Author: allison
Date: Sun Jan 25 19:21:17 2009
New Revision: 36012
Added:
trunk/languages/pod/
trunk/languages/pod/README
trunk/languages/pod/config/
trunk/languages/pod/config/makefiles/
trunk/languages/pod/config/makefiles/root.in
trunk/languages/pod/pod.pir
trunk/languages/pod/src/
trunk/languages/pod/src/parser/
trunk/languages/pod/src/parser/actions.pm
trunk/languages/pod/src/parser/grammar.pg
trunk/languages/pod/t/
trunk/languages/pod/t/harness
trunk/languages/pod/test.pod
Log:
[languages] Adding a PGE-based POD parser.
Added: trunk/languages/pod/README
==============================================================================
--- (empty file)
+++ trunk/languages/pod/README Sun Jan 25 19:21:17 2009
@@ -0,0 +1,30 @@
+This is a liberal Pod parser, intended to have the semantics checked and
+document structure built by an appropriate actions file. See the file test.pod
+for the syntax it can handle. It can be run with the '--target=parse' option,
+to dump the parse tree:
+
+ ../../parrot pod.pir --target=parse test.pod
+
+
+This parser mostly follows the specification in 'perlpodspec', with the
+following exceptions:
+
+ - It doesn't allow formatting codes to span multiple lines.
+
+ - It doesn't allow headers or item names to span multiple lines.
+
+ - It requires =cut at the end of a section of Pod, even at the end of the
+ file.
+
+ - It allows any Pod formatting within a =begin/=end section, with no special
+ treatment. (The semantic handling may impose constraints.)
+
+ - It doesn't truncate whitespace within a formatting code (leaving that up to
+ the semantic handling).
+
+ - No special parsing is done for L<> codes (leaving that up to the
formatter).
+
+The current actions file is a no-op, simply returning the parse tree (but left
+in place because HLLCompiler apparently can't handle a null actions file).
+
+The 'make test' target does nothing.
Added: trunk/languages/pod/config/makefiles/root.in
==============================================================================
--- (empty file)
+++ trunk/languages/pod/config/makefiles/root.in Sun Jan 25 19:21:17 2009
@@ -0,0 +1,118 @@
+## $Id$
+
+## arguments we want to run parrot with
+PARROT_ARGS =
+
+## configuration settings
+BUILD_DIR = @build_dir@
+LOAD_EXT = @load_ext@
+O = @o@
+
+## Setup some commands
+LN_S = @lns@
+PERL = @perl@
+RM_RF = @rm_rf@
+CP = @cp@
+PARROT = ../../par...@exe@
+CAT = $(PERL) -MExtUtils::Command -e cat
+BUILD_DYNPMC = $(PERL) $(BUILD_DIR)/tools/build/dynpmc.pl
+RECONFIGURE = $(PERL) $(BUILD_DIR)/tools/dev/reconfigure.pl
+#CONDITIONED_LINE(darwin):
+#CONDITIONED_LINE(darwin):# MACOSX_DEPLOYMENT_TARGET must be defined for OS X
compilation/linking
+#CONDITIONED_LINE(darwin):export MACOSX_DEPLOYMENT_TARGET := @osx_version@
+
+## places to look for things
+PARROT_DYNEXT = $(BUILD_DIR)/runtime/parrot/dynext
+PGE_LIBRARY = $(BUILD_DIR)/runtime/parrot/library/PGE
+PERL6GRAMMAR = $(PGE_LIBRARY)/Perl6Grammar.pbc
+NQP = $(BUILD_DIR)/compilers/nqp/nqp.pbc
+PCT = $(BUILD_DIR)/runtime/parrot/library/PCT.pbc
+
+PMC_DIR = src/pmc
+
+all: pod.pbc
+
+POD_GROUP = $(PMC_DIR)/pod_group$(LOAD_EXT)
+
+SOURCES = pod.pir \
+ src/gen_grammar.pir \
+ src/gen_actions.pir \
+# $(POD_GROUP)
+
+# PMCS = pod
+# PMC_SOURCES = $(PMC_DIR)/pod.pmc
+
+# the default target
+pod.pbc: $(PARROT) $(SOURCES)
+ $(PARROT) $(PARROT_ARGS) -o pod.pbc pod.pir
+
+src/gen_grammar.pir: $(PERL6GRAMMAR) src/parser/grammar.pg
+ $(PARROT) $(PARROT_ARGS) $(PERL6GRAMMAR) \
+ --output=src/gen_grammar.pir \
+ src/parser/grammar.pg \
+
+src/gen_actions.pir: $(NQP) $(PCT) src/parser/actions.pm
+ $(PARROT) $(PARROT_ARGS) $(NQP) --output=src/gen_actions.pir \
+ --target=pir src/parser/actions.pm
+
+$(POD_GROUP): $(PARROT) $(PMC_SOURCES)
+ cd $(PMC_DIR) && $(BUILD_DYNPMC) generate $(PMCS)
+ cd $(PMC_DIR) && $(BUILD_DYNPMC) compile $(PMCS)
+ cd $(PMC_DIR) && $(BUILD_DYNPMC) linklibs $(PMCS)
+ cd $(PMC_DIR) && $(BUILD_DYNPMC) copy --destination=$(PARROT_DYNEXT)
$(PMCS)
+
+# regenerate the Makefile
+Makefile: config/makefiles/root.in
+ cd $(BUILD_DIR) && $(RECONFIGURE) --step=gen::languages --languages=pod
+
+# This is a listing of all targets, that are meant to be called by users
+help:
+ @echo ""
+ @echo "Following targets are available for the user:"
+ @echo ""
+ @echo " all: pod.pbc"
+ @echo " This is the default."
+ @echo "Testing:"
+ @echo " test: Run the test suite."
+ @echo " testclean: Clean up test results."
+ @echo ""
+ @echo "Cleaning:"
+ @echo " clean: Basic cleaning up."
+ @echo " realclean: Removes also files generated by
'Configure.pl'"
+ @echo " distclean: Removes also anything built, in theory"
+ @echo ""
+ @echo "Misc:"
+ @echo " help: Print this help message."
+ @echo ""
+
+test: all
+ $(PERL) t/harness
+
+# this target has nothing to do
+testclean:
+
+CLEANUPS = \
+ pod.pbc \
+ src/gen_grammar.pir \
+ src/gen_actions.pir \
+ $(PMC_DIR)/*.h \
+ $(PMC_DIR)/*.c \
+ $(PMC_DIR)/*.dump \
+ $(PMC_DIR)/*$(O) \
+ $(PMC_DIR)/*$(LOAD_EXT) \
+ $(PMC_DIR)/*.exp \
+ $(PMC_DIR)/*.ilk \
+ $(PMC_DIR)/*.manifest \
+ $(PMC_DIR)/*.pdb \
+ $(PMC_DIR)/*.lib \
+
+
+clean: testclean
+ $(RM_RF) $(CLEANUPS)
+
+realclean: clean
+ $(RM_RF) Makefile
+
+distclean: realclean
+
+
Added: trunk/languages/pod/pod.pir
==============================================================================
--- (empty file)
+++ trunk/languages/pod/pod.pir Sun Jan 25 19:21:17 2009
@@ -0,0 +1,77 @@
+=head1 TITLE
+
+pod.pir - A Pod compiler.
+
+=head2 Description
+
+This is the base file for the Pod compiler.
+
+This file includes the parsing and grammar rules from
+the src/ directory, loads the relevant PGE libraries,
+and registers the compiler under the name 'Pod'.
+
+=head2 Functions
+
+=over 4
+
+=item onload()
+
+Creates the Pod compiler using a C<PCT::HLLCompiler>
+object.
+
+=cut
+
+.HLL 'pod'
+
+.namespace [ 'Pod';'Compiler' ]
+
+.loadlib 'pod_group'
+
+.sub '' :anon :load :init
+ load_bytecode 'PCT.pbc'
+ .local pmc parrotns, hllns, exports
+ parrotns = get_root_namespace ['parrot']
+ hllns = get_hll_namespace
+ exports = split ' ', 'PAST PCT PGE'
+ parrotns.'export_to'(hllns, exports)
+.end
+
+.include 'src/gen_grammar.pir'
+.include 'src/gen_actions.pir'
+
+.sub 'onload' :anon :load :init
+ $P0 = get_hll_global ['PCT'], 'HLLCompiler'
+ $P1 = $P0.'new'()
+ $P1.'language'('pod')
+ $P0 = get_hll_namespace ['Pod';'Grammar']
+ $P1.'parsegrammar'($P0)
+ $P0 = get_hll_namespace ['Pod';'Grammar';'Actions']
+ $P1.'parseactions'($P0)
+.end
+
+=item main(args :slurpy) :main
+
+Start compilation by passing any command line C<args>
+to the Pod compiler.
+
+=cut
+
+.sub 'main' :main
+ .param pmc args
+
+ $P0 = compreg 'pod'
+ $P1 = $P0.'command_line'(args)
+.end
+
+.include 'src/gen_builtins.pir'
+
+=back
+
+=cut
+
+# Local Variables:
+# mode: pir
+# fill-column: 100
+# End:
+# vim: expandtab shiftwidth=4 ft=pir:
+
Added: trunk/languages/pod/src/parser/actions.pm
==============================================================================
--- (empty file)
+++ trunk/languages/pod/src/parser/actions.pm Sun Jan 25 19:21:17 2009
@@ -0,0 +1,29 @@
+# $Id$
+
+=begin comments
+
+Pod::Grammar::Actions - ast transformations for Pod
+
+This file contains the methods that are used by the parse grammar
+to build the PAST representation of an Pod program.
+Each method below corresponds to a rule in F<src/parser/grammar.pg>,
+and is invoked at the point where C<{*}> appears in the rule,
+with the current match object as the first argument. If the
+line containing C<{*}> also has a C<#= key> comment, then the
+value of the comment is passed as the second argument to the method.
+
+=end comments
+
+class Pod::Grammar::Actions;
+
+method TOP($/) {
+ make $( $/ );
+}
+
+# Local Variables:
+# mode: cperl
+# cperl-indent-level: 4
+# fill-column: 100
+# End:
+# vim: expandtab shiftwidth=4:
+
Added: trunk/languages/pod/src/parser/grammar.pg
==============================================================================
--- (empty file)
+++ trunk/languages/pod/src/parser/grammar.pg Sun Jan 25 19:21:17 2009
@@ -0,0 +1,180 @@
+# $Id$
+
+=begin overview
+
+This is the grammar for Pod written as a sequence of Perl 6 rules.
+
+=end overview
+
+grammar Pod::Grammar is PCT::Grammar;
+
+rule TOP {
+ [ <skipped>
+ | <pod_section>
+ ]*
+ [ $ || <panic: 'Pod syntax error'> ]
+ {*}
+}
+
+rule skipped {
+ ^^ <![=]> \N*
+ {*}
+}
+
+rule pod_section {
+ [ <pod_directive> | <?before '='> ]
+ <pod_sequence>*
+ <cut_directive>
+ {*}
+}
+
+rule pod_sequence {
+ <back_directive> {*}
+ | <item_directive> {*}
+ | <over_directive> {*}
+ | <heading>
+ | <begin_directive>
+ | <end_directive>
+ | <for_directive>
+ | <encoding_directive>
+# | <literal_paragraph>
+ | <paragraph>
+}
+
+token pod_directive {
+ ^^ '=pod'
+ <.blank_line>
+ {*}
+}
+
+token cut_directive {
+ ^^ '=cut'
+ <.blank_line>?
+ {*}
+}
+
+token heading {
+ ^^
+ '=head'
+ <digit>
+ <block_title>?
+ <.blank_line>
+ {*}
+}
+
+token begin_directive {
+ ^^
+ '=begin'
+ <block_name>
+ <block_title>?
+ <.blank_line>
+ {*}
+}
+
+token end_directive {
+ ^^ '=end'
+ <block_name>
+ <.blank_line>
+ {*}
+}
+
+token for_directive {
+ ^^
+ '=for'
+ <block_name>
+ <block_title>?
+ <.blank_line>
+ {*}
+}
+
+token over_directive {
+ ^^
+ '=over'
+ [ <.pod_ws>
+ <digit>*
+ [ '.' <digit>+ ]?
+ ]?
+ <.blank_line>
+ {*}
+}
+
+token back_directive {
+ ^^ '=back'
+ <.blank_line>
+ {*}
+}
+
+token item_directive {
+ ^^
+ '=item'
+ [ <.pod_ws>
+ [ '*'
+ | <digit>* '.'?
+ ]
+ ]?
+ <.blank_line>
+ {*}
+}
+
+token encoding_directive {
+ ^^
+ '=encoding'
+ <block_name>
+ <.blank_line>
+ {*}
+}
+
+regex paragraph {
+ ^^
+ <!before '='> # Not a directive
+ <formatted_text>
+ [ \n <formatted_text> ]*
+ <.blank_line>
+ {*}
+}
+
+token literal_paragraph {
+ [ <.pod_ws> <formatted_text> \n ]+
+ <.blank_line>
+ {*}
+}
+
+token block_name {
+ <.pod_ws>
+ <[_]+alpha+digit>+
+ {*}
+}
+
+token block_title {
+ <.pod_ws>
+ <formatted_text>
+ {*}
+}
+
+regex formatted_text {
+ [ <format_code>
+ | <alpha>
+ | <digit>
+ | <punct>
+ | <pod_ws>
+ ]+
+ {*}
+}
+
+regex format_code {
+ <[BCEFILSXZ]>
+ [ '<<<' <formatted_text> '>>>'
+ | '<<' <formatted_text> '>>'
+ | '<' <formatted_text> '>'
+ ]
+ {*}
+}
+
+#token ws { <!ww> \s+ }
+
+token pod_ws {
+ [ ' ' | \t ]+ # Literal spaces or tabs, no newlines or other whitespace
+ {*}
+}
+
+token blank_line { \n <.pod_ws>? \n }
Added: trunk/languages/pod/t/harness
==============================================================================
--- (empty file)
+++ trunk/languages/pod/t/harness Sun Jan 25 19:21:17 2009
@@ -0,0 +1,13 @@
+#! perl
+
+# $Id$
+
+# pragmata
+use strict;
+use warnings;
+use 5.008;
+
+use lib qw( . lib ../lib ../../lib ../../lib );
+use Parrot::Test::Harness language => 'Pod',
+ compiler => 'pod.pbc';
+
Added: trunk/languages/pod/test.pod
==============================================================================
--- (empty file)
+++ trunk/languages/pod/test.pod Sun Jan 25 19:21:17 2009
@@ -0,0 +1,87 @@
+Foo
+Bar
+Baz
+
+=pod
+
+=over
+
+=item
+
+=back
+
+=cut
+
+Some blob of text.
+
+=over 4.5
+
+=item *
+
+=back
+
+=cut
+
+Another blob of text.
+
+=over 4
+
+=item 45
+
+=back
+
+=cut
+
+More text.
+
+=head1 Foo
+
+=cut
+
+
+
+
+=head2
+
+=cut
+
+=begin foo
+
+=end foo
+
+=cut
+
+=begin bar Some Text
+
+=end bar
+
+=cut
+
+=for baz
+
+=cut
+
+=head1 a paragraph
+
+Random paragraph.
+
+=cut
+
+=head2 more paragraphs
+
+Another random paragraph.
+Including a newline.
+
+=cut
+
+=head3 formatting codes
+
+Text containing E<formatting B<codes>>.
+
+A multi-bracket formating code containing C<<special -> characters>>.
+
+=cut
+
+=encoding utf8
+
+=cut
