place_close}) {
-      return $emit
+      $evt = { %$evt }; delete @{$evt}{qw(raw is_in_place_close)};
+      return [ $evt, $self->_stream_from_array(
+        @$events, { type => 'CLOSE', name => $evt->{name} }
+      ) ];
     }
-    return $self->_stream_concat($emit, $self->collect->(@_));
+    my $coll = &$coll_proto;
+    my $emit = $self->_stream_from_array(@$events);
+    return [ $coll->[0], $self->_stream_concat($coll->[1], $emit) ];
   };
 }
 
-sub collect {
-  my ($self, $into, $passthrough) = @_;
+sub replace {
+  my ($self, $replace_with, $options) = @_;
+  my $coll_proto = $self->collect($options);
   sub {
     my ($evt, $stream) = @_;
-    push(@$into, $evt) if $into;
-    if ($evt->{is_in_place_close}) {
-      return $evt if $passthrough;
-      return;
+    my $emit = $self->_stream_from_proto($replace_with);
+    my $coll = &$coll_proto;
+    # if we're replacing the contents of an in place close
+    # then we need to handle that here
+    if ($options->{content}
+        && ref($coll) eq 'HASH'
+        && $coll->{is_in_place_close}
+      ) {
+      my $close = $stream->next;
+      # shallow copy and nuke in place and raw (to force smart print)
+      $_ = { %$_ }, delete @{$_}{qw(is_in_place_close raw)} for ($coll, $close);
+      $emit = $self->_stream_concat(
+                $emit,
+                $self->_stream_from_array($close),
+              );
     }
-    my $name = $evt->{name};
-    my $depth = 1;
-    my $collector = $self->_stream_from_code(sub {
-      return unless $stream;
-      while (my ($evt) = $stream->next) {
-        $depth++ if ($evt->{type} eq 'OPEN');
-        $depth-- if ($evt->{type} eq 'CLOSE');
-        unless ($depth) {
-          undef $stream;
-          return $evt if $passthrough;
-          return;
-        }
-        push(@$into, $evt) if $into;
-        return $evt if $passthrough;
+    # For a straightforward replace operation we can, in fact, do the emit
+    # -before- the collect, and my first cut did so. However in order to
+    # use the captured content in generating the new content, we need
+    # the collect stage to happen first - and it seems highly unlikely
+    # that in normal operation the collect phase will take long enough
+    # for the difference to be noticeable
+    return
+      ($coll
+        ? (ref $coll eq 'ARRAY' # [ event, stream ]
+            ? [ $coll->[0], $self->_stream_concat($coll->[1], $emit) ]
+            : (ref $coll eq 'HASH' # event or stream?
+                 ? [ $coll, $emit ]
+                 : $self->_stream_concat($coll, $emit))
+          )
+        : $emit
+      );
+  };
+}
+
+sub replace_content {
+  my ($self, $replace_with, $options) = @_;
+  $self->replace($replace_with, { %{$options||{}}, content => 1 })
+}
+
+sub repeat {
+  my ($self, $repeat_for, $options) = @_;
+  $options->{into} = \my @into;
+  my @between;
+  my $repeat_between = delete $options->{repeat_between};
+  if ($repeat_between) {
+    $options->{filter} = sub {
+      $_->select($repeat_between)->collect({ into => \@between })
+    };
+  }
+  my $repeater = sub {
+    my $s = $self->_stream_from_proto($repeat_for);
+    # We have to test $repeat_between not @between here because
+    # at the point we're constructing our return stream @between
+    # hasn't been populated yet - but we can test @between in the
+    # map routine because it has been by then and that saves us doing
+    # the extra stream construction if we don't need it.
+    $self->_flatten_stream_of_streams(do {
+      if ($repeat_between) {
+        $s->map(sub {
+              local $_ = $self->_stream_from_array(@into);
+              (@between && $s->peek)
+                ? $self->_stream_concat(
+                    $_[0]->($_), $self->_stream_from_array(@between)
+                  )
+                : $_[0]->($_)
+            })
+      } else {
+        $s->map(sub {
+              local $_ = $self->_stream_from_array(@into);
+              $_[0]->($_)
+          })
       }
-      die "Never saw closing </${name}> before end of source";
-    });
-    return $passthrough ? [ $evt, $collector ] : $collector;
+    })
   };
+  $self->replace($repeater, $options);
+}
+
+sub repeat_content {
+  my ($self, $repeat_for, $options) = @_;
+  $self->repeat($repeat_for, { %{$options||{}}, content => 1 })
 }
 
 1;
+
+=head1 NAME
+
+HTML::Zoom::FilterBuilder - Add Filters to a Stream
+
+=head1 SYNOPSIS
+
+Create an L<HTML::Zoom> instance:
+
+  use HTML::Zoom;
+  my $root = HTML::Zoom
+      ->from_html(<<MAIN);
+  <html>
+    <head>
+      <title>Default Title</title>
+    </head>
+    <body bad_attr='junk'>
+      Default Content
+    </body>
+  </html>
+  MAIN
+
+Create a new attribute on the  C<body> tag:
+
+  $root = $root
+    ->select('body')
+    ->set_attribute(class=>'main');
+
+Add a extra value to an existing attribute:
+
+  $root = $root
+    ->select('body')
+    ->add_to_attribute(class=>'one-column');
+
+Set the content of the C<title> tag:
+
+  $root = $root
+    ->select('title')
+    ->replace_content('Hello World');
+
+Set content from another L<HTML::Zoom> instance:
+
+  my $body = HTML::Zoom
+      ->from_html(<<BODY);
+  <div id="stuff">
+      <p>Well Now</p>
+      <p id="p2">Is the Time</p>
+  </div>
+  BODY
+
+  $root = $root
+    ->select('body')
+    ->replace_content($body);
+
+Set an attribute on multiple matches:
+
+  $root = $root
+    ->select('p')
+    ->set_attribute(class=>'para');
+
+Remove an attribute:
+
+  $root = $root
+    ->select('body')
+    ->remove_attribute('bad_attr');
+
+will produce:
+
+=begin testinfo
+
+  my $output = $root->to_html;
+  my $expect = <<HTML;
+
+=end testinfo
+
+  <html>
+    <head>
+      <title>Hello World</title>
+    </head>
+    <body class="main one-column"><div id="stuff">
+      <p class="para">Well Now</p>
+      <p id="p2" class="para">Is the Time</p>
+  </div>
+  </body>
+  </html>
+
+=begin testinfo
+
+  HTML
+  is($output, $expect, 'Synopsis code works ok');
+
+=end testinfo
+
+=head1 DESCRIPTION
+
+Given a L<HTML::Zoom> stream, provide methods to apply filters which
+alter the content of that stream.
+
+=head1 METHODS
+
+This class defines the following public API
+
+=head2 set_attribute ( $attr=>value | {name=>$attr,value=>$value} )
+
+Sets an attribute of a given name to a given value for all matching selections.
+
+    $html_zoom
+      ->select('p')
+      ->set_attribute(class=>'paragraph')
+      ->select('div')
+      ->set_attribute(name=>'class', value=>'divider');
+
+
+Overrides existing values, if such exist.  When multiple L</set_attribute>
+calls are made against the same or overlapping selection sets, the final
+call wins.
+
+=head2 add_to_attribute ( $attr=>value | {name=>$attr,value=>$value} )
+
+Adds a value to an existing attribute, or creates one if the attribute does not
+yet exist.
+
+    $html_zoom
+      ->select('p')
+      ->set_attribute(class=>'paragraph')
+      ->then
+      ->add_to_attribute(name=>'class', value=>'divider');
+
+Attributes with more than one value will have a dividing space.
+
+=head2 remove_attribute ( $attr | {name=>$attr} )
+
+Removes an attribute and all its values.
+
+    $html_zoom
+      ->select('p')
+      ->set_attribute(class=>'paragraph')
+      ->then
+      ->remove_attribute('class');
+
+Removes attributes from the original stream or events already added.
+
+=head2 collect
+
+    TBD
+
+=head2 collect_content
+
+    TBD
+
+=head2 add_before
+
+    TBD
+
+=head2 add_after
+
+    TBD
+
+=head2 prepend_content
+
+    TBD
+
+=head2 append_content
+
+    TBD
+
+=head2 replace
+
+    TBD
+
+=head2 replace_content
+
+Given a L<HTML::Zoom/select> result, replace the content with a string, array
+or another L<HTML::Zoom> object.
+
+=head2 repeat
+
+    TBD
+
+=head2 repeat_content
+
+    TBD
+
+=head1 ALSO SEE
+
+L<HTML::Zoom>
+
+=head1 AUTHORS
+
+See L<HTML::Zoom> for authors.
+
+=head1 LICENSE
+
+See L<HTML::Zoom> for the license.
+
+=cut
+