The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
2003-12-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* utils/rdfproc.c: tweak help text

	* librdf/rdf_parser.h:
	Note parser syntax name, not raptor factory name, which is in the
	raptor parser context.

	* librdf/rdf_parser_raptor.c:
	Separate registered name from raptor parser name for all calls of
	raptor_new_parser.
	(librdf_parser_raptor_init): Convert the legacy parser name
	'raptor' into 'rdfxml' internally.
	(librdf_parser_raptor_constructor): Use raptor_parsers_enumerate
	and hard code mime types, URIs here for now.  Can use
	raptor_syntaxes_enumerate later.

	* raptor/raptor_rss.c, raptor/raptor_parse.c,
	raptor/ntriples_parse.c, raptor/n3_parser.y: Use expanded
	raptor_parser_register_factory with mime_type and uri_string args
	where appropriate.

	* raptor/libraptor.3, raptor/raptor.h: Added raptor_syntaxes_enumerate

	* raptor/raptor_general.c (raptor_parser_register_factory):
	Add mime_type and uri_string args, both optional.
	(raptor_syntaxes_enumerate): Added to get syntax name, label,
	mime_type or uri_string - all optional.
	(raptor_parsers_enumerate): Uses raptor_syntaxes_enumerate.

	* raptor/raptor_internal.h: Store parser mime_type, URI in
	raptor_parser_factory Update raptor_parser_register_factory to
	take mime_type, uri_string args.

	* utils/rdfproc.c: tweak help formatting

	* raptor/configure.ac: Bumped version to 1.2.0

	* raptor/configure.ac:
	Update RAPTOR_LIBTOOL_VERSION to reflect interfaces added,
	none removed giving current 2:0:1

	* raptor/libraptor.3: nroff/man style tweaks

	* raptor/libraptor.3: Style

2003-12-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac: words

	* raptor/configure.ac:
	Make flex version warnings mention N-Triples Plus more

	* librdf/rdf_model_storage.c, librdf/rdf_model.h,
	librdf/rdf_model.c (librdf_model_get_storage): Added, with factory
	method to match - optional.  The model_storage implementation
	implements it.

2003-12-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* utils/rdf-tree.c, utils/rdf-load.c: indented

	* raptor/libraptor.3: bump date

	* raptor/raptor_stringbuffer.c (main):
	Do not free as_string returned strings

	* raptor/raptor_stringbuffer.c: brackets

	* raptor/raptor_stringbuffer.c (raptor_free_stringbuffer):
	Free any constructed string.

	* configure.ac: Add AC_DEFINE for local MD5, SHA1, RIPEMD160

	* utils/rdfproc.c: Tweak help text to look better without getopt_long

	* librdf/rdf_serializer_rdfxml.c:
	Some more unsigned char fixes for URIs
	(librdf_serializer_print_statement_as_rdfxml): Try harder to find a
	maximal length legal property element name; now works for
	foaf:mbox_sha1sum - patch from Morten

	* docs/api.sgml.in, docs/Makefile.am:
	Added storage_mysql storage_tstore

	* raptor/raptor_xml_writer.c:
	Use raptor_stringbuffer to better grow the output cdata.

	* raptor/raptor_stringbuffer.c: stringbuffer now uses unsigned char
	Removed raptor_new|free_stringbuffer_node - used once, now inlined.
	(raptor_stringbuffer_append_string_common): Added with
	common append code merged here.
	(raptor_stringbuffer_append_counted_string,
	raptor_stringbuffer_append_string): Added do_copy arg.
	(main): Test code updated for api changes.

	* raptor/raptor_internal.h:
	Added prototypes for raptor_stringbuffer class to internal API

	* raptor/Makefile.am:
	Re-added raptor_stringbuffer.c and raptor_stringbuffer_test

2003-12-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_uri.c, librdf/rdf_stream.c, librdf/rdf_statement.c, librdf/rdf_query.c, librdf/rdf_node.c, librdf/rdf_model.c, librdf/rdf_iterator.c, librdf/rdf_hash.c:
	More C++ cast fixes mostly near URI and literal strings in test
	code, the main function.

	* librdf/Makefile.am: Added rdf_digest_sha1.c

	* librdf/rdf_digest_sha1.c: SHA1 digest code

2003-12-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rapper.1: Updated for 1.1.0, -a is gone.  Added ntriples-plus

	* raptor/libraptor.3: parser name is ntriples-plus

	* raptor/libraptor.3: Updated for 1.1.0

	* raptor/raptor.h:
	remove raptor_namespaces_end_namespace - does not exist.

2003-12-22  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/Makefile.am:
	Remove raptor_stringbuffer.c/test from dist - not used at present.

	* raptor/configure.ac:
	Added --with-xml-names to choose XML 1.0 name checking (default) or 1.1

	* raptor/raptor_utf8.c (raptor_unicode_is_namestartchar):
	Update to Namespaces in XML 1.1 WD
	http://www.w3.org/TR/2003/PR-xml-names11-20031105/#NT-NCNameStartChar
	and Extensible Markup Language (XML) 1.1 PR
	http://www.w3.org/TR/2003/PR-xml11-20031105/#NT-NameStartChar
	(raptor_unicode_is_namechar): Updated comment, no code changes
	needed.

	* raptor/tests/ntriplesplus/Makefile.am: No check-warn-rdf tests yet

	* raptor/raptor_utf8.c:
	Use RAPTOR_NFC_CHECK to wrap any use of the glib g_utf8_normalize.

	* raptor/configure.ac: Define RAPTOR_NFC_CHECK when NFC check is needed

	* raptor/configure.ac:
	Added --disable-nfc-check to disable Unicode NFC checking even if a
	suitable glib is present and providing it.  Otherwise, autodetects
	as before.

2003-12-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/Makefile.am: Removed REDLAND_LIBS

2003-12-18  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.PL: OSX old perl before 5.8

	* perl/Makefile.PL:
	Another attempt to handle MakeMaker changes since perl5.8

	* perl/Makefile.am:
	Compile perl without the -W flags; it's just too noisy.

	* librdf/redland.spec.in: Don't install perl twice.

2003-12-17  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_serializer_rdfxml.c (rdf_serializer_rdfxml_print_as_xml_content): 
	Redland strings are always UTF-8 and for now, the output encoding
	is always UTF-8 so don't encode high bits, emit them.

	* librdf/rdf_serializer_rdfxml.c: o

	* librdf/rdf_serializer_rdfxml.c:
	(librdf_serializer_rdfxml_serialize_model) Free rdf:RDF qname.

	* raptor/raptor_parse.c, librdf/rdf_parser_raptor.c:
	Update for changed raptor_generate_id handler calls - no const.

	* raptor/raptor_general.c, raptor/raptor.h, raptor/raptor_internal.h:
	raptor_generate_id handler does not take const string

	* raptor/raptor_general.c, raptor/raptor.h, raptor/raptor_internal.h:
	raptor_generate_id returns non const

	* raptor/ntriples_parse.c (raptor_ntriples_parse_line):
	Casts, unsigned char* for blank node IDs.

	* raptor/ntriples_parse.c (raptor_ntriples_parse_line):
	Enforce predicate must be URIref

2003-12-16  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/t/test1.t: Don't use last when not in a loop

	* raptor/ntriples_parse.c (raptor_ntriples_parse_line):
	Pass blank node identifier generation
	through raptor_generate_id.  Rewrite returns into setting rc and jump
	to cleanup to ensure allocated blank node IDs are freed.

	* utils/Makefile.am: Added EXTRA_PROGRAMS to CLEANFILES

	* utils/rdf-load.c, utils/rdf-tree.c:
	Put explicit \n in multiline strings

	* utils/rdf-tree.c:
	Updates from Morten to use openssl lhash - this isn't portable.

	* librdf/rdf_storage_mysql.c:
	(librdf_storage_mysql_context_serialise_next_statement):
	Fixes for over-eager use of multi-line string - add spaces before
	the newlines.
	Correct subject and object node creations.

	* librdf/redland.pc.in: Add @LIBRDF_LDFLAGS@ @LIBRDF_LIBS@ to libs

	* librdf/Makefile.am:
	Use LIBRDF_EXTERNAL_CPPFLAGS, LIBRDF_EXTERNAL_LIBS

	* redland-src-config.in: Put redland cflags, links first

	* redland-config.in: Use LIBRDF_EXTERNAL_CPPFLAGS, LIBRDF_EXTERNAL_LIBS

	* configure.ac: Added LIBRDF_EXTERNAL_CPPFLAGS,
	LIBRDF_EXTERNAL_LIBS used for system raptor

	* raptor/raptor.pc.in: Restore LDFLAGS, LIBS

	* perl/Makefile.PL: Don't need MEM_LIBS in OBJECT

	* utils/rdfproc.1: store to store-name in header

	* utils/rdfproc.c: Added file|uri to parsing help

	* utils/rdfproc.1: Document -n, -s, -t and parsing now turns files
	into URIs if they exist.

	* raptor/raptor-src-config.in: Removed --static-libs

	* utils/rdfproc.c:
	Added -n new - default is not to overwrite, i.e. parsing now appends
	Added -s storage
	Added -t storage-options
	Parsing now turns files into URIs if they exist.

2003-12-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* php/test.php: tidy prints

	* librdf/redland.pc.in: redland.pc.in

	* php/test.php:
	Better test code.  Serializing crashes it - to fix later.

	* php/Makefile.am:
	Reorder compile args so PHP libs are first and overridden if need be.

	* redland-src-config.in: Kill --static-libs
	Use @with_raptor@ to pick either raptor-src-config or raptor-config

	* configure.ac: AC_SUBST with_raptor

	* php/Makefile.am:
	Apply redland_wrap.c.patch when building redland_wrap.c
	Use  -d extension_dir=. for test-php

	* php/redland_wrap.c.patch:
	Patch to PHP wrapper to return empty string not C NULL

	* librdf/Makefile.am: Install redland.pc to libdir/pkgconfig

	* librdf/redland.spec.in: Updates:
	 - require raptor 1.1.0
	 - require libxml 2.4.0 or newer
	 - added pkgconfig redland.pc
	 - split redland/devel package shared libs correctly

	* configure.ac: Added librdf/redland.pc

	* librdf/Makefile.am: Added redland.pc.in

	* raptor/raptor.pc.in: Just link -lraptor

	* raptor/configure.ac: Remove use of have_redland for expat
	sources (no longer shipped with redland anyway)

	* raptor/configure.ac: Remove redland source check.

	* configure.ac, Makefile.am:
	Split SD subdirs into SD_PRE (must be made before librdf)
	and SD_POST (after).  raptor goes into PRE so that it is
	both compiled and installed before librdf.

	* raptor/Makefile.am: No need for librdf.la rule

	* raptor/rdfdump.c, raptor/raptor_stringbuffer.c,
	raptor/raptor_set.c, raptor/raptor_namespace.c,
	raptor/Makefile.am: Remove all RAPTOR_IN_REDLAND code use of
	REDLAND_LIBS, REDLAND_CFLAGS.  Now the test and rdfdump programs
	always just use raptor alone.

2003-12-14  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.am: Remove use of prefix; install where perl wants it

	* perl/Makefile.PL:
	Try again with the MakeMaker PM hash.  Remove double RDF/RDF

	* librdf/rdf_serializer_rdfxml.c (librdf_serializer_rdfxml_terminate):
	Allow raptor_free_namespaces to
	free the entire stack, including pcontext->rdf_ns

	* perl/Makefile.am, perl/update-perl-api.pl: Remove old API updater

	* librdf/rdf_node.h:
	librdf_node_get_literal_value_as_latin1 returning char*

	* librdf/rdf_node.c:
	(librdf_node_get_literal_value_as_latin1) Returning char*

	* examples/example4.c:
	Removed body of code since it was turned into rdfproc.

	* librdf/rdf_parser.h:
	Update librdf_parser_parse_string_as_stream prototype

	* librdf/rdf_model.h: Update librdf_model_query_string prototype

	* librdf/rdf_storage_mysql.c:
	(librdf_storage_mysql_node_hash) Casts for node char const strings.

	* utils/Makefile.am:
	Add mysql-utils target, don't build the mysql utilities by default.

	* utils/rdf-digest.c, utils/rdf-load.c, utils/rdf-tree.c:
	Redland MySQL storage utilities by Morten Frederiksen

	* utils/rdfproc.c:
	Updates for raptor changes - all URI strings are unsigned char*
	Update URI APIs here to match.
	Update all use of literal (UTF-8) strings to unsigned char* also.
	Add more casts to make C++ happy especially near string functions
	and malloc.

	* utils/Makefile.am: Add redland libs, cppflags, cflags

	* librdf/rdf_storage_mysql.c:
	(librdf_storage_mysql_node_hash) Fix to created digest to distinguish
	URIs 'R', literals 'L' and blank node identifiers 'B' in database.
	(librdf_storage_mysql_context_serialise_next_statement): Call error
	functions with the right world value.

	* examples/example4.c, examples/example3.c, examples/example2.c, examples/example1.c, librdf/rdf_uri.h, librdf/rdf_uri.c, librdf/rdf_stream.c, librdf/rdf_storage_mysql.c, librdf/rdf_storage_list.c, librdf/rdf_storage.c, librdf/rdf_statement.h, librdf/rdf_statement.c, librdf/rdf_serializer_rdfxml.c, librdf/rdf_serializer_raptor.c, librdf/rdf_serializer.h, librdf/rdf_serializer.c, librdf/rdf_query_triples.c, librdf/rdf_query.h, librdf/rdf_query.c, librdf/rdf_parser_raptor.c, librdf/rdf_parser.h, librdf/rdf_parser.c, librdf/rdf_node.h, librdf/rdf_node.c, librdf/rdf_model_storage.c, librdf/rdf_model.h, librdf/rdf_model.c, librdf/rdf_iterator.c, librdf/rdf_init.h, librdf/rdf_init.c, librdf/rdf_hash_memory.c, librdf/rdf_hash_bdb.c, librdf/rdf_hash.h, librdf/rdf_hash.c, librdf/rdf_concepts.c:
	Updates for raptor changes - all URI strings are unsigned char*
	Update URI APIs here to match.
	Update all use of literal (UTF-8) strings to unsigned char* also.
	Add more casts to make C++ happy especially near string functions
	and malloc.

	* raptor/raptor_general.c (raptor_print_statement_part_as_ntriples):
	Update call of
	raptor_print_ntriples_string

	* raptor/raptor.h: Update raptor_print_ntriples_string prototype

	* raptor/raptor_general.c (raptor_print_ntriples_string):
	Take unsigned const char*

2003-12-08  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: mysql version reporting

	* configure.ac: make --with-threestore, --with-mysql take CONFIG|yes|no

	* configure.ac: Add BDB 4.2 to the list.
	Fixes for 3store (--with-threestore) else search for 3store-config;
	add glib to cflags/libs using pkg-config.
	Report raptor version.
	Report triple stores found.

2003-12-06  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_uri.h, librdf/rdf_uri.c, librdf/rdf_parser_raptor.c:
	Updates for raptor API use of unsigned char* for URI strings.
	Change redland URIs to take/return unsigned char* for URI strings

	* librdf/rdf_hash.c (librdf_hash_print): replaced fprintf

	* raptor/Makefile.am:
	Narrower yyerrlabl fix since bison 1.875a no longer requires it

	* raptor/examples/grapper.c:
	Casts and updates to use URI strings from raptor as unsigned char*
	and converting to/from gchar*

	* raptor/rdfdump.c, raptor/raptor_xml_writer.c, raptor/raptor_xml.c, raptor/raptor_www_libxml.c, raptor/raptor_www.c, raptor/raptor_uri.c, raptor/raptor_sax2.c, raptor/raptor_rss.c, raptor/raptor_parse.c, raptor/raptor_namespace.c, raptor/n3_parser.y, raptor/n3_lexer.l:
	Lots of casts for str* function args, return values between unsigned
	char* as used for UTF8 and URI strings and char* used by str* functions.
	Some further casts for strings from XML.

	* raptor/raptor_www_test.c (write_bytes_fh): No return value.
	(main) Test code casts unsigned char* for URI strings
	Fix write_content_type and write_bytes_fh set handler methods.

	* raptor/raptor_utf8.c (raptor_unicode_char_to_utf8):
	unsigned char* arg.
	(raptor_utf8_is_nfc): unsigned int, cast for unsigned char*

	* raptor/raptor_sequence.c (raptor_sequence_ensure):
	Use void** not void*.  C++ cares.
	(main): Test code casts to void* for args.

	* raptor/raptor_qname.c (raptor_qname_string_to_uri):
	cast for signed/unsigned int
	comparison.

	* raptor/raptor_locator.c (raptor_format_locator):
	Use raptor_uri_as_counted_string to save a strlen.

	* raptor/raptor_identifier.c:
	(raptor_identifier_print) fputs arg cast [function for RAPTOR_DEBUG only]

	* raptor/raptor_general.c (raptor_parsers_enumerate): Unsigned int.
	(raptor_parse_uri_write_bytes): unsigned char* cast.
	Declare raptor_xml_literal_datatype_uri_string.
	(raptor_print_statement_detailed): Replace several fprintf with fputs
	and fputc.
	(raptor_print_ntriples_string): Use unsigned long/size_t for counts.
	(raptor_statement_part_as_counted_string): unsigned char*
	Use raptor_xml_literal_datatype_uri_string.
	(raptor_statement_part_as_string): unsigned char*
	(raptor_print_statement_part_as_ntriples): unsigned char*.  Replace
	several fprintf with fputs and fputc.
	Use raptor_xml_literal_datatype_uri_string
	(raptor_default_generate_id_handler): Casts for str* functions
	(raptor_check_ordinal): unsigned char*

	* raptor/ntriples_parse.c:
	(raptor_ntriples_generate_statement,raptor_ntriples_string_as_utf8_string,
	raptor_ntriples_parse_line):
	Updated to take/return unsigned char* args and internals.  Update
	raptor_uri calls for similar changes.
	Use raptor_xml_literal_datatype_uri_string
	Lots of casts for str* functions char* arguments.

	* raptor/Makefile.am:
	Remove maintainer only n3 lex/yacc rules (flex/bison)
	Post process the bison output to remove unused label to make g++ happier.

	* raptor/raptor_internal.h:
	Added raptor_xml_literal_datatype_uri_string for the RDF datatype
	literal URI string, used several times.
	raptor_check_ordinal takes unsigned char*
	A couple of lengths, counts become unsigned int.
	raptor_unicode_char_to_utf8, raptor_format_sax2_element,
	raptor_xml_writer_cdata, raptor_xml_writer_comment,
	raptor_xml_writer_as_string changed to
	take/return unsigned char* for UTF8 strings.

	* raptor/raptor.h:
	raptor_new_uri_func, raptor_new_uri_from_local_name_func,
	raptor_new_uri_relative_to_base_func, raptor_uri_as_string_func,
	raptor_uri_as_counted_string_func URI factory methods
	changed to all take/return unsigned char* for URI strings
	raptor_statement_part_as_counted_string,
	raptor_statement_part_as_string,
	raptor_new_uri, raptor_new_uri_from_uri_local_name,
	raptor_new_uri_relative_to_base, raptor_uri_as_string,
	raptor_uri_as_counted_string
	URI methods changed to take/return unsigned char* for URI strings
	raptor_ntriples_string_as_utf8_string changed to return unsigned
	char* for UTF8 string
	raptor_uri_resolve_uri_reference, raptor_uri_filename_to_uri_string,
	raptor_uri_uri_string_to_filename,
	raptor_uri_uri_string_to_filename_fragment, raptor_uri_is_file_uri
	Changed to use unsigned char* for URI strings, char* for filenames

2003-12-01  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_set.c (raptor_set_stats_print): Debug printing tweak.

2003-11-28  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_internal.h:
	raptor_check_ordinal with now unsigned char* arg
	raptor_sax2 content_cdata now unsigned char*

	* raptor/raptor_general.c (raptor_check_ordinal): unsigned char *name

	* raptor/raptor_parse.c: Lots of char/unsigned char casting.

	* raptor/n3_common.h: Undo n3_syntax_error back to raptor_parser arg.

	* raptor/n3_parser.y: (n3_parser_error) aka yyerror; take a void arg.

	* raptor/n3_common.h:
	The n3_syntax_error aka yyerror function takes a void arg.

	* raptor/n3_parser.y: Casts for rdf_parser, strings.

	* raptor/n3_lexer.l:
	Define INPUT_FN as yyinput (C++) or input (otherwise)
	Add more casts for C++ near rdf_parser, yytext.
	(copy_token): More casts for malloc and string functions.
	(copy_string_token): More size_t, casts.

	* raptor/ntriples_parse.c (raptor_ntriples_term):
	Use a size_t; cast for C++.

	* raptor/ntriples_parse.c (raptor_ntriples_term_valid,raptor_ntriples_term):
	Change
	argument names from class to term_class to avoid C++ keyword.

	* raptor/raptor_getopt.h: Protect prototypes for C++

	* raptor/raptor_stringbuffer.c: casts for RAPTOR_*ALLOC returns

	* raptor/n3_parser.y: oops, raptort -> raptor

	* raptor/n3_parser.y: another cast for RAPTOR_MALLOC return

	* raptor/n3_parser.y: casts for RAPTOR_CALLOC returns

	* raptor/configure.ac:
	Move adding pkg-config glib2.0 cppflags/libs till after the other
	libraries since that's more likely to be system wide and least
	effect libxml2, other libraries that might be elsewhere.  Mostly
	affects OSX which is more sensitive to link order.

	* raptor/raptor_stringbuffer.c (main): declare at start of block

	* raptor/configure.ac: Check for libxml/SAX2.h - the new SAX2 API

	* raptor/raptor_libxml.c:
	Added a pile of libxml2_* macros to use the libxml2 SAX2 functions
	when they are present, otherwise the SAX1.  Not using the SAX2 parts
	of the libxml2 API at present.

	* raptor/strcasecmp.c: Add debug message arg casts.

	* raptor/configure.ac:
	Include libxml/parser.h when checking for other libxml2 headers

	* raptor/ntriples_parse.c: Add some debug message arg casts.
	(raptor_ntriples_parse_chunk): At end of input, check that there is
	no remaining junk.

	* raptor/ntriples_parse.c (raptor_ntriples_parse_chunk):
	Fix line counting problems when \r\n
	crosses a buffer or when a line ends at the end of a buffer.
	last_char recorded in state.

	* raptor/raptor_stringbuffer.c (main): Free returned strings.

	* raptor/Makefile.am:
	Added raptor_stringbuffer.c raptor_stringbuffer_test code

	* raptor/raptor.h: Added raptor_stringbuffer.

	* raptor/raptor_stringbuffer.c: Stringbuffer class for growing strings

	* raptor/raptor_libxml.c (raptor_libxml_init):
	With libxml2 use raptor_xml_characters_handler
	for sax->characters.

	* raptor/raptor_parse.c (raptor_cdata_grammar): Added is_cdata arg.
	(raptor_xml_characters_handler): Added, calling raptor_cdata_grammar.
	(raptor_xml_cdata_handler): Updated to call raptor_cdata_grammar with
	is_cdata=1.
	(raptor_xml_parse_init): With expat use raptor_xml_characters_handler
	with XML_SetCharacterDataHandler.
	(raptor_cdata_grammar): Tidy error reporting; do not use
	raptor_xml_writer_as_string for a simple print.

	* raptor/raptor_internal.h:
	Added raptor_xml_characters_handler prototype.

2003-11-26  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.PL:
	List the .pm files explicitly since perl 5.8.1 broke scanning

2003-11-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/manifest.pl: Add withdrawn tests check

2003-11-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/ntriplesplus/Makefile.am:
	Added more N-Triples Plus tests (1-7) and bad (0-3)

	* raptor/tests/ntriplesplus/test-07.out, raptor/tests/ntriplesplus/test-07.ntp, raptor/tests/ntriplesplus/test-06.out, raptor/tests/ntriplesplus/test-06.ntp, raptor/tests/ntriplesplus/test-05.out, raptor/tests/ntriplesplus/test-05.ntp, raptor/tests/ntriplesplus/test-04.out, raptor/tests/ntriplesplus/test-04.ntp, raptor/tests/ntriplesplus/test-03.out, raptor/tests/ntriplesplus/test-03.ntp, raptor/tests/ntriplesplus/test-02.out, raptor/tests/ntriplesplus/test-02.ntp, raptor/tests/ntriplesplus/test-01.out, raptor/tests/ntriplesplus/test-01.ntp:
	More N-Triples Plus tests

	* raptor/tests/ntriplesplus/bad-00.ntp, raptor/tests/ntriplesplus/bad-01.ntp, raptor/tests/ntriplesplus/bad-02.ntp, raptor/tests/ntriplesplus/bad-03.ntp:
	N-Triples Plus bad tests

	* raptor/n3_lexer.l:
	In prefix state, always return to initial on matching '.', then
	error out.
	(n3_syntax_error): Copy removed here.
	(main): Init enough more of parser/locator so that n3_syntax_error
	in main library can be used.

2003-11-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.c: extra newline in version

	* raptor/rdfdump.c: extra newline in help

2003-11-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_parser_raptor.c:
	(librdf_parser_raptor_error_handler,librdf_parser_raptor_warning_handler):
	Do not pass on buffer directly since it is not a format string and
	might contain % such as when it is a URI.  This causes libc to abort
	in the middle of vsnprintf with an assertion failure :/

2003-11-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* demos/ntriples.pl: Style, HTML fixes.
	Counting wasn't right.

	* raptor/tests/Makefile.am:
	Added check-ntriples-plus to check N-Triples Plus with the N-Triples
	tests.

	* raptor/n3_parser.y, raptor/n3_lexer.l, raptor/n3_common.h:
	Added lineno to raptor_n3_parser context to track newlines for dos,
	unix, mac since flex doesn't do that by default with yylineno.

	* raptor/rdfdump.c: tidy help messages

	* raptor/ntriples_parse.c:
	Track newlines correctly for \r\n across chunks; move last_nl into
	ntriples parser context.
	(raptor_ntriples_parse_chunk,raptor_ntriples_parse_start): Use and
	init last_nl in ntriples parser context.

2003-11-16  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h: Added raptor_ntriples_string_as_utf8_string

	* raptor/ntriples_parse.c (raptor_ntriples_term):
	Added allow_utf8 argument and new class
	RAPTOR_TERM_CLASS_FULL which uses all the passed in string.
	(raptor_ntriples_string_as_utf8_string): Added, using
	raptor_ntriples_term as above.
	(raptor_ntriples_parse_line): Use size_t len args.

	* utils/rdfproc.c: Replace a bunch of fprintf(..., "\n") with fputc

	* raptor/rdfdump.c:
	Use f/puts instead of print/fprintf when there are no %s

	* librdf/rdf_heuristics.c (main):
	Make test code silent unless LIBRDF_DEBUG > 1
	Tidy code a little.

	* librdf/rdf_digest.c, librdf/rdf_hash.c:
	Replace a bunch of fprintf(..., "\n") with fputc

	* raptor/raptor_uri.c (raptor_default_new_uri_for_rdf_concept):
	Remove duplicate strlen(base_uri)

2003-11-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_lexer.l: Document more ntriples plus
	Error out if @prefix's name doesn't match name

	* raptor/raptor_utf8.c (raptor_unicode_is_namestartchar):
	Two | changed to ||.  Likely
	worked anyway

2003-11-11  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/libraptor.3: Noted raptor_set_feature with non_nfc_fatal

	* raptor/raptor.h:
	Added RAPTOR_FEATURE_NON_NFC_FATAL to pick between NFC errors or warnings.

	* raptor/raptor_general.c (raptor_set_feature):
	Added RAPTOR_FEATURE_NON_NFC_FATAL to pick
	between NFC errors or warnings.
	(raptor_set_parser_strict): Set feature_non_nfc_fatal default off.

	* raptor/raptor_internal.h:
	Added feature_non_nfc_fatal to pick between NFC errors or warnings.

	* raptor/raptor_parse.c:
	Put 'quotes' around element/attribute names in messages.
	Use feature_non_nfc_fatal to pick between NFC errors or warnings.

	* raptor/tests/Makefile.am: check-bad-nfc-rdf - use strict mode

	* raptor/rdfdump.c: Document -m/--mode arg

2003-11-10  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac: Added tests/ntriplesplus

	* raptor/tests/ntriplesplus/Makefile.am: N-Triples plus tests

	* raptor/tests/ntriplesplus/test-00.ntp, raptor/tests/ntriplesplus/test-00.out:
	N-Triples Plus default namespace test

	* raptor/tests/Makefile.am: added ntriplesplus dir

	* raptor/raptor_qname.c (raptor_qname_string_to_uri):
	Keep original name around for error reporting.

	* raptor/rdfdump.c, raptor/n3_parser.y: parser now called ntriples-plus

	* raptor/n3_parser.y: parser called ntriplesplus

	* raptor/n3_parser.y (production directive):
	Fix declaring default namespace prefix

2003-11-09  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am: added bad-20.rdf

	* raptor/tests/bad-20.rdf:
	check non-namespaced element does not crash parser

	* raptor/raptor_qname.c (raptor_new_qname,raptor_new_qname_from_namespace_local_name):
	Do not
	die if no URI for qname is available.  It might be <a xmlns="">
	which is at least needed for some error reports or for embedded
	qnames.  Caused unnecessary crashes when parsing failed.

2003-11-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac:
	Added --with-mysql-config and automatic search for mysql_config program.
	Added STORAGE_OBJS

	* librdf/rdf_storage_mysql.c, librdf/rdf_storage_mysql.h:
	RDF Storage in MySQL DB

	* AUTHORS: Added Morten as MySQL store author

	* librdf/rdf_storage.h, librdf/rdf_storage.c: Added mysql store

	* librdf/Makefile.am: Added STORAGE_OBJS
	Added rdf_storage_mysql.c rdf_storage_mysql.h

	* librdf/librdf.h: Added LIBRDF_ERROR3

	* raptor/raptor_xml.c (main): Make tests less chatty on success

	* raptor/raptor_uri.c (main): Make it less chatty on success

	* raptor/raptor_uri.c (raptor_uri_uri_string_to_filename_fragment):
	Allow file://a|/ and file://a:/
	(main): For WIN32, check the above works.

	* librdf/rdf_parser_raptor.c:
	Add mapping of rdf:nodeID values to redland generated IDs.
	(librdf_parser_raptor_init): Create the userid->redlandID hash map.
	(librdf_parser_raptor_terminate): Added to delete the hash map.
	(librdf_parser_raptor_generate_id_handler): Do the mapping.

	* raptor/autogen.sh, autogen.sh:
	remove ltmain.sh libtool before libtoolize

	* python/redlandtest.py (class RedlandGetsCase):
	Add test for making a new anon (no arg) blank node

	* raptor/manifest.pl:
	Emit test-results.rdf in the style of http://www.w3.org/2003/08/owl-systems/test-results-out
	for machine processing of test runs

	* raptor/raptor_general.c (raptor_parse_uri_with_connection):
	Fail before parsing if
	raptor_www_fetch failed.

	* raptor/raptor_internal.h:
	Added raptor_www_error_varargs internal prototype.

	* raptor/raptor_www_libxml.c (raptor_www_libxml_http_error):
	This was just all wrong, printing to
	stderr and then exit(1).  Change to use the proper
	raptor_www_error_varargs callback.

	* raptor/raptor_www.c (raptor_www_error_varargs):
	Added, with va_list signature.

	* raptor/raptor_general.c (raptor_parse_file):
	When uri is given and base_uri is NULL, copy
	the uri and free it later - fix to match the function documentation.

2003-10-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_general.c (raptor_parse_file):
	fclose only when fh is not NULL

2003-10-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_serializer_rdfxml.c (librdf_serializer_rdfxml_terminate):
	Free namespace before stack.

	* librdf/rdf_serializer_rdfxml.c:
	Use updated raptor_(new|free)_namespaces
	(librdf_serializer_rdfxml_init, librdf_serializer_rdfxml_terminate):
	Make the rdf namespace
	(librdf_serializer_rdfxml_serialize_model): declare rdf:RDF
	using raptor_new_qname_from_namespace_local_name

	* raptor/raptor_parse.c (raptor_xml_end_element_handler):
	For RAPTOR_DEBUG, declare
	element_name at start of function.

2003-10-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h:
	raptor_namespace(s)_(new|free) renamed to raptor_(new|free)_namespace(s)
	Added raptor_new_qname_from_namespace_local_name

	* raptor/raptor_qname.c (raptor_new_qname_from_namespace_local_name):
	Added.

	* raptor/raptor_namespace.c:
	raptor_namespaces_(new|free) renamed to raptor_(new|free)_namespaces

	* raptor/raptor_namespace.c:
	raptor_namespace_(new|free) renamed to raptor_(new|free)_namespace

	* librdf/rdf_serializer_rdfxml.c:
	Added a raptor_namespace_stack to the context.
	(librdf_serializer_rdfxml_init): Updated with stack
	(librdf_serializer_rdfxml_terminate): Added, to free stack.
	(rdf_serializer_rdfxml_print_as_xml_attribute): Deleted.
	(librdf_serializer_print_statement_as_rdfxml): Do less work in
	finding max XML-name character for a property URI.
	Replace use of rdf_serializer_rdfxml_print_as_xml_attribute
	with calls to raptor_xml_escape_string.
	(librdf_serializer_rdfxml_serialize_model): Use namespace stack
	to declare rdf namespace - initial code.
	(librdf_serializer_rdfxml_register_factory): Add terminate call.

	* raptor/raptor_xml_writer.c:
	raptor_namespaces_free renamed to raptor_namespaces_clear

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Use raptor_new_sax2_element.
	raptor_namespaces_free renamed to raptor_namespaces_clear

	* raptor/raptor_internal.h: Added raptor_new_sax2_element

	* raptor/raptor_sax2.c (raptor_new_sax2_element): Added.

	* raptor/raptor.h:
	Added prototypes for raptor_namespaces_new, raptor_namespaces_clear

	* raptor/raptor_namespace.c
	(raptor_namespaces_new,raptor_namespaces_free): 
	Added for constructor and destructor.
	raptor_namespaces_free renamed to raptor_namespaces_clear for
	emptying a statically allocated namespace stack.

	* raptor/n3_parser.y:
	raptor_namespaces_free renamed to raptor_namespaces_clear

	* librdf/rdf_serializer_rdfxml.c
        (librdf_serializer_rdfxml_raptor_error_handler): Added.
	(rdf_serializer_rdfxml_print_xml_attribute): Take world arg.
	Use raptor_xml_escape_string to do the escaping.
	(librdf_serializer_print_statement_as_rdfxml): pass world arg to
	calls to rdf_serializer_rdfxml_print_xml_attribute

	* librdf/rdf_utf8.h: Make librdf_unicode_char_to_utf8 match code

2003-10-16  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y: Moved n3 lexer/parser stuff to n3_common.h
	Added uri field to union.
	URI_LITERAL and QNAME_LITERAL now are uri.
	PREFIX now expects an IDENTIFIER to follow.
	Remove all make qname/uri and free(copied token) sequences since the
	lexer does it.
	Added fake yy_init_globals to stop dumb warning.

	* raptor/n3_lexer.l:
	Make raptor_uris here from lexer tokens (qnames or URIs)
	For @prefix, recognise following token as an identifier specially
	rather than try to make it a URI.
	(n3_token_free): Free string or raptor_uri.
	(main): Lots of fixups to fake enough n3 parser structure to
	get it working fairly standalone.

	* raptor/Makefile.am: Added n3_common.h
	Fixup free of null inside flex-generated lexer.

	* raptor/n3_common.h: N3 parser/lexer shared internals

	* raptor/raptor_internal.h: Moved n3 lexer/parser stuff to n3_common.h

	* raptor/n3_parser.y (n3_qname_to_uri):
	Replace with call to raptor_qname_string_to_uri
	and added length parameter.

	* raptor/raptor_qname.c (raptor_new_qname):
	Replace raptor_namespace_local_name_to_uri
	with use of raptor_new_uri_from_uri_local_name.
	(raptor_qname_string_to_uri): Added, making only the URI
	equivalent to the qname and handling N3/RDQL-style special
	cases such as "prefix:", ":" and NULL.

	* raptor/raptor_internal.h, raptor/raptor.h:
	Moved raptor_qname, raptor_namespace, raptor_namespace_stack
	classes into public API.
	Added raptor_qname_string_to_uri

	* raptor/raptor_namespace.c (raptor_namespace_local_name_to_uri):
	Removed - only used once
	internally and was never public.

2003-10-09  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_lexer.l (n3_token_free): Added, for cleanup in debugging.
	(main): Init and clear token/lval.

	* raptor/tests/Makefile.am: Added ex-53

	* raptor/tests/ex-53.out, raptor/tests/ex-53.rdf:
	Check allowing optional rdf:RDF

	* raptor/rdfdump.c:
	--assume/-a feature_assume_is_rdf deleted; rdf:RDF is optional.

	* raptor/raptor_general.c, raptor/raptor_parse.c:
	feature_assume_is_rdf deleted; rdf:RDF is optional.

	* raptor/raptor_internal.h: feature_assume_is_rdf deleted

2003-10-08  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_node.c: Fix debug calls when LIBRDF_DEBUG > 1

2003-10-06  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/wine.out, raptor/tests/wine.rdf:
	OWL Wine Ontology from http://www.w3.org/TR/owl-guide/wine.rdf

	* raptor/tests/Makefile.am:
	Added OWL Wine ontology from http://www.w3.org/TR/owl-guide/wine.rdf
	as wine.rdf wine.out

2003-09-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y:
	Remove n3_parser_lex; re-#define yylex to call direct

	* raptor/n3_lexer.l (copy_string_token):
	Destroy malloced string on error return.

	* raptor/n3_parser.y (n3_parse):
	Don't delete buffer, pop buffer state; a successful lex
	does that.

	* raptor/raptor_internal.h: Remove n3_token_print

	* raptor/n3_parser.y: Use reentrant yacc parser.
	Store the lexer lval in the n3_parser context.
	Lots of #define trickery to get flex/bison to talk nicely.
	Make n3_parser_error take an rdf_parser arg (this isn't configurable
	by bison itself, so is likely fragile).
	Remove use of extern in lineno; get it from the lexer.
	Remove N3_Parser global; use rdf_parser local.
	(n3_parser_error): Update for having rdf_parser arg, update locator
	lineno from scanner.
	(n3_syntax_error, n3_qname_to_uri): Get lineno from scanner.
	(n3_parse): Remove fixmes, no need for protecting globals.
	(main): Update for reentrant parser; init locator from standalone
	args.

	* raptor/n3_lexer.l:
	Remove n3_lexer.c/.h prototypes no longer(?) needed with re-entrant lexer.
	Remove use of lineno; let lexer do it.
	Change lexer call to pass in lval from reentrant parser.
	(n3_token_print): Pass in lval.
	(main): Update for api changes.

2003-09-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_internal.h: Updates for reentrant lexer.

	* raptor/n3_parser.y: Use reentrant lexer API.
	Define YYLEX_PARAM to be scanner arg, from current grammar.
	(n3_parser_lex): Take scanner arg.
	(n3_syntax_error): Add rdf_parser arg.
	(n3_parse): Init and destroy reentrant lexer.
	(raptor_n3_parse_terminate): Tidy up any lexer stuff.
	(main): Check for file not found, report it.

	* raptor/n3_lexer.l:
	Switch to reentrant lexer.  Pass rdf_parser into code, yyextra
	internally.
	(yywrap): Add scanner arg.
	(copy_string_token, n3_syntax_error): Add rdf_parser arg.
	(main): Use reentrant calls for lexer to set yyin, get_text.  Use
	yylex_init/yylex_destroy.

	* raptor/Makefile.am: n3_lexer_test depends on raptor_utf8

	* raptor/raptor_parse.c:
	Add EXPAT_UTF8_BOM_CRASH fix updates for sax2 changes.

	* raptor/configure.ac: Tweak for old flex version output

	* raptor/configure.ac: Try to check flex is new enough.

2003-09-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: 3store linking fixes.

	* librdf/rdf_storage_tstore.c: comment out unused context

	* configure.ac: Bumped raptor min to 1.1.0
	Added --with-threestore.

	* librdf/rdf_storage.c (librdf_init_storage):
	Call librdf_init_storage_tstore if HAVE_TSTORE
	(librdf_storage_add_statements): Do add statements over a stream here
	if lower level just has add_statement method.

	* librdf/rdf_storage_tstore.h: RDF Storage using 3store

	* librdf/Makefile.am: Added rdf_storage_tstore.c rdf_storage_tstore.h

	* librdf/rdf_storage_tstore.c: RDF Storage using 3store

2003-09-22  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_init.h: Added internal librdf_error_varargs

	* librdf/rdf_init.c (librdf_error_varargs): Added, internal

2003-09-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y: Minor C reformatting

2003-09-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h: Added raptor_parsers_enumerate prototype

	* raptor/n3_lexer.l: minor reformatting

	* raptor/n3_lexer.l (copy_string_token): Make \r, \n and \t work

	* raptor/raptor_general.c (raptor_init):
	Ensure rdf/xml is default parser.
	(raptor_parsers_enumerate): Added, to enumerate parsers, returning
	their name & label.

	* raptor/examples/grapper.c:
	Use raptor_parsers_enumerate to get parser names, labels.

	* raptor/n3_lexer.l: flex archaeology for options

	* raptor/examples/Makefile.am: Don't build examples by default

2003-09-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_www.c, raptor/raptor_general.c:
	Revert to old API for raptor_uri_uri_string_to_filename

	* raptor/raptor.h, raptor/raptor_uri.c
	(raptor_uri_uri_string_to_filename): Restored to old API.
	(raptor_uri_uri_string_to_filename_fragment): Added with fragment arg.

	* librdf/rdf_hash_bdb.c: Use bdb->set_flags only when present

	* configure.ac: Check for BDB DB->set_flags

2003-09-17  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y (n3_qname_to_uri):
	Handle NULL (":" in N3) returning the default
	namespace.  It's not quite clear if this is legal.

	* raptor/n3_parser.y (raptor_n3_generate_statement):
	Do nothing if some part of the triple
	is NULL.

	* raptor/raptor_internal.h: n3_syntax_error now takes varargs

	* raptor/n3_parser.y (n3_syntax_error): Now takes varargs

	* raptor/n3_lexer.l: n3_syntax_error now takes varargs
	(copy_string_token): Added \u, \U.  Fixed, \r, \n, \t

	* raptor/n3_parser.y: Wrap a debugging printf

	* raptor/n3_lexer.l:
	For blank literal "_:abc", don't include _: in the id passed to the parser.

	* raptor/n3_parser.y:
	Throughout replace raptor_new_uri with raptor_copy_uri when copying
	existing base URI.

	* raptor/n3_parser.y:
	Throughout: Handle NULL uri string meaning use the base URI

	* raptor/n3_lexer.l:
	Handle <> (returning NULL) as well as <> with content.
	(n3_token_print): Update to match.
	(n3_syntax_error): Simple standalone copy here.

	* raptor/n3_parser.y (n3_parser_error):
	Set lineno and call raptor_parser_simple_error to
	pass on the parsing error.
	(n3_syntax_error): Added. Set lineno and call raptor_parser_error
	to pass on a general syntax error.
	(n3_qname_to_uri): Init locator line before calling raptor_new_qname
	that may fail, calling raptor_parser_simple_error.

	* raptor/raptor_internal.h: Add n3_syntax_error

	* raptor/n3_lexer.l: Call n3_syntax_error on a syntax error

	* raptor/examples/grapper.c: Add n3 syntax

	* raptor/n3_parser.y (propertyList): Handle NULL verb, two cases.
	(raptor_n3_parse_start): No locator column, byte values just yet.

	* raptor/n3_parser.y (n3_qname_to_uri):
	Call raptor_new_qname with rdf_parser for errors

2003-09-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y:
	Don't raptor_free_uri shared uris made from qnames

	* raptor/n3_parser.y (n3_parse): Tidy up flex buffers.
	(raptor_n3_parse_terminate): Destroy any flex state on exit.

	* raptor/n3_parser.y (n3_parse): delete buffer after parse.

	* raptor/n3_parser.y: Free uri strings returned from URI_LITERAL.

	* raptor/raptor_identifier.c (raptor_new_identifier):
	Note uri, literal_datatype are shared and not copied.

	* raptor/n3_parser.y: Free strings returned from QNAME_LITERAL.

	* raptor/raptor_sequence.c (raptor_new_sequence): Use RAPTOR_MALLOC.
	(raptor_free_sequence): Free the raptor_sequence.

	* raptor/raptor_identifier.c (raptor_new_identifier):
	Note id, literal, literal_language are
	shared and not copied.

	* raptor/n3_parser.y (statement): Free identifier used for subject
	(propertyList): Free identifier used for verb

	* raptor/n3_parser.y:
	Track when an identifier is copied using is_malloced
	(raptor_free_triple): Actually free the triple.

	* raptor/raptor_internal.h, raptor/n3_parser.y:
	raptor_triple now just has 3 items.

	* raptor/n3_parser.y (raptor_n3_parse_terminate): Free namespaces

	* raptor/n3_parser.y (n3_parse): Do not parser NULL or empty string.
	(raptor_n3_parse_chunk): Do not parser empty buffer.

	* raptor/rdfdump.c: (main) Don't free NULL uri

	* raptor/n3_parser.y: More debugging messages.
	Recover from errors, don't generate partial triples.
	(n3_qname_to_uri): Return NULL if raptor_new_qname does not give a
	URI (some error happened).

	* raptor/n3_lexer.l: Remove END token, should only use EOF
	Count lines right for \r\n|\r|\n
	Handle EOF in comments

	* raptor/n3_parser.y: Remove END token, should only use EOF
	Add more debugging statements.
	Throughtout, change raptor_new_uri to raptor_new_uri_relative_to_base.
	(statement): Handle empty propertyList ("[]").
	(propertyList 1): Copy verb into objectList, then append
	propertyList items (if not empty "[]" again).
	(objectList): Add empty item for "[]", returning NULL.
	(resource): For [], handle NULL and generate statements here
	before returning the generated id.
	(n3_parser_print_statement): Added for test code
	(main): In test code, init URI module only, create fake
	static rdf_parser and n3_parser and initialise enough (base URI,
	and context) so that it works.

	* raptor/n3_parser.y: Add error recovery at '.'

2003-09-14  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/n3_parser.y:
	raptor_print_triple renamed to raptor_triple_print

	* raptor/n3_parser.y: Only define raptor_print_triple if debugging.

	* raptor/raptor_identifier.c, raptor/n3_parser.y:
	raptor_print_identifier renamed to raptor_identifier_print

	* raptor/n3_parser.y: Raptor N-Triples+/N3 parser

	* raptor/n3_lexer.l: Raptor N-Triples+/N3 lexer

	* raptor/raptor_identifier.c (raptor_identifier_print):
	Added for debugging

	* raptor/raptor_internal.h: When debugging, raptor_identifier_print

	* raptor/raptor_identifier.c (raptor_new_identifier):
	Add literal, literal_datatype,
	literal_language args and handle them.
	(raptor_init_identifier): Deleted, not used (enough).
	(raptor_copy_identifier, raptor_free_identifier): Updated for
	literal, literal language and literal datatype.

	* raptor/rdfdump.c:
	Use the syntax name (after validation) to intialise the parser rather
	than an ever-growing set of flags.

	* raptor/raptor_uri.c (raptor_default_new_uri):
	If the filename had a fragment, re-append
	it to the file:URI after updating it to be correct.
	(raptor_uri_uri_string_to_filename): Add fragment_p arg to return
	the URI fragment after a discovered filename in a file:URI.
	(assert_uri_to_filename): Update call to
	raptor_uri_uri_string_to_filename.

	* raptor/raptor_parse.c (raptor_xml_parser): Add namespaces.
	Elsewhere change rdf_parser->namespaces to
	rdf_xml_parser->namespaces.
	(raptor_xml_parse_start): Initialise the namespaces for rdf/xml.

	* raptor/raptor_sequence.c: Raptor sequence ADT

	* raptor/raptor_namespace.c (raptor_namespace_init):
	Add defaults arg to control which
	namespaces are added by default. 0=none, 1=xml, 2=... others
	(main): Update test code to give new arg.

	* raptor/raptor_general.c (raptor_init): Call raptor_init_parser_n3.
	(raptor_start_parser, raptor_free_parser): Remove namespace code from
	here; moves into specific parser context code.
	(raptor_parse_file): Update for raptor_uri_uri_string_to_filename
	extra arg.

	* raptor/raptor_xml_writer.c:
	(raptor_new_xml_writer) Update for raptor_namespaces_init defaults
	arg.

	* raptor/raptor_rss.c (raptor_rss_emit):
	Use raptor_new_identifier rather than
	raptor_init_identifier and make items dynamically allocated.

	* raptor/raptor_www.c:
	(raptor_www_file_fetch) Update for raptor_uri_uri_string_to_filename
	extra argument.

	* raptor/raptor.h:
	Add literal, literal_datatype, literal_language to raptor_identifier
	structure.
	Add above arguments to raptor_new_identifier.
	Remove raptor_init_identifier - not used.
	Add fragment_p argument to raptor_uri_uri_string_to_filename.

	* raptor/raptor_internal.h:
	Remove namespaces from raptor_parser; now in per-syntax contexts.
	Updated raptor_namespaces_init to take defaults arg.
	Added N3 class prototypes, for n3_token_print, raptor_init_parser_n3,
	n3_parser_lex.
	Added raptor_triple structure.
	Added sequence class prototypes.

	* raptor/Makefile.am:
	Add n3_lexer.c n3_lexer.h both generated from n3_lexer.l by flex;
	add maintainer-only rules to do that.
	Add n3_parser.tab.c n3_parser.tab.h generated from n3_parser.y by
	yacc; add maintainer-only rules to do that.
	Add raptor_sequence.c and test.

	* raptor/configure.ac: Add lex (flex required) and yacc

2003-09-12  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Redland.i:
	Add prototypes for internal functions librdf_error, librdf_warning
	along with warning that these aren't in the API.

2003-09-11  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* python/RDF.py (Model): Added sync() calling librdf_model_sync

	* python/redlandtest.py (RedlandModelGeneral): Test model.sync()

2003-09-08  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* python/RDF.py:
	(Node:_get_literal_value) Use dt_uri=Uri(string=... not (uri_string=...

	* configure.ac: Bumped version to 0.9.15

	* Snapshotted redland_0_9_14 for 0.9.14 release

	* librdf/redland.spec.in: Require raptor 1.0.0


	* librdf/rdf_hash_bdb.c (librdf_hash_bdb_open):
	Restore previous call to db_create.
	Add an extra clause to do nothing inside big #ifdef chain
	if HAVE_DB_CREATE is present.

	* configure.ac: Min raptor version is now 1.0.0

	* raptor/configure.ac: Bumped version to 1.1.0

	* Snapshotted raptor_1_0_0 for 1.0.0 release

	* raptor/libraptor.3: new date

	* raptor/ntriples.h,raptor/Makefile.am: Removed old header ntriples.h

	* raptor/raptor_general.c, raptor/raptor.h, raptor/ntriples_parse.c:
	Removed deprecated functions as promised. Changes are described in
	the libraptor.3 man page.

	* raptor/configure.ac: Updated for Raptor 1.0.0
	Shared library soname major now 1

	* configure.ac:
	With only --with-bdb=PREFIX given, set the lib & include from that
	value.
	Use 'unknown' rather than none when some BDB bits given and no
	guessing.

	* raptor/win32_config.h: Added R_OK define for access()

	* raptor/tests/Makefile.am: test wording for failures

	* librdf/rdf_hash_bdb.c (librdf_hash_bdb_open):
	Probably restore early DB3.0 to working with db_create.

2003-09-05  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am: Added bad-05.nt

	* raptor/tests/bad-05.nt: Bad Unicode character #x110000

	* raptor/tests/test.out, raptor/tests/test.nt: fixes

	* raptor/tests/test.nt:
	Removed resource18-20 - illegal Unicode chars.  Added \U0010FFFF

	* raptor/ntriples_parse.c (raptor_ntriples_term):
	Forbid Unicode characters outside #x0-#x10FFFF

2003-09-04  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* java/org/librdf/redland/World.java: Use librdf_version/copyright*

	* Redland.i: Use librdf_version* and both librdf_copyright strings

	* tcl/Makefile.am: clean more

	* raptor/libraptor.3: Updated for 1.0.0
	Added raptor_parse_file_stream.
	Added new feature normalize_language.
	Added list of all static variables exported.

	* configure.ac: Remove a bunch of unused LIBRDF_VERSION_ AC_SUBST.
	Remove LIBRDF_VERSION_INFO - not used
	Define LIBRDF_VERSION_DECIMAL here and make it an AC_SUBST.

	* librdf/rdf_init.c: Use LIBRDF_VERSION_DECIMAL define.

	* raptor/configure.ac:
	Define RAPTOR_VERSION_DECIMAL here and make it an AC_SUBST.

	* raptor/raptor_general.c: Use RAPTOR_VERSION_DECIMAL define.

	* redland-config.in: Added --version-decimal.

	* redland-config.1, raptor/raptor-config.1:
	Document --version-decimal and --libtool-libs

	* raptor/raptor-config.in: Added --version-decimal.

	* raptor/rdfdump.c: Allow filename "-" to be used as standard input. 
	When a filename is given, use raptor_parse_file.
	Adjust the error messages to mention file names when using
	raptor_parse_file

	* raptor/raptor.h: Added raptor_parse_file_stream

	* raptor/raptor_general.c (raptor_parse_file_stream):
	Added, allowing passing in of an existing
	FILE* stream (with optional filename) and parsing rather than
	raptor doing the fopen/fclose.
	(raptor_parse_file): A NULL uri argument now means stdin.

	* raptor/raptor_internal.h, raptor/raptor_parse.c:
	Remove rdf_parser->fh

	* utils/rdfproc.1:
	Update parse, serailize commands to note optional syntax names, list
	valid names and mention the default RDF/XML.

	* utils/rdfproc.c:
	Make syntax name option to parser commands, defauting to (default
	parser) rdf/xml
	Hide 'query' command further.
	Print short copyright in usage, help.
	Tidy help messages.
	Delete rotten parser features

	* librdf/rdf_parser_raptor.c (librdf_parser_raptor_constructor):
	Register "rdfxml" name and mime
	type, URL http://www.w3.org/TR/rdf-syntax-grammar.
	Keep raptor name around

	* raptor/rdfdump.c:
	Use raptor_short_copyright_string in usage/help messages

	* raptor/raptor_general.c, raptor/raptor.h:
	Added raptor_short_copyright_string

	* utils/rdfproc.c: Free options hash

	* librdf/redland.spec.in: Add rdfproc.
	Require raptor 0.9.12 min

	* utils/rdfproc.1: tweak

	* utils/Makefile.am: Added rdfproc.1

	* utils/rdfproc.1: rdfproc manual page

	* utils/rdfproc.c: help

	* utils/rdfproc.c:
	Tidy usage message - use triples & graphs, explain source, target, arc.
	Add short copyright

	* librdf/rdf_init.c, librdf/librdf.h:
	Added librdf_short_copyright_string.
	Made long one have line breaks and more URIs

	* utils/rdfproc.c: statements now find
	Added optional CONTEXT to add/add-typed/remove, warning if not enabled.

	* utils/rdfproc.c: Use rdfproc_getopt.h

	* configure.ac:
	Enable public domain getopt in getopt.c & rdfproc_getopt.h for
	rdfproc if there is no getopt or getopt_long

	* utils/Makefile.am:
	Enable public domain getopt in getopt.c & rdfproc_getopt.h

	* utils/getopt.c, utils/rdfproc_getopt.h:
	public domain getopt for rdfproc

	* configure.ac: Look for getopt.h

	* librdf/rdf_hash.h: Added librdf_hash_put_strings

	* librdf/rdf_hash.c (librdf_hash_put_strings):
	Added helper to add strings to hash easier.

	* librdf/rdf_storage.c (librdf_new_storage_with_options):
	Added, taking a librdf_hash
	argument.  Similar to librdf_new_model_with_options.

	* librdf/rdf_storage.h: Added librdf_new_storage_with_options

	* librdf/rdf_init.c:
	Renamed public statics to librdf_version* and librdf_copyright_string.
	Added librdf_version_decimal.

	* librdf/librdf.h: Added REDLAND_API
	Added REDLAND_DEPRECATED
	Added public statics librdf_version* and librdf_copyright_string.

	* utils/Makefile.am: Added rdfproc

	* utils/rdfproc.c:
	Redland RDF processor - converted from examples/example4.c

2003-09-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Makefile.PL: Updated for 0.9.13
	Uses configure.ac
	Pick the perl prefix to pass to configure.
	make clean needs a little help else it just returns 'exit 2'.
	install: Installs just the shared libraries for raptor and
	redland, then perl.
	Have a hand-crafted MANIFEST and MANIFEST.SKIP now.

	* raptor/raptor_general.c (raptor_set_feature):
	Add new feature normalize_language
	(raptor_set_parser_strict): Set default for feature
	normalize_language to true.

	* raptor/raptor.h, raptor/raptor_internal.h:
	Add new feature normalize_language

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Normalize language to lowercase.
	After http://www.w3.org/TR/rdf-concepts/#dfn-language-identifier
	Controlled by a new parser feature 'normalize_language'.

	* raptor/ntriples_parse.c (raptor_ntriples_parse_line):
	Normalize language to lowercase.
	After http://www.w3.org/TR/rdf-concepts/#dfn-language-identifier

2003-09-01  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/Makefile.am:
	In AM_LDFLAGS, use @LIBRDF_LIBS@ for -L/-l libraries and
	@LIBRDF_INTERNAL_LIBS@ for libtool things.
	Remove not-needed $(LIBS) from test binaries link lines.

	* examples/Makefile.am:
	In AM_LDFLAGS, use @LIBRDF_LIBS@ for -L/-l libraries and
	@LIBRDF_INTERNAL_LIBS@ for libtool things.

	* Makefile.am: raptor subdirectory is now not always built

	* configure.ac: TIdy some configure --with/--enable messages
	Added RAPTOR_CONFIG program (default raptor-config)
	Added --with-raptor=system or internal.  By default will pick the best.
	Checks for minmum versions and warns if too old, but proceeds.
	raptor subdirectory is now not always built.

	* raptor/raptor-config.in: Oops, -lraptor with --libs

	* redland-config.in:
	Added --libtool-libs for compiling with librdf using libtool.

	* configure.ac:
	Added LIBRDF_LIBTOOLLIBS for compiling with librdf using libtool.

	* raptor/configure.ac:
	Added RAPTOR_LIBTOOL_LIBS for compiling with raptor using libtool.

	* raptor/raptor-config.in:
	Added exec_prefix to make --libs generate the right -L
	Added --libtool-libs for compiling with raptor using libtool.

2003-08-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac, raptor/configure.ac:
	dmalloc enabled only if dmalloc.h is present

	* raptor/tests/Makefile.am: Removed warn-01 re-added accidently.

	* raptor/tests/warn-00.out, raptor/tests/warn-00.rdf:
	Added rdf:bagID warning check

	* raptor/tests/Makefile.am:
	Added scanning tests and ex-52.svg/out for inside SVG

	* raptor/tests/ex-52.out, raptor/tests/ex-52.svg:
	Check scanning for rdf/xml in SVG

	* raptor/raptor_general.c (raptor_set_parser_strict):
	Scanning and assuming are never default
	on, must be enabled

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Fix scanning for rdf:RDF.  Do
	parent->child processing if the grammar has a state set up,
	in this case it is expecting a list of node elements.

	* raptor/rdfdump.c: Set strict before setting other features

	* raptor/rdfdump.c: Use strict_mode

	* raptor/tests/warn-00.out, raptor/tests/warn-00.rdf,
	raptor/tests/warn-01.out, raptor/tests/warn-01.rdf,
	raptor/tests/warn-03.rdf: These are now errors not wrarnings

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	 Non-namespaced elements are now
	an error.
	(raptor_process_property_attributes): Tidy non-namespaced element name.
	(raptor_start_element_grammar): Give errors if an attempt is made
	to proceed dealing with elements with no namespace for property or
	node elements - attributes are caught above.

	* raptor/tests/Makefile.am:
	Add bad-18.rdf, bad-19.rdf for non-namespaced elements

	* raptor/tests/bad-18.rdf, raptor/tests/bad-19.rdf:
	Test node/property elements without namespaces fail

	* raptor/raptor_internal.h:
	Add raptor_parser field 'magic' for libxml2 error/warning callback
	validation.  and declare RAPTOR_LIBXML_MAGIC to set use there

	* raptor/raptor_general.c (raptor_new_parser):
	Set RAPTOR_LIBXML_MAGIC field in structure for
	libxml2 error/warning callback validation.

	* raptor/ntriples_parse.c (raptor_ntriples_term):
	Check that the string/URI term was terminated
	before the end of the string.

	* raptor/tests/Makefile.am: Added bad-04.nt

	* raptor/tests/bad-04.nt: Test for non-terminated URI

	* raptor/raptor_libxml.c (raptor_libxml_warning,raptor_libxml_error):
	Validate the ctx pointer
	returned since sometimes it is a ctx, sometimes ctx->userData.  The
	latter is what is expected.

	* raptor/tests/Makefile.am:
	Oops, run bad ntriples tests in N-Triples mode

	* raptor/rapper.1: Added --version/-v

	* raptor/rdfdump.c: Tidied up option error handling, messages.
	Added --version/-v

	* librdf/rdf_storage.c:
	(librdf_storage_add_statement, librdf_storage_add_statements,
	librdf_storage_remove_statement,
	librdf_storage_context_add_statement,
	librdf_storage_context_remove_statement)
	For these optional methods that can change the store, return failure
	if the factory does not implement them.
	(librdf_storage_context_add_statements): Added either using the
	factory method or librdf_storage_context_add_statement if that
	is missing.
	(librdf_storage_context_remove_statements): Added either using the
	factory method or librdf_storage_context_remove_statement if that
	is missing.
	(librdf_storage_context_as_stream): Added, deprecating
	librdf_storage_context_serialise.

	* librdf/rdf_storage.h:
	Added new context_add_statements and context_remove_statements
	factory methods.  Note that those are optional and so are
	all other context methods or methods that change the store (could be
	read only)
	Added prototypes for librdf_storage_context_add_statements,
	librdf_storage_context_remove_statements.
	Added prototype for librdf_storage_context_as_stream, deprecating
	librdf_storage_context_serialise.

	* librdf/rdf_model_storage.c:
	(librdf_model_storage_context_add_statements) Added, calling
	librdf_storage_context_add_statements
	(librdf_model_context_remove_statements) Added, calling
	librdf_storage_context_remove_statements

	* librdf/rdf_model.c:
	(librdf_model_context_add_statements) Use factory method if
	present otherwise use existing code and
	librdf_model_context_add_statement method.
	(librdf_model_context_remove_statements) Use factory method if
	present otherwise use existing code and
	librdf_model_context_remove_statement method.

	* librdf/rdf_model.h: model_factory - note sync is optional.
	Add context_add_statements and context_remove_statements, both
	optional that librdf_model will implement using the singular forms if
	they are missing.

	* librdf/rdf_node.c (librdf_node_equal): Remove unused var

2003-08-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_stream.c (librdf_stream_update_current_element):
	Do not do excessive work -
	only do a lower level check is_updated is false.
	(librdf_stream_next): Reset is_updated after an
	stream->next_method is performed.
	(librdf_stream_get_object): Do less work, just return
	librdf_stream_update_current_element, always the current element or
	NULL at the end.

	* librdf/rdf_iterator.c (librdf_iterator_update_current_element):
	Do not do excessive work -
	only do a lower level check is_updated is false.
	(librdf_iterator_next): Reset is_updated after an
	iterator->next_method is performed.
	(librdf_iterator_get_object): Do less work, just return
	librdf_iterator_update_current_element, always the current element or
	NULL at the end.

	* librdf/rdf_stream.h, librdf/rdf_iterator.h: Add is_updated flag

	* librdf/rdf_node.c (librdf_node_equals): Actually use interned nodes.

	* librdf/rdf_uri.c (librdf_uri_equals): Actually use interned URIs.

	* librdf/rdf_model.c (Main):
	In test code, use librdf_parser_parse_string_as_stream rather
	than write files, delete them.

	* librdf/rdf_node.c:
	Updates to share all librdf_node objects like for librdf_uri - usage
	counts and shared pointers.
	(librdf_init_node): Initialise the three hashes for resource,
	literal, blank nodes.
	(librdf_finish_node): Clean up the three hashes.
	(librdf_new_node_from_uri_string_or_uri): Internal helper
	function to do all resource node creation.  Does the
	minimum of uri_string to/from uris objects conversion.
	If the resource node isn't in the resource hash (H_RESOURCE), add it
	otherwise just bump the usage count.
	(librdf_new_node_from_uri_string, librdf_new_node_from_uri,
	librdf_new_node_from_uri_local_name,
	librdf_new_node_from_normalised_uri_string): Use
	librdf_new_node_from_uri_string_or_uri.
	(librdf_new_node_from_typed_literal): If the librdf_node_encode-ed
	buffer of the literal node isn't in the literal hash (H_LITERAL),
	add it, otherwise just bump the usage count.
	(librdf_new_node_from_blank_identifier): If the identifier string
	isn't in the blank hash (H_BLANK), add it, otherwise just bump the
	usage count.
	(librdf_new_node_from_node): Now 1 line.
	(librdf_free_node): Decrement usage and then free and remove from the
	appropriate hashes if the usage was 0.
	(librdf_node_get_li_ordinal): Get the ordinal from the URI string
	since it isn't/wasn't ever stored.

	* librdf/rdf_node.h:
	librdf_node - add usage count. Remove ordinal union part, never used.
	Add key & size to literal part.

	* librdf/rdf_statement.c, librdf/rdf_query.c,
	librdf/rdf_concepts.c (main): Call librdf_init_node and
	librdf_finish_node in the test code.

	* librdf/rdf_init.h:
	Added node mutext (when threading) and node hashes to world structure.

	* librdf/rdf_init.c (librdf_world_open): Call librdf_init_uri.
	(librdf_world_set_uris_hash): Deleted
	(librdf_init_world): Don't use arg2.

	* librdf/rdf_uri.c (librdf_new_uri): Fail if uri_string is NULL.
	Tidy up cleanup and thread un-/locking.
	(librdf_new_uri_from_uri): Actually use the interning of URIs.
	(librdf_new_uri_from_uri_local_name): Fail if URI is NULL.
	Tidy up cleanup and thread un-/locking.
	(librdf_new_uri_normalised_to_base, librdf_new_uri_relative_to_base):
	Fail if URI is NULL.

	* raptor/configure.ac, configure.ac:
	Added --with-dmalloc option default auto for maintainer, no otherwise.

2003-08-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/lib/RDF/Redland/Parser.pm:
	Use raptor in the example. Add a mime type example.

	* demos/demo.pl: s/statement/triple/ in html

	* demos/demo.pl: Update parser labels.
	Add owl namespace.
	Trim rdf content whitespace.
	Become read only if db is not writable to this user.

	* docs/Makefile.am: Added .pod.html rule

2003-08-28  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* docs/fix-python-xhtml:
	Remove remaining broken fragments, example urls

	* docs/fix-python-xhtml: Exclude __builtin__ class from links

	* Released RPMs version 0.9.13-3 (patch redland-0.9.13-perl2.patch)

	* librdf/redland.spec.in:
	(0.9.13-2, 0.9.13-3) patches added to fix broken perl UNIVERSAL::isa

	* perl/lib/RDF/Redland/URI.pm, perl/lib/RDF/Redland/Node.pm:
	Use URI::URL in the isa

	* Released RPMs version 0.9.13-2 (patch redland-0.9.13-perl1.patch)

	* perl/lib/RDF/Redland/URI.pm, perl/lib/RDF/Redland/Node.pm:
	Fix totally broken use of UNIVERSAL::isa

	* perl/lib/RDF/Redland/URI.pm (new):
	Fix arg promotion from URI objects.

	* perl/lib/RDF/Redland/URI.pm (new): Fix cloning.

	* configure.ac: Bumped version to 0.9.14

	* Snapshotted redland_0_9_13 for 0.9.13 release

	* docs/redland.pod: Note out of dateness

	* docs/redland.pod, librdf/rdf_serializer_rdfxml.c,
	librdf/rdf_serializer_raptor.c, utils/db_upgrade.c,
	tcl/example.tcl, ruby/example.rb: Use librdf_model_as_stream

	* java/org/librdf/redland/Model.java (as_stream):
	Added - use Use librdf_model_as_stream or with a context
	node, use librdf_model_context_as_stream

	* python/RDF.py (Model.as_stream): Use librdf_model_context_as_stream

	* perl/lib/RDF/Redland/Model.pm:
	Fix uses of RDF::Redland::Node::_ensure($node) to use it as a redland
	node.
	(as_stream): Take optional context node, promote as necessary,
	using librdf_model_context_as_stream. Add documentation.

	* Redland.i: Added librdf_model_context_as_stream

	* librdf/rdf_model.c, librdf/rdf_model.h:
	Added librdf_model_context_as_stream, deprecating
	librdf_model_context_serialize as confusing with the
	librdf_serializer class and mixing US/UK spelling.

	* perl/serialize.pl, demos/demo.pl, perl/example.pl: Use as_stream

	* perl/lib/RDF/Redland/Model.pm (as_stream):
	Added, deprecating serialise and serialise.

	* librdf/rdf_model.c:
	Added librdf_model_as_stream, deprecating librdf_model_serialise
	as confusing with the librdf_serializer class.

	* librdf/rdf_model.h, Redland.i: Added librdf_model_as_stream

	* tcl/Makefile.am: TCL_INCLUDES

	* php/php.ini: load shared library from .

	* php/Makefile.am:
	Remove libtool again.  Try "if OSX, else rest of world" approach.

	* configure.ac:
	TCL_INCLUDE now TCL_INCLUDES and blank if not found, otherwise -Idir

	* tcl/Makefile.am:
	Remove libtool again.  Try "if OSX, else rest of world" approach.

	* configure.ac: default TCL_INCLUDE to /usr/include

	* ruby/Makefile.am: minor depends

	* ruby/Makefile.am:
	Remove libtool again.  Try "if OSX, else rest of world" approach.
	Package now 'redland' to match ruby convention.

	* ruby/example.rb: Now redland

	* tcl/Makefile.am: Don't -I-I@TCL_INCLUDE@

	* configure.ac:
	tcl version tweak -nonewline since OSX does silly things

	* tcl/Makefile.am: Removed README

	* tcl/README: Moved into ../docs/tcl.html

	* tcl/Makefile.am:
	Use @TCL_VERSION@ for the library name and -I@TCL_INCLUDE@

	* configure.ac:
	--with-perl/python/java/tcl/php/ruby now either enable the language
	or enable and set the binary.  Always try to find the version of
	the language before building the API automatically.
	Hunt for Tcl tcl.h and Java JDK JNI headers in various places.
	Report the languages available and language APIs built.

	* php/Makefile.am: comments.
	testing is optional, php program might not exist

	* php/Makefile.am:
	Very complex build and install using a libfake.la so I don't have to
	care about details of building shared object and dynamically loaded
	libraries.  No OSX specific fixes here since I can't see what
	the convention is.

	* tcl/Makefile.am: Be verbose in pkgIndex.tcl

	* ruby/Makefile.am: docs

	* tcl/Makefile.am: No -module here

2003-08-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* tcl/Makefile.am:
	Very complex build and install using a libfake.la so I don't have to
	care about details of building shared object and dynamically loaded
	libraries (such as on OSX).  The OSX specific naming is burnt in here
	since the dynamically loaded library must be called Redland.dylib

	* ruby/Makefile.am:
	Very complex build and install using a libfake.la so I don't have to
	care about details of building shared object and modules (such
	as on OSX).  OSX specific naming have to be burnt in here since
	shared modules must be called Redland.bundle

	* ruby/Makefile.am, tcl/Makefile.am:
	Make build and installing work again, without libtool.

	* tcl/README: Update Tcl building info

	* tcl/test.tcl: Back to auto_path .

	* configure.ac: BDB check update again.
	Report discovered BDB version in summary

	* configure.ac: Use bdbc_ for temporary variables in BDB checking code.

	* perl/Makefile.PL: Revert to working version.  No need for
	builddir/srcdir things.

	* perl/Makefile.am: Revert: Remove inplace editing of PERL_MAKEFILE

	* configure.ac: Another major BDB hunt and validate.  This time
	check for db.h with version numbers.

	* perl/Makefile.PL: Set FULLEXT to RDF/Redland

	* perl/Makefile.am: Make sure PERL really uses the right FULLEXT,
	BASEEXT and BOOTSTRAP values

	* librdf/rdf_hash_bdb.c:
	Remove BDB_CLOSE/FD_2_ARGS ifdef sequence at start - now tested in
	configure.
	Rename to HAVE_BDB_CLOSE/FD_2_ARGS

	* configure.ac: bdb_dbname test was inverted

	* configure.ac: use "if test ...; then", not "if [ ... ]; then"

	* configure.ac: I give up.  Put OSX fink hints here - look for /sw/bin.
	Add more Sleepycat/Berkeley DB checks, more guesswork and hunting for
	BDB headers in /usr/include/db* subdirectories.

	* perl/Makefile.am: Use CFLAGS, not MEM.

	* perl/Makefile.PL: Use CFLAGS, not MEM.
	Use redland-src-config --libs not --static-libs

	* librdf/rdf_hash_bdb.c (librdf_hash_bdb_open):
	More BDB #ifdef mess for db->open 6 or 7
	arguments.

	* librdf/rdf_utf8.c, librdf/rdf_uri.c, librdf/rdf_stream.c,
	librdf/rdf_storage.c, librdf/rdf_statement.c,
	librdf/rdf_serializer.c, librdf/rdf_query.c, librdf/rdf_parser.c,
	librdf/rdf_node.c, librdf/rdf_model.c, librdf/rdf_iterator.c,
	librdf/rdf_heuristics.c, librdf/rdf_hash.c, librdf/rdf_files.c,
	librdf/rdf_digest.c, librdf/rdf_concepts.c:
	When running a standalone test which has a STANDALONE #define set
	then do not compile the code from the module, but use the code in the
	linked librdf - all the tests link with this.  This prevents
	complaints about dual symbols from the OSX linker, and crashes with
	statics that store the pointers to factories (models, storages,
	parsers).

2003-08-26  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: Added AC_HEADER_TIME and checks for time.h, sys/time.h
	for gettimeofday

	* librdf/rdf_parser_raptor.c
	(librdf_parser_raptor_generate_id_handler): Added to use
	librdf_world_get_genid to get a unique genid if the user does not
	provide one.
	(librdf_parser_raptor_parse_file_as_stream,
	librdf_parser_raptor_parse_as_stream_common,
	librdf_parser_raptor_parse_into_model_common): Use
	librdf_parser_raptor_generate_id_handler to generate IDs.

	* librdf/rdf_node.c (librdf_new_node_from_blank_identifier):
	If a NULL identifier
	is passed in, use librdf_world_get_genid to get a unique genid.

	* librdf/rdf_init.c:
	Added a slew of time related headers for gettimeofday
	(librdf_new_world): Get the startup time to init world genid_base.
	Set genid_counter to 1.
	(librdf_world_set_feature): Support
	LIBRDF_WORLD_FEATURE_GENID_BASE, LIBRDF_WORLD_FEATURE_GENID_COUNTER
	both integers.
	(librdf_world_get_genid): Added to return a new
	genid of the form "r"+base+"r"+(counter++)

	* librdf/rdf_init.h: World gains genid_base and genid_counter.
	New world features LIBRDF_WORLD_FEATURE_GENID_BASE,
	LIBRDF_WORLD_FEATURE_GENID_COUNTER
	New internal librdf_world_get_genid

	* docs/fix-python-xhtml: +email

	* docs/fix-python-xhtml: -email

	* docs/fix-python-xhtml: words

	* docs/fix-python-xhtml: New pydoc, new bad use of ID.
	Skip the huge #commented heading in the source.

	* examples/example2.c:
	Updated to a slightly more useful example - parsing from
	string content and adding/checking/removing a statement.

	* python/redlandtest.py (testIterateModel, testContains): Added.

	* librdf/rdf_storage.c: Revert, remove contexts check here.

	* librdf/rdf_storage_list.c (librdf_storage_list_contains_statement):
	Handle contains with
	contexts using librdf_storage_list_find_statements.

	* librdf/rdf_storage_hashes.c
	(librdf_storage_hashes_contains_statement): Handle contains with
	contexts using librdf_storage_hashes_find_statements.

	* librdf/rdf_parser_raptor.c: removed unused variable

	* librdf/rdf_storage.c (librdf_storage_hashes_contains_statement):
	Revert fix.

	* librdf/rdf_storage_hashes.c
	(librdf_storage_hashes_contains_statement): Move the fix for
	stores with contexts into librdf_storage_contains_statement

	* librdf/rdf_storage_list.c
	(librdf_storage_list_contains_statement): Build a
	librdf_storage_list_node to use to run librdf_list_contains.

	* librdf/rdf_storage.h:
	Added index_contexts flag to librdf_storage_factory_s.

	* python/RDF.py (Model.__iter__): Fix to return an iter

	* python/setup.py:
	Remove MEM, MEM_LIBS code - rely on CFLAGS and LDFLAGS being passed in.
	Move version code to top.

	* python/Makefile.am: Pass on CFLAGS and LDFLAGS to python setup.py

	* python/example.py: empty statement

	* python/RDF.py: words in copyright

	* python/test/test.py, python/example.py:
	Updated to use 0.9 api features

	* ruby/Makefile.am: test-ruby fix shared dir

	* ruby/Makefile.am: LIBRUBYARG works 1.6-1.8

	* python/redlandtest.py:
	Update by Edd Dumbill to test the 0.9 RDF.py changes.

	* python/RDF.py:
	Major update to use Python 2.2+ idioms and features by Edd Dumbill.

	General remarks

	  * Removed string exceptions and replaced with RedlandError ones

	  * Where underlying Redland C objects were unexpectedly null, raise
	    exceptions rather than returning None, "" or [] as appropriate.

	  * Raise exceptions if Redland C constructors fail in Python
	    constructors.

	  * Did quite a lot of documentation updating, where possible
	    demonstrating preferred Pythonic usages.

	  * Added more unit tests to test changed functionality.

	RDF.Node

	  * Constructor now accepts Uri to make a resource/property node, or
	    a string to make a string literal

	  * Removed usage of node_type() in favour of is_resource(), is_blank()
	    etc methods.

	  * DEPRECATED get_blank_identifier() etc methods in favour of
	    node.blank_identifier etc properties.  Raise exception if properties
	    inconsistent with the node type are requested.

	RDF.Statement

	  * No need to name subject, predicate, object in constructor any more.
	    Uris or string literals accepted in place of nodes in constructor
	    argument.

	  * __getattr__ method deleted, using new-style property() for
	    subject, predicate, object

	RDF.Model

	  * Deleted get_*_iterator methods

	  * Renamed sources() to get_sources(), targets() to get_targets(),
	    arcs() to get_predicates().  Retained old method names as aliases.

	  * Added get_sources_context(), get_targets_context(),
	    get_predicates_context() which return (node, context) tuples.

	  * Amended get_source(), get_sources() etc methods so that Uri and
	    string literals were acceptable as shortcut types for nodes.

	  * Added find_statements_context() which returns (statement, context)
	    tuples.

	  * Renamed context_remove_statements() to
	    remove_statements_with_context(). Retained old method name as an
	    alias.

	  * Added __delitem__ method so del[statement] and del[statement,
	    context] work.

	  * Added append() method so append(statement) and append(statement,
	    context) work.  Suggest that append() is used in preference to
	    add_statement() for Python idiomatic usage.

	  * DEPRECATED serialise() in favour of as_stream().

	  * Added as_stream_context() to serialise (statement, context) tuples.

	  * Exception raised if len() is attempted on model with a non-countable
	    storage.

	  * DEPRECATED Model.add(), Model.add_typed_literal_statement() usages:
	    now that constructing Statements is easier, it's incongruous to have
	    these in the model.  Additionally it saves reproducing various bits
	    of error checking that properly belong in the Node() constructor.

	  * Implemented __contains__() so "if statement in model" and
	    "if (statement, context) in model" work as expected using the more
	    efficient Redland C methods where possible.

	RDF.Stream

	  * DEPRECATED Stream.context_iter().  Use in preference the _context
	    variants of the appropriate Model methods.  This leads to fewer
	    lines of code.

	RDF.Uri

	  * Amended constructor so strings or Uri instances can be passed as
	    the first argument, without named parameters being required, and
	    the right thing still happen.

	----
	I did some indenting and white space editing for consistency
	with the existing code.

	* librdf/rdf_list.c (librdf_list_iterator_get_method): Oops

	* librdf/rdf_hash.c (librdf_hash_get_all_iterator_get_method): oops

	* java/Makefile.am:
	Use check-local again, since we don't actually make programs called
	test1 and test2.

	* Redland.i: Add prototypes for static, internal methods

	* python/Makefile.am: Pass in CFLAGS to python

	* librdf/rdf_hash.c (librdf_hash_get_all_iterator_get_method): Give
	an error for an unknown method and return NULL.

	* librdf/rdf_list.c (librdf_list_iterator_get_method):
	Check method is get_object
	and return an error if it isn't.

	* librdf/rdf_storage.c
	(librdf_storage_stream_to_node_iterator_get_method): Provide
	iterator get_context method as well as get_object, using the
	internal stream.

	* ruby/Makefile.am: Made an rubylib make definition

	* tcl/Makefile.am: Made an tcllib make definition

	* tcl/Makefile.am:
	Interrogate tclsh for versions, and then guess include paths and
	library name.

	* tcl/test.tcl, tcl/example.tcl: Back to redland

	* tcl/test.tcl, tcl/example.tcl: Tcl package now Redland

	* configure.ac: Look for tclsh for TCL

	* ruby/Makefile.am: More libtool flags

	* ruby/Makefile.am: Build shared objects again.

	* ruby/Makefile.am: Don't depend on build-ruby

	* librdf/rdf_storage_hashes.c
	(librdf_storage_hashes_contains_statement): Use
	librdf_storage_hashes_find_statements to make this work when
	contexts are being used.

	* librdf/rdf_node.c (librdf_node_equals): Doc fix

	* librdf/rdf_storage_list.c (librdf_storage_list_node_equals):
	When not using contexts, don't
	check the context nodes.

	* Redland.i (librdf_call_python_message, librdf_call_perl_message):
	Fix use of
	vsnprintf in perl and python message callbacks to properly copy
	va_list using va_copy.  Same bug that happened in raptor.

2003-08-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* java/Makefile.am: Added test-java again

	* tcl/Makefile.am: Added test-tcl again

	* ruby/Makefile.am: Added test-ruby again

	* python/example.py: Modernise the python to 2.2+

	* python/RDF.py: Throughout:
	Updated to use new Python classes (2.2+)
	Replace __getattr__/__setattr__ with property() method.

	* python/Makefile.am: check-local now also does unittest-python

	* raptor/configure.ac: Bumped version to 0.9.13

	* raptor/raptor_parse.c (raptor_start_element_grammar):
	With rdf:datatype, do not lose the
	URI string pointer.  For rdf:ID, do not allocate the URI twice.

	* raptor/raptor_parse.c (raptor_generate_statement):
	Do not set language when a datatype is given.

	* raptor/raptor_xml_writer.c (raptor_new_xml_writer):
	Initialise writer buffer to an empty string
	to start (i.e. just \0).
	(raptor_xml_writer_start_element): Now assume buffer is always
	present, remove empty buffer case.
	(raptor_xml_writer_end_element,raptor_xml_writer_cdata): Handle 0
	length case, no strncpy.

	* raptor/tests/Makefile.am: Added ex-51.
	Fix daml+oil test.

	* raptor/tests/ex-51.out, raptor/tests/ex-51.rdf:
	Check empty XML literal works

	* raptor/tests/ex-41.out: No language for datatyped literals.

	* raptor/tests/daml-oil.rdf, raptor/tests/daml-oil.out,
	raptor/tests/Makefile.am: Updated to DAML+OIL schema 2001-03 as
	defined in http://www.daml.org/2001/03/daml+oil-index.html

	* raptor/tests/daml-oil.out, raptor/tests/daml-oil.rdf: Added
	2000-11-30 http://www.cs.man.ac.uk/%7Ehorrocks/DAML-OIL/daml-oil.rdf

	* raptor/tests/owl-schema.rdf, raptor/tests/owl-schema.out:
	Updated OWL schema http://www.w3.org/2002/07/owl to match
	that given in OWL Reference 2003-08-18 CR
	at http://www.w3.org/TR/2003/CR-owl-ref-20030818/#appB

	* raptor/libraptor.3: Updated for 0.9.12

	* python/RDF.py (Statement._get_subject):
	Missed accidently when updating to do more copies.

	* configure.ac: No languages by default.

2003-08-24  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.PL: Include -I builddir librdf dir to get rdf_config.h

	* Makefile.am: make languages after core libs

	* perl/Makefile.PL: More attempts at src/builddir fix.

	* perl/Makefile.am: Fix src/builddir for output of MakeMaker

	* tcl/Makefile.am: again

	* tcl/Makefile.am: oops, -c to compile

	* ruby/Makefile.am:
	This dir is now compiled and installed conditional on ruby being
	enabled with --with-ruby (off by default).
	Move -local targets into standard 'make' and 'make install' ones.
	Try to compile with the right Ruby shared library and install in
	the Ruby libs dir.

	* tcl/Makefile.am: Use @TCL@

	* tcl/Makefile.am:
	This dir is now compiled and installed conditional on tcl being
	enabled with --with-tcl (off by default).
	Move -local targets into standard 'make' and 'make install' ones.
	Try to compile the Tcl shared library and install in the Tcl
	libs dir.  Finding the tcl.h includes is impossible, so note
	that in the README.

	* tcl/README: Tcl building issues

	* php/Makefile.am: php installs now

	* configure.ac:
	In checking for progs, hunt for perl, python etc. rather than  names.
	Added --with-perl/python/java/php/tcl/ruby with optional
	argument =PATH to set the script.  This enables the particular
	language to be built.  Present defaults are perl and python are
	enabled.
	Move --with-ecma-cli to be nearby and accept 'no' also.
	Moved --with-jdk to be nearby.
	The subdirectories called with standard 'make' and 'make install'
	now are based on languages.
	Report the languages configured in the build summary.

	* python/Makefile.am:
	This dir is now compiled and installed conditional on python being
	enabled with --with-python (by default).
	Move -local targets into standard 'make' and 'make install' ones
	and create a fake build and install to keep automake happy, since
	we are using python's Distutils to do the real work.

	* perl/Makefile.am:
	This dir is now compiled and installed conditional on perl being
	enabled with --with-perl (by default).
	Move -local targets into standard 'make' and 'make install' ones
	and create a fake build and install to keep automake happy, since
	we are using perl's MakeMaker to do the real work.

	* java/Makefile.am:
	This dir is now compiled and installed conditional on java being
	enabled with --with-java, so move -local targets into standard
	'make' and 'make install' ones.
	Install both the shared library and jar.

	* java/Makefile.am: core_wrap.c without makefile variables

	* java/Makefile.am:
	EXTRA_LTLIBRARIES is broke - it won't build shared libraries.  So
	fake it using lib_LTLIBRARIES and then override the installation for
	install-javalibLTLIBRARIES: to do nothing.

	* java/Makefile.am:
	Shared library now librdf-java and built using libtool.

	* java/org/librdf/redland/World.java, java/redland-fragment.java:
	Now rdf-java (aka librdf-java under unixen)

	* librdf/rdf_init.c: Now 2003

2003-08-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* java/org/librdf/redland/Node.java:
	(constructors) Update to always copy rather than deal with shared
	nodes.
	(isResource, isLiteral, isBlank): Added.

	* java/org/librdf/redland/Statement.java:
	(constructors, getSubject, getPredicate, getObject) Update node and
	statements to always copy rather than deal with shared.

	* java/org/librdf/redland/Stream.java:
	Returning copied statements is default.

	* java/org/librdf/redland/Iterator.java:
	Returning copied nodes is default.

	* configure.ac: For --with-jdk, hunt for the machine-dependent subdir

	* java/Makefile.am: Use JDK_INCLUDES

	* java/org/librdf/redland/Parser.java:
	(parse) Added methods parsing content ina  string into a model
	or as a stream of statements.

	* java/org/librdf/redland/Model.java: sync has no return

	* configure.ac: Added --with-jdk to specify where Java lives

	* java/org/librdf/redland/Model.java,
	perl/lib/RDF/Redland/Model.pm (sync): Added calling
	librdf_model_sync

	* Redland.i: Added librdf_model_sync

	* librdf/rdf_model_storage.c (librdf_model_storage_sync):
	Added, invoking librdf_storage_sync

	* librdf/rdf_model.c (librdf_model_sync):
	Added, invoking it in the implementation via new
	factory method sync.

	* librdf/rdf_model.h: Added librdf_model_sync and factory method.

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_sync):
	Added, implementing sync method.

	* librdf/rdf_storage.c: Added librdf_storage_sync

	* librdf/rdf_storage.h: Added librdf_storage_sync and factory method.

	* python/Makefile.am, perl/Makefile.am: clean more

	* perl/serialize.pl: Updated API tweaks

	* perl/lib/RDF/Redland/Node.pm (new_from_literal):
	Ensure is_wf_xml is an integer.

	* perl/lib/RDF/Redland/Model.pm (remote_statement):
	Allow NODE NODE NODE [CONTEXT] args

	* perl/lib/RDF/Redland/Statement.pm (new):
	Promote node arguments to redland nodes using
	RDF::Redland::Node::_ensure.

	* perl/lib/RDF/Redland/Node.pm (_ensure):
	Added, to promote other objects to redland nodes

	* perl/lib/RDF/Redland/Model.pm (add): Now call add_statements.
	(add_statements,remove_statement,add_statements,remove_statements,
	remove_context_statements): Promote node arguments to redland nodes
	using RDF::Redland::Node::_ensure.

	* perl/lib/RDF/Redland/Statement.pm (new_from_nodes):
	Fix call to constructor

	* perl/example.pl:
	Test add_statement(NODE, NODE, NODE) form - with same NODE

	* perl/lib/RDF/Redland/Model.pm (add_statement):
	Accept NODE NODE NODE [CONTEXT] as arguments too.

2003-08-22  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Redland.i: typo

	* perl/lib/RDF/Redland/Parser.pm:
	(parse_string_as_stream, parse_string_into_model):
	Added for parsing strings.  Note that the base_uri is required.

	* python/RDF.py:
	(Parser.parse_string_as_stream, Parser.parse_string_into_model):
	Note base_uri is required in docs, raise an error if it isn't there.

	* python/RDF.py:
	(Parser.parse_string_as_stream, Parser.parse_string_into_model):
	Added.

	* librdf/rdf_parser_raptor.c
	(librdf_parser_raptor_parse_as_stream_common,
	(librdf_parser_raptor_parse_uri_into_model_common): Ensure
	source_uri is set up correctly.  Refuse to work without a
	base_uri.

	* python/test/test.py: Added test with parser.parse_string_as_stream

	* librdf/redland.spec.in: try to get perl install working again

	* Redland.i: Added librdf_parser_parse_string_as_stream,
	librdf_parser_parse_string_into_model 

	* librdf/rdf_parser_raptor.c: Added new methods for implementing
	librdf_parser_parse_string_as_stream,
	librdf_parser_parse_string_into_model
	(librdf_parser_raptor_parse_as_stream_common): New, core
	work for both string and uri parsing into a stream.
	(librdf_parser_raptor_parse_uri_into_model_common): New, core
	work for both string and uri parsing into a model.

	* librdf/rdf_parser.h (librdf_parser_parse_string_as_stream,
	librdf_parser_parse_string_into_model): Added

	* librdf/rdf_parser.c (librdf_parser_parse_string_as_stream,
	librdf_parser_parse_string_into_model): Added

2003-08-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/Makefile.am: Remove -static from test links

	* raptor/raptor_uri.c (main):
	Test xmlbase and retrievable URI tranforms.

	* raptor/raptor_uri.c (raptor_uri_resolve_uri_reference):
	Handle #s relative to a uri-reference with a #fragment.

	* raptor/raptor_www.c:
	(raptor_www_fetch) Use raptor_new_uri_for_retrieval to ensure
	that the URI-reference fragments are removed, and the URI path exists.

	* raptor/raptor.h: Added raptor_new_uri_for_retrieval

	* raptor/raptor_uri.c (raptor_new_uri_for_retrieval):
	Added, strips fragments and ensures / path is present.

	* raptor/raptor_xml_writer.c (raptor_xml_writer_start_element,
	raptor_xml_writer_end_element): Use size_t for lengths.

	* raptor/raptor_xml_writer.c (raptor_xml_writer_start_element):
	Set content_element_seen in parent
	only if there is a parent.
	(raptor_xml_writer_end_element): Change current_element to parent
	only if there is a current element.

	* raptor/tests/Makefile.am:
	Pull out may-fail NFC checks into a separate set and don't exit 1
	if they do fail.  Failure is possible since it requires GNOME glib2
	which isn't always available.

	* librdf/redland.spec.in: Fix perl CORE dir

	* librdf/redland.spec.in: Removed duplicate perl CORE shared objects

	* librdf/redland.spec.in: Added redland-db-upgrade.1 to devel

2003-08-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.PL:
	Use .. for redland-src-config and add check it is there.

	* python/setup.py:
	Use .. for redland-src-config and add an os.access check for it

	* redland-src-config.in: test ! -r

	* redland-src-config.in: Stop if cannot find raptor-src-config

	* python/setup.py: append define_macros only on match.

	* python/setup.py: Oh yes it is.

	* python/setup.py: redland-src-config is not in top_srcdir

	* python/redlandtest.py, python/setup.py, python/example.py:
	Set #! to /usr/bin/python

	* python/RDF.py:
	Update classes to always return copies of Nodes, Statements
	from streams, iterators such as model.find_statements.
	The user should now not need care about shared versus copying
	of returned values but just use them.

	* python/setup.py, python/Makefile.am:
	Pass MEM, MEM_LIBS to python compiling C

	* perl/lib/RDF/Redland/Statement.pm:
	(subject,predicate,object) Copy redland node for librdf set methods.

	* perl/lib/RDF/Redland/Node.pm: POD fixes

	* perl/example.pl: Minor updates to use newer API calls.

	* perl/lib/RDF/Redland/Statement.pm:
	(subject,predicate,object) Docs - note no copying neeeded.
	(predicate): Minor typo

	* perl/lib/RDF/Redland/RSS.pm (Model.DESTROY): Call super to tidy up
	(Node.new): Debugging
	(Node.DESTROY): Call super to tidy up
	(properties): Do less work, no need to copy.

	* perl/lib/RDF/Redland/Node.pm (new): Invoke clone properly.

	* perl/lib/RDF/Redland/URI.pm (new):
	Constructor now copies as well as building URIs.  Takes
	string literals or URIs by sniffing for classes.
	(new_from_uri): Replaced by new
	(clone): Another way to copy a URI.
	Document these changes

	* perl/lib/RDF/Redland/Iterator.pm:
	Update for Node API change and no need to copy here
	(current, context): Do not copy returned node.

	* perl/lib/RDF/Redland/Stream.pm:
	Update for Statement API change and no need to copy here
	(current, context): Do not copy returned statement.

	* perl/lib/RDF/Redland/Model.pm:
	Update for Node, Statement API changes and no need to copy here
	(sources,targets,arcs): Don't copy statements to new lists.

	* perl/lib/RDF/Redland/Statement.pm (new):
	Constructor now copies as well as building statements.  Takes
	string literals, URIs by sniffing for classes.
	(new_from_statement,new_from_nodes): Replaced by new.
	(clone): Another way to do new_from_statement
	(subject,predicate,object): Do less work, always return a copy.
	Document these changes

	* perl/lib/RDF/Redland/Node.pm (new):
	Constructor now builds resources, literals, URIs by sniffing
	for classes.
	(new_from_literal,new_from_typed_literal): Replaced with new_literal
	(new_xml_litearl): Added
	(new_from_node): Replaced with clone
	Document these changes

	* perl/rss-dump.pl:
	Comment out (working) code that lets perl do the URI retrieval,
	assume raptor has that.
	Slight updates for changed perl API.

	* librdf/rdf_node.c (main):
	Test: Revert accidently removed librdf_free_uri

	* perl/Makefile.am, perl/Makefile.PL:
	Pass on MEM to CFLAGS when building perl.

	* librdf/rdf_parser_raptor.c:
	Use LIBRDF_FREE for data returned from raptor.  This is OK
	as long as raptor is in the redland sources.  Otherwise it'll
	need to be SYSTEM_FREE.

	* librdf/rdf_hash_bdb.c:
	Use SYSTEM_FREE for data allocated by BDB via malloc()

	* librdf/rdf_init.c: Use SYSTEM_MALLOC/FREE for mutexes.
	(librdf_system_malloc, librdf_system_free): Added for dmalloc
	debugging.

	* librdf/librdf.h:
	Added SYSTEM_MALLOC and SYSTEM_FREE for routines that return
	genuine malloc() and free()ed memory - outside Redland typically.

	* raptor/raptor_general.c (raptor_check_ordinal):
	parentheses just for gcc

2003-08-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* python/setup.py: license now license apparently

	* utils/redland-db-upgrade.1: fix db names

	* utils/Makefile.am: Added redland-db-upgrade.1

	* utils/redland-db-upgrade.1: Manpage for redland-db-upgrade

	* librdf/rdf_node.c:
	Make nodes immutable - remove all node class set methods.
	(librdf_new_node_from_literal): Now just calls
	librdf_new_node_from_typed_literal.
	(librdf_new_node_from_typed_literal):
	librdf_new_node_set_typed_literal_value inlined here.
	(librdf_new_node_from_blank_identifier):
	librdf_set_blank_identifier inlined here.
	(librdf_new_node_from_node): Call the appropriate node constructors
	rather than malloc a new node and use the set methods.
	(librdf_node_set_uri, librdf_node_set_type,
	librdf_node_set_typed_literal_value, 
	librdf_node_set_blank_identifier): Deleted and inlined where used.
	(librdf_node_set_li_ordinal,librdf_node_set_literal_value): Deleted,
	never used in this module.
	(librdf_node_decode): Call the appropriate node constructors
	rather than malloc a new node and use the set methods.
	(main): Remove node set methods from test code.

	* librdf/rdf_node.h:
	Make nodes immutable - removed all public set methods into rdf_node.c

	* java/org/librdf/redland/Node.java:
	Make nodes immutable - remove set methods setURI, setType,
	setLiteralValue.

	* perl/lib/RDF/Redland/Node.pm:
	Make nodes immutable.  Remove set parts of methods uri,
	blank_identifier and type.  Remove method set_literal_value.

	* python/RDF.py (Class Node):
	Make nodes immutable - remove set methods.

	* Redland.i: Make nodes immutable - remove node class set methods.

	* Redland.i: comment out calling python exceptions for now

2003-08-17  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: Remove MEM=-DLIBRDF_MEMORY_DEBUG if no dmalloc present.

	* librdf/rdf_uri.c, librdf/rdf_stream.c, librdf/rdf_storage.c,
	librdf/rdf_statement.c, librdf/rdf_serializer.c,
	librdf/rdf_query.c, librdf/rdf_parser.c, librdf/rdf_node.c,
	librdf/rdf_model.c, librdf/rdf_iterator.c,
	librdf/rdf_heuristics.c, librdf/rdf_hash.c, librdf/rdf_files.c,
	librdf/rdf_digest.c, librdf/rdf_concepts.c (main): In test code,
	remove calls to librdf_memory_report for old internal memory
	debugging code.

	* librdf/librdf.h:
	Remove LIBRDF_MEMORY_DEBUG and internal memory loss debugging code.
	External tools such as dmalloc work a lot better.

	* librdf/Makefile.am: Removed rdf_memory.c

	* librdf/rdf_memory.c: Memory debugging code superceeded by better
	external tools such as dmalloc

	* librdf/rdf_hash_bdb.c, configure.ac:
	Added BDB 4.0 and 4.1 open interface support - 6 or 7 arguments.

	* raptor/raptor_xml_writer.c:
	(raptor_xml_writer_end_element) Reset the current_element pointer
	on finishing.  Makes any succeeding cdata do the right thing.

	* librdf/redland.spec.in: Updates for new perl module names

	* raptor/rdfdump.c: Inside redland, don't call raptor_init/finish,
	it's done by redland's world.

2003-08-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_rss.c (raptor_rss_parse_chunk):
	Stop working aftera a user abort of the parser.

	* raptor/raptor_general.c (raptor_check_ordinal): c is not const

	* raptor/rdfdump.c, raptor/raptor_xml_writer.c,
	raptor/raptor_xml.c, raptor/raptor_www_libwww.c,
	raptor/raptor_utf8.c, raptor/raptor_uri.c, raptor/raptor_set.c,
	raptor/raptor_sax2.c, raptor/raptor_qname.c,
	raptor/raptor_parse.c, raptor/raptor_namespace.c,
	raptor/raptor_locator.c, raptor/raptor_libxml.c,
	raptor/raptor_identifier.c, raptor/raptor_general.c,
	raptor/ntriples_parse.c: Move dmalloc includes into
	raptor_internal.h and use everywhere.

	* raptor/raptor_internal.h:
	Add raptor dmalloc includes here to ensure all raptor code uses it
	or not consistently.

	* librdf/rdf_parser_raptor.c: (librdf_parser_raptor_error_handler,
	librdf_parser_raptor_warning_handler):
	Remove inadvisable use of raptor_parser as a variable name.
	(librdf_parser_raptor_parse_uri_into_model): Zap unused world.

	* librdf/rdf_hash.c (librdf_hash_from_array_of_strings):
	Fix return value.

	* examples/Makefile.am: Added redland_dbus

	* examples/redland_dbus.c: Initial import

2003-08-12  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_uri.c, librdf/rdf_stream.c,
	librdf/rdf_storage_list.c, librdf/rdf_storage_hashes.c,
	librdf/rdf_storage.c, librdf/rdf_statement.c,
	librdf/rdf_serializer_rdfxml.c, librdf/rdf_serializer.c,
	librdf/rdf_query.c, librdf/rdf_parser_raptor.c,
	librdf/rdf_parser.c, librdf/rdf_node.c, librdf/rdf_model.c,
	librdf/rdf_init.c, librdf/rdf_hash_memory.c,
	librdf/rdf_hash_bdb.c, librdf/rdf_hash.c, librdf/rdf_digest.c,
	librdf/rdf_concepts.c:
	Replaced most LIBRDF_FATAL1/2 calls with LIBRDF_ERROR1/2; both now
	take world arguments.  The remaining fatal errors are either in
	redland startup when there is no sensible way to recover or in out of
	memory situations.  The LIBRDF_ERROR1/2 calls now give the error and
	recover/fail the method if appropriate.  The documentation on several
	methods were updated to note this.  The error notification isn't
	always possible since some methods have way to return errors.
	A new method on the world class is needed to test for this state.

	* librdf/librdf.h:
	Added LIBRDF_ERROR1 and LIBRDF_ERROR2 that take world argument and call
	either librdf_error(world, ...) or #ifdef LIBRDF_DEBUG,  abort()

	* librdf/rdf_parser_repat.c: Out of date, gone

	* librdf/rdf_stream.c, librdf/rdf_storage_list.c,
	librdf/rdf_storage.c, librdf/rdf_serializer_rdfxml.c,
	librdf/rdf_parser_repat.c, librdf/rdf_parser_raptor.c,
	librdf/rdf_node.c, librdf/rdf_hash_memory.c,
	librdf/rdf_hash_bdb.c, librdf/rdf_digest.c: Replace most abort()
	with calls to LIBRDF_FATAL2 which aborts but gives a message too.

	* librdf/rdf_init.c (librdf_world_get_feature,
	librdf_world_set_feature): Return failure
	rather than crash - not implemented or exposed in APIs yet.

2003-08-11  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_model.c (main):
	Add test that librdf_model_add_statement fails on a partial statement.

	* librdf/rdf_model.c (librdf_model_size):
	Added warning that it might not return an answer.
	(librdf_model_add_statement, librdf_model_add,
	librdf_model_remove_statement, librdf_model_contains_statement,
	librdf_model_context_add_statement,
	librdf_model_context_remove_statement): Added check that the
	statement used is complete, using new librdf_statement_is_complete

	* librdf/rdf_statement.h: Added librdf_statement_is_complete

	* librdf/rdf_statement.c (librdf_statement_is_complete):
	Added to check that a statement has
	all the subject, predicate, object fields.

2003-08-08  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am: Added warn-03

	* raptor/tests/warn-03.rdf: Handle deleting of default namespaces

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Handle when a name has a
	namespace but that namespace has no URI such as xmlns="".  In
	that case, the element has non-namespaced parts too, so skip.

	* raptor/ntriples_parse.c (raptor_ntriples_parse_line):
	Casts so isspace calls get int args.

	* raptor/raptor_uri.c (raptor_uri_is_absolute):
	Cast so isalpha and isalnum get int args.

	* raptor/tests/Makefile.am: Addex ex-50

	* raptor/tests/ex-50.out, raptor/tests/ex-50.rdf:
	Check parseType with unknown value

	* raptor/raptor_parse.c (raptor_start_element_grammar):
	Handle parseType="Literal"
	without duplicating code.

	* raptor/raptor_parse.c (raptor_start_element_grammar):
	Handle parseType="...." which
	isn't any of the other known types identically to
	parseType="Literal".

	* raptor/raptor_general.c (raptor_check_ordinal):
	Return <0 on failure such as no legal
	characters at all.

	* raptor/raptor_internal.h: Added raptor_check_ordinal.

	* raptor/ntriples_parse.c (raptor_ntriples_generate_statement): Make
	RAPTOR_IDENTIFIER_TYPE_ORDINAL predicates for property URI strings
	that match the rdf:_<n> pattern with n a decimal integer>0.

	* raptor/raptor_parse.c:
	Use raptor_check_ordinal for checking <n> in rdf:_<n>

	* raptor/raptor_general.c (raptor_check_ordinal):
	Check the <n> in rdf:_<n>

	* raptor/raptor_general.c (raptor_vsnprintf):
	Non-portable use of va_list fixed by copying the
	arguments with va_copy before passing to vsnprintf calls.  The
	symptom was crashes on some architectures where this mattered, such
	as powerpc.

2003-08-07  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_parser_raptor.c: Track errors and warnings in parsing.
	(librdf_parser_raptor_error_handler(): Call raptor_parse_abort

	* librdf/rdf_statement.c (librdf_statement_decode_parts):
	Updated for new librdf_node_decode API

	* librdf/rdf_node.c (librdf_node_decode):
	Updated for new API - returns a new allocated node
	and sets the size pointer as a side effect.  This removes the only
	use of librdf_new_node outside the node class.
	(main): Updated the test code to match.

	* librdf/rdf_node.h (librdf_node_decode): Updated for new API

	* raptor/raptor_general.c (raptor_parse_uri_with_connection):
	Return failure status.

	* librdf/rdf_node.c (librdf_new_node): No args means make a blank node.
	(librdf_new_node_from_blank_identifier): Docs

	* python/RDF.py (Model.serialise):
	Add optional contexts arg, to serialise contexts
	as stream of statements.
	(NS): New class to cache handy nodes from a namespace URI

2003-08-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am:
	Added bad N-Triples tests bad-0[0-3].nt and checks

	* raptor/tests/bad-00.nt, raptor/tests/bad-01.nt,
	raptor/tests/bad-02.nt, raptor/tests/bad-03.nt: Bad N-Triples

	* raptor/ntriples_parse.c: raptor_ntriples_term_class Added for:
	(raptor_ntriples_term_valid): Checking validity of a ntriples term -
	this could be inlined.
	(raptor_ntriples_string) Renamed to:
	(raptor_ntriples_term) Use raptor_ntriples_term_valid.
	(raptor_ntriples_parse_line): Add more checks that whitespec exists
	between ntriples terms.
	Error to have typed literals with languages.

	* perl/Makefile.PL: Remove -Dbool=int - this can only cause trouble

	* perl/lib/RDF/Redland/Node.pm:
	Added methods is_resource, is_literal, is_blank using
	librdf_node_is_resource, librdf_node_is_literal, librdf_node_is_blank

	* python/RDF.py: Removed NODE_TYPE_LI
	(Node) is_resource, is_literal, is_blank now use
	librdf_node_is_resource, librdf_node_is_literal, librdf_node_is_blank

	* Redland.i, librdf/rdf_node.h, librdf/rdf_node.c:
	Added librdf_node_is_resource, librdf_node_is_literal,
	librdf_node_is_blank

	* librdf/rdf_node.h, librdf/rdf_node.c:
	Delete LIBRDF_NODE_TYPE_LI - never used.  Reserve the allocated type 3.

	* raptor/tests/test.out, raptor/tests/test.nt:
	Updated to remove language from typed literals

	* raptor/raptor_general.c (raptor_print_statement_detailed):
	Fix datatype uri output <uri> not <uri<

2003-07-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_general.c (raptor_parser_simple_error):
	Call raptor_parser_error_varargs, don't
	lose the arguments.

2003-07-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: ECMA CLR=>ECMI CLI since it's not just the runtime.
	ECMA CLI= ECMA  Common  Language  Infrastructure.

	* configure.ac:
	No need for CSHARP_COMPILE, dotGNU compiler accepts same args as mcs.

2003-07-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.ac: Also substitute CSHARP_COMPILE for csharp

	* configure.ac:
	Added --with-ecma-clr for mono or dotgnu versions of toolchain.
	Sets CLRRUN and CSHARPCC for use in csharp dir

	* Makefile.am: Added csharp dir

	* raptor/configure.ac: Bumped version to 0.9.12

	* raptor/raptor_sax2.c: struct nsd: use size_t for length.

	* raptor/configure.ac:
	Ensure the libxml2 xmlReader API is new enough (2.5.0+) such as
	having xmlParserSeverities.

2003-07-28  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h:
	Update raptor_generate_id_handler to take user_bnodeid arg.

	* raptor/raptor_rss.c: Update calls of raptor_generate_id with
	user_bnodeid (NULL for existing calls)

	* raptor/raptor_parse.c:
	Update calls of raptor_generate_id with user_bnodeid (NULL for
	existing calls)
	Use it to wrap the rdf:nodeID values for subject and object cases.

	* raptor/raptor_internal.h:
	raptor_generate_id updated to add user_bnodeid

	* raptor/raptor_general.c (raptor_set_generate_id_handler):
	Document final argument
	user_bnodeid from the rdf:nodeID attribute value.
	(raptor_default_generate_id_handler): Add user_bnodeid, return it if
	present.
	(raptor_generate_id): Add user_bnodeid and pass on.

	* raptor/raptor_internal.h: Added raptor_xml_writer_comment

	* raptor/raptor_namespace.c (raptor_namespace_copy):
	Don't copy uri and then lose it.

	* raptor/raptor_xml_writer.c:
	Added current_element for tracking empty/not empty elements.
	(raptor_xml_writer_comment): Added, just concatenating the content
	via raptor_xml_writer_cdata.

	* raptor/raptor_parse.c (raptor_xml_comment_handler):
	Call raptor_xml_writer_comment inside
	parseType="Literal"

	* raptor/raptor_xml_writer.c: raptor_xml_writer gains stack depth.
	(raptor_free_xml_writer): Clear any content_cdata before finishing.
	(raptor_xml_writer_start_element)
	Add depth to raptor_format_sax2_element calls.  Increase it
	(raptor_xml_writer_start_element,raptor_xml_writer_end_element):
	Add depth to raptor_format_sax2_element calls.  Decrease it
	and raptor_namespaces_end_for_depth each time.

	* raptor/raptor_namespace.c:
	Moved  error_handler and error_data arguments around.
	(raptor_namespaces_start_namespace) Gets those as arguments
	(raptor_namespaces_start_namespace) Added, simpler version of _full
	(raptor_namespaces_start_namespace_full) Added, was the old interface
	but less error arguments.
	(raptor_namespaces_namespace_in_scope): Fix namespace URI comparison.
	(raptor_namespace_new) Looses error arguments.
	(raptor_namespace_copy) Added, copy to a new stack with a new depth.

	* raptor/raptor_sax2.c (raptor_format_sax2_element):
	Gain stack depth argument.
	Only use namespace declarations when there is a namespace stack
	present.
	copy namespaces to new stack when new ones are needed.

	* raptor/raptor_internal.h:
	Add error_handler and error_data to namespace_stack.
	raptor_namespaces_start_namespace gets those as arguments
	raptor_namespace_new looses them
	raptor_namespaces_start_namespace takes less args
	raptor_namespaces_start_namespace_full added
	raptor_namespace_copy added
	content_cdata_seen and content_element_seen back into sax2_element

	* raptor/raptor_parse.c:
	Moved content_cdata_seen and content_element_seen back into
	sax2_element.
	Update for new raptor_namespaces_start_namespace calling convention.

	* raptor/raptor_xml_writer.c: Debug

	* raptor/raptor_namespace.c (raptor_namespaces_format):
	Fix missing counting : when present

	* raptor/tests/ex-11.rdf, raptor/tests/ex-11.out:
	Updated to declare the html namespace as default, expect it in the
	N-Triples output.

	* raptor/raptor_namespace.c (raptor_namespaces_namespace_in_scope):
	Added, checking if a given
	namespace is declared in scope.
	(raptor_namespaces_format): Added, returning a string to declare
	the given namespace.

	* raptor/raptor_xml_writer.c:
	(raptor_xml_writer_start_element,raptor_xml_writer_end_element):
	Updated for raptor_format_sax2_element new arguments

	* raptor/raptor_sax2.c (raptor_format_sax2_element):
	Add raptor_namespace_stack argument.
	Create xmlns declarations for elements not declared in the current
	stack state, using raptor_namespaces_format to create the string.

	* raptor/raptor_internal.h:
	Added prototypes for raptor_namespaces_namespace_in_scope,
	raptor_namespaces_format
	raptor_format_sax2_element now takes a raptor_namespace_stack

2003-07-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_internal.h:
	raptor_xml_writer prototypes take unsigned char*

	* raptor/raptor_parse.c:
	Move the code building parseType="Literal" strings to
	raptor_xml_writer class.
	(raptor_cdata_grammar): Added, just for symmetry mostly, with most
	code taken from raptor_xml_cdata_handler.

	* raptor/raptor_xml_writer.c:
	Move the code building parseType="Literal" strings to
	raptor_xml_writer class.

	* raptor/Makefile.am: Added raptor_xml_writer.c

	* raptor/raptor_xml_writer.c: Initial version

	* raptor/raptor_parse.c:
	Split content_cdata fields between sax2_element & (RDF/XML
	specific) element.
	raptor_element: Add xml_writer field.
	Various calls changed to use the new raptor_simple_message_handler
	for error handling implemented as raptor_parser_simple_error here.
	(raptor_xml_parser_simple_error_handler): Added, matching the
	raptor_simple_message_handler API and calling raptor_parser_error.
	(raptor_start_element_grammar): When parseType="Literal" appears,
	create a new raptor_xml_writer.
	(raptor_end_element_grammar): When parseType="Literal" ends, delete
	the raptor_xml_writer.

	* raptor/raptor_sax2.c (raptor_format_sax2_element):
	Use raptor_simple_message_handler.

	* raptor/raptor_general.c (raptor_start_parse):
	Use raptor_parser_simple_error with raptor_namespaces_init.
	(raptor_parser_simple_error): Added, matching the
	raptor_simple_message_handler API but same as raptor_parser_error.

	* raptor/raptor_xml.c (raptor_xml_escape_string):
	Use raptor_simple_message_handler.

	* raptor/raptor_qname.c (raptor_new_qname):
	Use raptor_simple_message_handler.

	* raptor/raptor.h: raptor_sax2_element moved here, semi-public.
	Re-ordered URI functions earlier.
	Various methods changed to use (public) raptor_simple_message_handler
	for error handling.

	* raptor/raptor_internal.h: Delete raptor_internal_message_handler.
	Added prototype raptor_parser_simple_error, implementing
	raptor_simple_error_handler API.
	Various methods changed to use (public) raptor_simple_message_handler
	for error handling.
	raptor_sax2_element moved to semi-public raptor.h
	rdf/xml-specific cdata parts moved from raptor_sax2_element to
	raptor_element.
	Added raptor_xml_writer functions.

	* raptor/raptor_namespace.c:
	Use (public) raptor_simple_message_handler for error handling.

2003-07-24  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_parse.c (raptor_process_property_attributes):
	NFC error message tidy.
	(raptor_end_element_grammar): Fix NFC error reporting and recovery.
	Report NFC validation failures for XML Literals

	* raptor/tests/Makefile.am: Added bad-15 bad-17 for bad NFC checking

	* raptor/tests/bad-16.rdf, raptor/tests/bad-17.rdf,
	raptor/tests/bad-15.rdf: Bad NFC tests for property attribute,
	element, element ptl

	* raptor/configure.ac:
	Added check for g_utf8_normalize in glib 2.0 using pkgconfig.
	Defines HAVE_G_UTF8_NORMALIZE if present.

	* raptor/raptor_internal.h: Added raptor_utf8_is_nfc

	* raptor/raptor_parse.c (raptor_process_property_attributes):
	Check for valid NFC on property
	attribute values.
	(raptor_end_element_grammar): Check for valid NFC on plain literal
	property element values.

	* raptor/raptor_utf8.c (raptor_utf8_is_nfc):
	Added Normal Form C checking, using GNOME glib
	2.0 g_utf8_normalize initially.

	* raptor/raptor_parse.c (raptor_start_element_grammar):
	Forbid property attributes and all
	rdf:* attributes (except rdf:ID) with rdf:parseType

	* raptor/raptor_general.c:
	(raptor_statement_part_as_counted_string,
	raptor_print_statement_part_as_ntriples): Do not emit language for
	datatyped literals.

	* raptor/raptor_parse.c (raptor_process_property_attributes):
	rdf:li is forbidden as a property attribute

	* raptor/raptor_parse.c: rdf_syntax_terms_info table: rdf:li is
	forbidden as a property attribute

	* raptor/libraptor.3: Updated raptor_set_feature for
	RAPTOR_FEATURE_ALLOW_BAGID and
	RAPTOR_FEATURE_ALLOW_RDF_TYPE_RDF_LIST

	* raptor/tests/owl-schema.out, raptor/tests/ex-39.out:
	Updated to remove the <idList> rdf:type rdf:List triples

	* raptor/raptor_general.c (raptor_set_feature,
	raptor_set_parser_strict): Added a new feature
	RAPTOR_FEATURE_ALLOW_RDF_TYPE_RDF_LIST (user argument) and
	feature_allow_rdf_type_rdf_List (internal) to generate the rdf:type
	rdf:List triple from rdf:parseType="Collection".  The default is no
	after latest RDF/XML revisions.  Not relevant for daml:Collection
	which get the daml:List always.

	* raptor/raptor.h:
	Added a new feature feature RAPTOR_FEATURE_ALLOW_RDF_TYPE_RDF_LIST
	to control rdf:type rdf:List triple generation from
	rdf:parseType="Collection" (default no)

	* raptor/raptor_internal.h:
	Added a new feature feature_allow_rdf_type_rdf_List
	to control rdf:type rdf:List triple generation from
	rdf:parseType="Collection" (default no)

	* raptor/raptor_parse.c (raptor_start_element_grammar):
	Remove rdf:type rdf:List triple
	generation from rdf:parseType="Collection" by default.  Not
	for daml:Collection.  Add a new feature
	feature_allow_rdf_type_rdf_List to control this.

2003-07-22  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_xml.c (raptor_valid_xml_ID, raptor_xml_escape_string):
	unsigned long for
	all unichars.

	* raptor/raptor_internal.h:
	Update raptor_utf8_to_unicode_char to use unsigned long output.

	* raptor/raptor_utf8.c (raptor_utf8_to_unicode_char):
	Take and use unsigned long for unichars.

	* raptor/raptor_rss.c:
	namespace->nspace since might be a C/C++ keyword sometime

	* raptor/raptor_www_curl.c (raptor_www_curl_header_callback):
	Turn void* into char*

	* raptor/raptor_set.c: Casts

	* raptor/rdfdump.c (rdfdump_error_handler):
	Cast data into raptor_parser*

	* raptor/raptor_www_curl.c:
	(raptor_www_curl_write_callback,raptor_www_curl_header_callback):
	Return unsigned int 0 on failure, cannot return -1 :)

	* raptor/raptor_www.c: Some casts near mallocs

	* raptor/tests/owl-schema.out: Updated to match 2003-03-18 version.

	* raptor/tests/owl-schema.rdf:
	Updated to 2003-03-18 version (just changed DOS line endings)

	* raptor/ntriples_parse.c (raptor_ntriples_parse_chunk):
	Handle just the end marker being given
	i.e. len=0 (and possibly s=NULL)

	* raptor/raptor_parse.c (raptor_xml_end_element_handler):
	When parsing has been aborted
	(rdf_parser->failed), clean up used memory rather than just return.
	element_name is not used except when debugging, so #ifdef it.

2003-07-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h:
	Export global statics raptor_copyright_string, raptor_version_string,
	raptor_version_major, raptor_version_minor, raptor_version_release
	and raptor_version_decimal

	* raptor/raptor_general.c:
	Added statics raptor_copyright_string, raptor_version_string,
	raptor_version_major, raptor_version_minor, raptor_version_release
	and raptor_version_decimal

	* raptor/Makefile.am:
	Removed raptor_cc code since ISO may charge a commercial use fee for
	this list.

	* raptor/raptor_cc.gperf: ISO 3166-1 'The use of ISO 3166-1 in
	commercial products may be subject to a licence fee.' says the
	maintenance agency.  Goodbye code.

	* raptor/raptor_parse.c:
	Free former rdf:about, rdf:resource attribute string values before
	they are zapped.

	* raptor/raptor_parse.c (raptor_xml_parse_terminate):
	Delete the sax2 object when cleaning
	up.

	* raptor/rdfdump.c:
	Free new uri_string when it's allocated for a filename

	* raptor/rdfdump.c: Tidied usage and help information.

	* raptor/rapper.1: Updated to match current rapper arguments.

	* raptor/Makefile.am: Added raptor_identifier.c

	* raptor/raptor_general.c, raptor/raptor_identifier.c:
	Moved raptor_identifer classes to raptor_identifier.c

2003-07-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.c:
	If first argument is a filename, make it into a file:/// uri.

	* raptor/raptor_uri.c (raptor_default_new_uri):
	Turn probably-bad file:filename 'URIs' into
	proper file:///... etc. ones.

	* raptor/raptor_parse.c:
	Make use of forbidden rdf-namespaced property attributes into errors,
	as they should be.
	(raptor_forbidden_propertyAttribute_name): Now used.
	Reword some error messages.

	* raptor/tests/Makefile.am (check-bad-rdf):
	Note when bad test succeeds instead of failing

	* raptor/tests/Makefile.am: Added bad-13, bad-14

	* raptor/tests/bad-14.rdf:
	rdf:Description is not a legal property attribute

	* raptor/tests/bad-13.rdf: A property element cannot take rdf:about

	* raptor/raptor_uri.c (raptor_new_uri_for_xmlbase): Docs

2003-07-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/libraptor.3: formatting

	* raptor/libraptor.3: Added
	raptor_set_default_generate_id_parameters,
	raptor_set_generate_id_handler

	* raptor/raptor_general.c (raptor_set_default_generate_id_parameters):
	Fix it right this time.

	* raptor/raptor_general.c (raptor_set_default_generate_id_parameters):
	Fix base so the next
	generated ID uses the integer given, not integer+1

	* raptor/raptor_general.c (raptor_free_parser):
	Free any user-set genid prefix.

2003-07-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h:
	raptor_genid_type enum added - for RAPTOR_GENID_TYPE_BNODEID, BAGID
	Added prototypes for raptor_set_generate_id_handler,
	raptor_set_default_generate_id_parameters.

	* raptor/raptor_general.c (raptor_set_generate_id_handler):
	Added, to sent the generate ID
	handler implementation.
	(raptor_set_default_generate_id_parameters): Added, to sent the
	generate ID handler parameters for the default implementation
	("gen"+integer).
	(raptor_default_generate_id_handler): Added, moved default code from
	raptor_generate_id
	(raptor_generate_id): Use handler if it exists, otherwise the default
	implementation.

	* raptor/raptor_internal.h:
	Added generate_id_handler data parts to raptor_parser internals.

2003-06-24  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* redland-config.in: Make --libs pick the right libdir - add
	exec_prefix variable declaration.

	* utils/Makefile.am, librdf/Makefile.am, examples/Makefile.am:
	Use @LIBS@ in LIBS to pick up any system-wide use.

	* configure.ac:
	Pick up any external CPPFLAGS, LDFLAGS, LIBS to start with.

	* raptor/tests/Makefile.am: use top_builddir not ..

	* raptor/Makefile.am: add libraptor.la to rapper dependencies

2003-06-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* python/RDF.py: Statements, Model: more docs, usage.

	* python/Makefile.am: Added redlandtest.py

	* python/redlandtest.py: Redland Python unit tests

	* python/RDF.py: v0.8
	Updates from Matt Biddulph.
	(class Model) Now takes optional storage, if omitted, a memory
	storage model is created
	(class StreamWithContextIter, class IteratorWithContextIter,
	class IteratorIter, class StreamIter): Internal classes to create
	python iterators out of redland stream/iterators.
	(class Stream): New method context_iter that returns a python iterator
	returning (stream, context node) tuple on each iteration.
	(class HashStorage): Simple class to simplify creating hashed storage.
	(class MemoryStorage): Simple class to simplify creating in-memory
	storage.
	(class Parser): Name is now optional, defaults to raptor and RDF/XML.
	(class Parser): Methods parse_as_stream, parse_into_model will
	autoconvert uri arguments that are strings into redland URIs.

	* python/test/test.py: s/if not thing/if thing is None/

	* raptor/rdfdump.c:
	Added -e/--ignore-errors otherwise rapper stops parsing after 1st
	error using raptor_parse_abort.

2003-06-14  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_rss.c (raptor_rss_insert_identifiers):
	Back to being legal C99.

	* raptor/libraptor.3: Fix changes for 0.9.11

	* raptor/libraptor.3: Updated for stuff since 0.9.10

	* raptor/raptor_rss.c (raptor_rss_insert_identifiers):
	Init identifier after item

	* raptor/rdfdump.c, raptor/configure.ac:
	Tweak RSS Tag Soup parser words

2003-06-10  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Emit an error for namespace
	declarations that are RDF namespace URI plus some chars.  Emit
	a warning if a namespace is declared same as RDF one but 1 char
	short.

	* raptor/tests/Makefile.am: Added bad-12, warn-02

	* raptor/tests/bad-12.rdf, raptor/tests/warn-02.out,
	raptor/tests/warn-02.rdf: Check for bad rdf namespace URI
	declarations and warn if last char of RDF namespace URI omitted

2003-06-09  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_uri.c (librdf_new_uri_from_filename): Removed unused vars.

2003-06-08  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac:
	Fix the check for RSS parser requirements and report it more
	verbosely.

	* raptor/configure.ac:
	RSS parser only if libxml/reader.h present (for now).

	* raptor/configure.ac: libcurl reporting

	* raptor/raptor_parse.c (raptor_xml_parse_chunk_): Use sax2->first_read

	* raptor/raptor_parse.c (raptor_xml_parse_start,raptor_xml_parse_chunk_):
	Don't use first_read on newer libxml2.

2003-06-06  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_internal.h:
	raptor_sax2_s: Add first_read #if LIBXML_VERSION < 20425

	* raptor/raptor_rss.c (raptor_rss_insert_identifiers):
	Fix GCC-ism, declare variable at
	start of block.

	* raptor/raptor_parse.c (raptor_xml_parse_init):
	Move declaration of expat xp to start of
	function.

	* raptor/examples/raptor_abort.c, raptor/examples/grapper.c,
	raptor/strcasecmp.c, raptor/rdfdump.c, raptor/raptor_xml.c,
	raptor/raptor_www_test.c, raptor/raptor_www_libxml.c,
	raptor/raptor_www_libwww.c, raptor/raptor_www_curl.c,
	raptor/raptor_www.c, raptor/raptor_win32.c, raptor/raptor_utf8.c,
	raptor/raptor_uri.c, raptor/raptor_set.c, raptor/raptor_sax2.c,
	raptor/raptor_rss.c, raptor/raptor_qname.c, raptor/raptor_parse.c,
	raptor/raptor_namespace.c, raptor/raptor_locator.c,
	raptor/raptor_libxml.c, raptor/raptor_general.c,
	raptor/ntriples_parse.c, raptor/configure.ac: Merged patch from
	Jose Kahan to switch to use raptor_config.h (helps people
	compiling from source with multiple config.h)

2003-06-05  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/grapper.c (fs_ok_button_callback):
	Use raptor_uri_filename_to_uri_string
	to make the file URI.
	(main): If the argument is a filename, make a URI string out of it
	via raptor_uri_filename_to_uri_string and use it instead of
	assuming it is a URI.

2003-06-04  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Redland.i, librdf/rdf_uri.h: Added librdf_new_uri_from_filename

	* librdf/rdf_uri.c (librdf_new_uri_from_filename):
	Added using raptor_uri_filename_to_uri_string

	* autogen.sh: Report min SWIG version

2003-05-26  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* demos/rss-show.pl: words

2003-05-12  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_xml.c (raptor_xml_escape_string):
	Changed API - does not require a parser arg.

	* raptor/raptor_internal.h: Moved SAX2 parts and prototypes here.

	* raptor/raptor_general.c:
	Use RAPTOR_PARSER_RSS to wrap init of RSS parser

	* raptor/raptor.h:
	Changed prototype of raptor_xml_escape_string - does not require a
	parser arg.

	* raptor/raptor_parse.c:
	Split raptor_element/raptor_rdf_xml_parser into SAX2/RDF bits.
	Lots of function and structure renaming.

	* raptor/configure.ac:
	Define RAPTOR_PARSER_RSS only when libxml is around

	* raptor/raptor_sax2.c: SAX2 API

	* raptor/Makefile.am: Added raptor_sax2.c

2003-04-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/lib/RDF/Redland/Model.pm:
	Same as before but with right package names
	(find_statements): Return copies of shared statements
	(sources,arcs,targets): Return copies of shared nodes

	* perl/lib/RDF/Redland/Model.pm (find_statements):
	Return copies of shared statemetns
	(sources,arcs,targets): Return copies of shared nodes

2003-04-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_statement.c (librdf_statement_to_string):
	Handle buffer size counting when parts are null

	* librdf/rdf_heuristics.c (librdf_heuristic_gen_name):
	Check loop doesn't go before start.

	* librdf/rdf_utf8.c (librdf_utf8_print):
	check length first before looking in what might
	be a byte beyond the length

2003-04-28  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_rss.c (raptor_rss_emit): Use raptor_free_identifier

	* raptor/raptor_rss.c (raptor_rss_emit):
	Properly init the raptor_identifier items

2003-04-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_rss.c, raptor/raptor_parse.c, raptor/ntriples_parse.c:
	Use raptor_parser_register_factory with label param.

	* raptor/raptor.h: Added raptor_get_name, raptor_get_label

	* raptor/raptor_general.c: Added parser label to factory
	(raptor_parser_register_factory): Add label param, copy it.
	(raptor_get_name): Added, return name of parser.
	(raptor_get_label): Added, return label of parser.

	* raptor/raptor_internal.h: Added parser label to factory

	* raptor/examples/grapper.c: Remove some g_printfs
	Remove use of display qnames - not impl.

	* raptor/raptor_general.c (raptor_statement_part_as_counted_string):
	Init len for literals with
	the literal len included

	* raptor/examples/grapper.c: Don't use triples_list for now.
	(grapper_model_set_syntax): Fix output.
	(grapper_model_statements_handler): Remove newlines from literals.

	* raptor/raptor_general.c (raptor_statement_part_as_counted_string):
	Init len for literals.

2003-04-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_set.c: fix doccumment

2003-04-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/grapper.c:
	Replace N-Triples / RDF/XML with dropdown menu and add RSS tag soup.

2003-04-22  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/redland.spec.in: Updated for Redhat 9 and RPM 4

	* python/RDF.py: Added blank node is_blank() method example
	(Node.__set_attr): Fix set_type
	(Node.get_blank_identiifer, Node.set_blank_identiifer): Added

	* python/RDF.py: some perlisms in the docs/comments

	* python/RDF.py:
	this one file is now also under BSD without advertising/MIT license

2003-04-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.c: Added -i/--input for rdfxml, ntriples, rss

	* raptor/raptor_internal.h: More RAPTOR_DEBUG macros

	* raptor/raptor_general.c:
	(raptor_init) Added rss parser via raptor_init_parser_rss when
	HAVE_LIBXML_XMLREADER_H

	* raptor/Makefile.am: Added raptor_rss.c

	* raptor/raptor_rss.c: Raptor RSS parser

2003-04-17  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac: Bumped version to 0.9.11

	* raptor/libraptor.3: Added raptor_set_parser_strict
	Added raptor_www_no_www_library_init_finish

	* raptor/raptor.spec.in: Require curl
	Added raptor-config, raptor.pc

	* raptor/tests/Makefile.am:
	Move rdf:bagID tests to list of tests with warnings for now (while
	testing in lax mode)

2003-04-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* docs/fix-pod-xhtml: and a newline

	* docs/fix-pod-xhtml: added xml decl

	* docs/fix-pod-xhtml: More fixing

	* docs/fix-pod-xhtml: New style POD to html cleanup

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_register):
	Allow NULL name

	* perl/lib/RDF/Redland/Storage.pm: add rest of storage options to docs

	* raptor/rdfdump.c (print_statements):
	Print the program name not "rdfdump" hardcoded.

	* librdf/rdf_node.h, librdf/rdf_node.c:
	Added librdf_node_get_literal_value_as_counted_string

2003-04-14  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac: Check for libxml/xmlreader.h

2003-04-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/grapper.c: Added about box, triples count.
	Free some allocated memory.  Rest seems lost in gtk.

	* raptor/examples/Makefile.am: Add AM_CFLAGS, LIBS for debugging

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Don't copy an empty attributes array.

	* raptor/configure.ac: Added raptor.pc

	* raptor/Makefile.am:
	Added raptor.pc pkgconfig file installing to $(libdir)/pkgconfig

	* raptor/raptor.pc.in: pkgconfig for raptor

2003-04-12  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/Makefile.am: Added rdf_model_storage.h

	* librdf/rdf_model_storage.h: RDF Model Storage Interface definition

2003-04-10  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_model.c:
	Split model/storage into model/model factory/model storage factory
	just like hashes, storage classes.
	(librdf_init_model): Call librdf_init_model_storage
	(librdf_finish_model): Call librdf_delete_model_factories
	(librdf_delete_model_factories, librdf_model_register_factory,
	librdf_get_model_factory):  Added, after rdf_storage.c similar names.
	(librdf_new_model_with_options): Use storage factory create method.
	(librdf_new_model_from_model): Use clone method.
	(librdf_free_model): Use destroy method.
	The remaining methods that depend on storage now use factory methods.
	Some methods have slight edits to use the model method not the
	storage one when it doesn't matter, or when no factory method is
	present, and it is emulated at the model level.

	* librdf/Makefile.am: Added rdf_model_storage.c

	* librdf/rdf_model.h:
	Added internal librdf_model_factory struct with factory methods
	Added librdf_model_register_factory, librdf_get_model_factory

	* librdf/librdf.h: added librdf_model_factory

	* librdf/rdf_model_storage.c: RDF model over a storage

2003-04-05  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h: void arg

	* raptor/examples/raptor_abort.c: Zap curl cleanup.

	* raptor/rdfdump.c:
	Added -c flag to getopts - oops, missed in last release.
	Added -m/--mode flag to set strict/lax.  Check the values and die
	with usage.
	Check the legal values of -o/--output and die with usage.
	Zap curl cleanup.

	* raptor/raptor_www.c (raptor_www_no_www_library_init_finish):
	To control global WWW
	library init/finish

	* raptor/raptor_parse.c:
	Make bagID optional - removed from language and gives errors
	(strict), allowed with warnings (lax).
	lax/strict controlled by feature_allow_bagID

	* raptor/raptor_internal.h: Added feature_allow_bagID

	* raptor/raptor_general.c: Several more docucomments for functions.
	(raptor_new_parser): Use raptor_set_parser_strict
	(raptor_set_feature): Added RAPTOR_FEATURE_ALLOW_BAGID
	(raptor_set_parser_strict): Added to set strict/lax mode flags.

	* raptor/raptor.h: Added RAPTOR_FEATURE_ALLOW_BAGID
	Added raptor_set_parser_strict
	Added raptor_www_no_www_library_init_finish

2003-04-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am: Added warn-01

	* raptor/tests/warn-01.out, raptor/tests/warn-01.rdf:
	Check warning on non-prefixed property elements

	* raptor/raptor_parse.c, raptor/tests/Makefile.am
	(raptor_xml_start_element_handler): Error recovery - try to hide that
	a bad element was found from a parent element, it thinks the element
	is empty.  See tests/warn-00.rdf

	* raptor/tests/warn-00.out, raptor/tests/warn-00.rdf: Check warnings

2003-04-02  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Makefile.am, raptor/configure.ac, raptor/Makefile.am, configure.ac:
	debian dir elsewhere

2003-04-01  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_parser_raptor.c
	(librdf_parser_raptor_new_statement_handler): (re)add storing in a
	model direct.
	(librdf_parser_raptor_parse_uri_as_stream_write_bytes_handler): Added
	to support librdf_parser_raptor_parse_uri_as_stream via raptor_www
	(librdf_parser_raptor_parse_uri_as_stream): Added.  Calls
	librdf_parser_raptor_parse_file_as_stream if that's possible.
	(librdf_parser_raptor_parse_uri_into_model): Added, doing all of
	librdf_parser_raptor_parse_file_into_model.
	(librdf_parser_raptor_constructor): Register new functionality.

	* raptor/raptor_uri.c (raptor_uri_uri_string_to_filename):
	Use raptor_strcasecmp
	(raptor_uri_is_file_uri): Use raptor_strncasecmp

	* librdf/rdf_parser_libwww.c: No libwww parser

2003-03-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/fix-groff-xhtml: more cols

	* raptor/fix-groff-xhtml: More fixes

	* raptor/Makefile.am: Added libraptor.html, fix-groff-xhtml

	* raptor/fix-groff-xhtml: Fix groff output HTML

	* raptor/libraptor.3: deleted repeated URI METHODS

	* raptor/libraptor.3: Updated for 0.9.6->present

2003-03-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/grapper.c: Use N-triples output style.
	Pass in URL command line argument

	* raptor/Makefile.am: Added raptor_xml.c

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	do XML attribute value
	normalization for libxml2.  Cannot be done properly since the type of
	the attribute is lost.  expat gets it right.
	(raptor_xml_parse_init): Cast for expat XML_SetExternalEntityRefHandler

	* raptor/raptor_xml.c:
	Updated for C14N text/attribute node encoding rules.  Don't de-UTF8
	at the same time.

	* raptor/tests/ex-49.out: > in attribute appears raw

	* raptor/raptor_general.c:
	(main) Moved raptor_validate_xml_ID, raptor_xml_escape_string test
	code to raptor_xml.c

	* raptor/raptor_xml.c: Raptor XML routines

	* raptor/raptor_general.c:
	Moved raptor_validate_xml_ID, raptor_xml_escape_string to raptor_xml.c

	* raptor/tests/Makefile.am: Added ex-49

	* raptor/tests/ex-49.out, raptor/tests/ex-49.rdf:
	Checking escaping in parseType Literal values with XML attributes

	* raptor/raptor_parse.c (raptor_format_element):
	Pass in parser for UTF-8 error handling.
	Use raptor_xml_escape_string for attribute values but
	only malloc/free if lengths changed.
	(raptor_xml_end_element_handler): Update to new
	raptor_xml_escape_string API and only malloc/free it if lengths
	changed.

	* raptor/raptor_parse.c (raptor_xml_cdata_handler):
	Use updated raptor_xml_escape_string API

	* raptor/raptor.h: Added raptor_statement_part_as_counted_string,
	raptor_statement_part_as_string
	Changed raptor_xml_escape_string API

	* raptor/raptor_general.c
	(raptor_statement_part_as_counted_string,
	raptor_statement_part_as_string): Added, making N-triples style
	output from parts of raptor_statement.
	(raptor_xml_escape_string): Change API to take an existing
	buffer/calculate length
	(main): Update for raptor_xml_escape_string.

	* raptor/examples/Makefile.am: Also clean grapper binary

2003-03-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/grapper.c: Added N-Triples parsing

	* raptor/examples/Makefile.am: Don't build grapper usually

	* raptor/examples/grapper.c: Raptor GTK GUI example code

	* python/RDF.py, librdf/rdf_parser_raptor.c, librdf/rdf_parser.h,
	librdf/rdf_parser.c, librdf/Makefile.am, docs/redland.pod,
	docs/api.sgml.in, docs/Makefile.am, demos/parser-tests.pl,
	demos/demo.pl, configure.ac: Removed repat

	* librdf/rdf_uri.c (librdf_free_uri): Move world defn inside ifdef

	* raptor/examples/Makefile.am: Added grapper.c

	* raptor/examples/Makefile.am: more deps

	* raptor/raptor_internal.h:
	Added raptor_print_statement_part_as_ntriples prototype

	* raptor/raptor_general.c:
	raptor_print_ntriples_string moved to raptor_general.c
	(raptor_print_statement_part_as_ntriples): Added, internal.
	(raptor_print_statement_as_ntriples): Now uses above.

	* Makefile.am: Removed repat

	* raptor/ntriples_parse.c:
	raptor_print_ntriples_string moved to raptor_general.c

	* librdf/Makefile.am: die recursive make

2003-03-28  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/examples/Makefile.am: another deps attempt

	* raptor/examples/Makefile.am: typo

	* raptor/examples/Makefile.am: Use LDADD and hunt for @REDLAND_LIBS@
	Added $(top_builddir)/../librdf/librdf.la

	* raptor/configure.ac: Hunt for librdf.la in abs dir

	* raptor/rapper.1: added -c/--count

	* raptor/examples/Makefile.am: Fixes to make cross-dir building work.

	* raptor/configure.ac, raptor/Makefile.am: Added examples dir

	* raptor/raptor_general.c (raptor_parse_uri_with_connection): Added.
	(raptor_parse_abort): Added.

	* raptor/raptor.h: Added raptor_parse_uri_with_connection.
	Added raptor_parse_abort.

	* raptor/examples/raptor_abort.c, raptor/examples/Makefile.am: examples

	* raptor/rdfdump.c:
	Call curl_global_cleanup if using curl to free it's resources.

	* raptor/raptor_www.c (raptor_www_init,raptor_www_finish):
	Don't init/cleanup curl, we
	can't guarantee doing this at most once if a handle is passed in.

	* raptor/raptor_www_curl.c (raptor_www_curl_free): Tidy

	* raptor/raptor_www_curl.c (raptor_www_curl_init,raptor_www_curl_free):
	Use/mark field
	curl_init_here to note when to cleanup a handle - don't destroy one
	that was provided.

	* raptor/raptor_internal.h:
	for curl, record when curl_easy_init was done in raptor

	* raptor/raptor_general.c (raptor_init): Call raptor_www_init
	(raptor_finsh): Call raptor_www_finish
	(raptor_parse_uri): Delete www object on failure.

	* raptor/configure.ac: Bumped version to 0.9.10

	* librdf/rdf_model.c: unused var

	* librdf/rdf_stream.c, librdf/rdf_statement.c,
	librdf/rdf_serializer.c, librdf/rdf_query.c, librdf/rdf_parser.c,
	librdf/rdf_node.c, librdf/rdf_iterator.c, librdf/rdf_hash.c,
	librdf/rdf_concepts.c:
	Changes to test code after initial locking for threads support.
	(main) Testcode - use librdf_world_init_mutex

	* librdf/rdf_uri.c: Initial locking for threads from Seth Ladd
	(librdf_new_uri): Lock/unlock world for new URI.
	(librdf_free_uri): Lock/unlock world for destroying URI.
	(main) Testcode - use librdf_world_init_mutex

	* librdf/rdf_init.h: Initial locking for threads from Seth Ladd
	Added mutex to world.
	Added prototype for internal librdf_world_init_mutex

	* librdf/rdf_init.c: Initial locking for threads from Seth Ladd
	(librdf_free_world): Delete mutex on shutdown.
	(librdf_world_init_mutex): Added (internal) to be used by tests
	rather than initing all of redland.
	(librdf_world_open): Create mutex on startup.

	* configure.ac: Initial locking for threads from Seth Ladd
	Added --with-threads for POSIX threads

2003-03-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* demos/demo.pl: utf-8 output

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_grow_buffer):
	Fix return code sense.
	(librdf_storage_hashes_add_remove_statement): Use calculate length
	for key/value len, not buffer sizes.
	(librdf_storage_hashes_context_add_statement): Return failure if
	contexts aren't enabled and this is called.

	* librdf/rdf_model.c:
	(main) Test code - check get_sources, get_targets, get_arcs

	* python/RDF.py (Stream.context): Use self.stream

2003-03-26  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_set.c:
	Only use raptor_set_stats_print with RAPTOR_DEBUG

	* raptor/raptor_general.c: Some comment tidying.
	(raptor_start_parse): Docucomment.  Copy the uri into the parser
	structure (base_uri, locator uri), don't just keep the pointer.
	(raptor_free_parser): Free the base URI in the structure, if
	present.

	* raptor/raptor_parse.c (raptor_xml_parse_start):
	Remove uri arg.   Fail if no base URI is
	given - stored in the parser object

	* raptor/ntriples_parse.c (raptor_ntriples_parse_start):
	Remove uri arg.

	* raptor/raptor_internal.h:
	raptor_parser_factory start method: Remove (base) uri arg, it's in
	the object data.

	* raptor/raptor_www.c: Use RAPTOR_FREE,MALLOC,CALLOC
	(raptor_www_free): Free the www object.  Doh.

	* raptor/libraptor.3: raptor_parse_chunk: Takes unsigned char buffer.

	* raptor/libraptor.3:
	raptor_start_parse: Note NULL base URI ok for ntriples

	* raptor/libraptor.3: Typo: raptor_start_parse not raptor_parse_start

	* raptor/raptor_general.c (raptor_xml_escape_string):
	Make it work with 10ffff again

	* raptor/raptor_parse.c (raptor_xml_cdata_handler):
	Cast around raptor_xml_escape_string

	* raptor/raptor_general.c (raptor_xml_escape_string):
	Null terminate new string

	* raptor/raptor_www_libxml.c, raptor/raptor_www.c:
	Use RAPTOR_WWW_BUFFER_SIZE for I/O buffers

	* raptor/raptor_internal.h:
	Define RAPTOR_WWW_BUFFER_SIZE for I/O buffers

	* raptor/rdfdump.c: Declare raptor_stats_print when RAPTOR_DEBUG

	* raptor/raptor_general.c: (raptor_stats_print) C99

	* librdf/rdf_model.c (librdf_model_add_typed_literal_statement):
	Don't free object here,
	librdf_model_add does it as part of freeing the statement.

2003-03-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* demos/rss-show.pl, demos/parser-tests.pl, demos/ntriples.pl,
	demos/demo.pl: Updated to 0.9.12 API

2003-03-24  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.c: Call raptor_stats_print with RAPTOR_DEBUG

	* raptor/raptor_set.c: Record set hits/misses with RAPTOR_DEBUG
	(raptor_set_stats_print): Defined with RAPTOR_DEBUG

	* raptor/raptor_internal.h:
	Move raptor_xml_parser typedef here (still internal).
	raptor_xml_parser_stats_print, raptor_set_stats_print: Defined with
	RAPTOR_DEBUG

	* raptor/raptor_parse.c (raptor_xml_parser_stats_print):
	Defined with RAPTOR_DEBUG

	* raptor/raptor_general.c (raptor_stats_print):
	Defined with RAPTOR_DEBUG

	* raptor/raptor_parse.c (raptor_xml_cdata_handler):
	Use raptor_xml_escape_string when content
	type is an XML literal

	* raptor/raptor_general.c (raptor_xml_escape_string):
	Now takes and returns lengths
	Fix assumption of '\0' terminated strings.
	(main): Update for counted strings

	* raptor/raptor.h:
	raptor_xml_escape_string now takes and returns lengths

	* raptor/raptor.h: Added raptor_xml_escape_string

	* raptor/Makefile.am: Added raptor_general_test

	* raptor/raptor_general.c (raptor_xml_escape_string):
	Added, XML-escapes UTF-8 strings.
	(main): Added set of tests for raptor_xml_escape_string

2003-03-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/ex-48.out, raptor/tests/ex-48.rdf:
	ex-48 parse type literal with entity encoding

	* raptor/tests/Makefile.am: Added ex-48

	* raptor/raptor_www_libxml.c (raptor_www_libxml_fetch):
	Make this work again

	* raptor/raptor_parse.c (raptor_record_ID): Dealloc item after adding

	* librdf/rdf_parser_raptor.c: raptor_free now raptor_free_parser
	(librdf_raptor_uri_as_counted_string): Added
	Use it to register the raptor_uri

	* raptor/raptor_parse.c: typo

	* raptor/raptor_parse.c:
	Remove raptor_id_list implementation for ID checking to use raptor_set.
	(raptor_xml_parse_init): Init raptor_set for ids.
	(raptor_xml_parse_terminate): Use raptor_free_set.
	(raptor_record_ID): Use raptor_set_add to check for unique "ID base-URI"
	(raptor_free_ID_list): Gone

	* raptor/raptor_uri.c (raptor_default_uri_as_counted_string,
	(raptor_uri_as_counted_string): Added and used in default factory.

	* raptor/raptor.h: Added raptor_uri_as_counted_string.

	* raptor/raptor_internal.h:
	Added raptor_set and raptor_new_set constructor, raptor_free_set
	destructor and raptor_set_add only method

	* raptor/Makefile.am: Added raptor_set.c, raptor_set_test

	* raptor/raptor_set.c: Raptor sets for ID checking

2003-03-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* demos/rss-show.pl: sort feeds case insensitively

	* demos/rss-show.pl: Added Annotate module

	* librdf/rdf_storage_hashes.c:
	Add key/value buffers to hash context and share/grown them as needed
	to prevent a lot of unncessary malloc/frees.
	(librdf_storage_hashes_terminate): Delete shared key/value buffers.
	(librdf_storage_hashes_grow_buffer): Added to grow them when needed.
	(librdf_storage_hashes_add_remove_statement): Use
	librdf_storage_hashes_grow_buffer and shared key/value buffers.

	* librdf/rdf_hash_memory.c (librdf_hash_memory_expand_size):
	Revert bucket rehashing code to the older, simpler, slightly
	slower (2x work) since changed version failed.

	* librdf/rdf_hash_memory.c:
	Switched to same hash algorithm as perl, inlined it.
	(librdf_hash_memory_expand_size): When expanding, only move the nodes
	that need it, don't move everything.

2003-03-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* examples/example4.c: Free nodes on failure of get_source/target/arc

	* raptor/tests/Makefile.am: Added OWL namespace document / rdf
	schema owl-schema.rdf, owl-schema.out

	* raptor/raptor_www.c: No ghttp, init w3c libwww

	* librdf/rdf_statement.c (librdf_statement_to_string):
	Init statement_string_len to 0

	* librdf/rdf_hash_memory.c (librdf_hash_memory_expand_size):
	Quick hack to make hash grow.
	Needs a proper fix.

	* librdf/rdf_statement.c (librdf_statement_to_string):
	Use librdf_node_to_counted_string to save on strlen calls.

	* librdf/rdf_node.c (librdf_node_to_string):
	Now uses librdf_node_to_counted_string.
	(librdf_node_to_counted_string): Added, returns already known length.
	Uses librdf_uri_to_counted_string to save on strlen calls.
	(librdf_node_encode): Use librdf_uri_as_counted_string to save on
	strlen calls.

	* librdf/rdf_node.h: Added librdf_node_to_counted_string

	* librdf/rdf_uri.c (librdf_uri_as_counted_string,
	librdf_uri_to_counted_string): Added returning the already
	stored/calculated length of the string.

	* librdf/rdf_uri.h:
	Added librdf_uri_as_counted_string, librdf_uri_to_counted_string

	* raptor/tests/owl-schema.out, raptor/tests/owl-schema.rdf:
	OWL namespace schema

	* raptor/raptor_www_libwww.c: add fatal does-not-work error

	* raptor/configure.ac, raptor/raptor_internal.h,
	raptor/raptor_www_ghttp.c: Delete libghttp

2003-03-18  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_www_libwww.c: Raptor WWW with W3C libwww

	* raptor/rdfdump.c:
	Added -c/--count option to just count triples, don't print anything.

	* raptor/raptor_www_test.c: Use raptor_www_init/finish

	* raptor/raptor_www_libxml.c, raptor/raptor_www_ghttp.c:
	Handle www->failed and aborting transfer.

	* raptor/raptor_www_curl.c: Handle www->failed and aborting transfer.
	(raptor_www_curl_init): Use passed-in connection if available.

	* raptor/raptor_www.c (raptor_www_init): Added, for once-only init.
	(raptor_www_finish): Added, for once-only tidy.
	(raptor_www_new): Now uses new constructor
	(raptor_www_new_with_connection): Added, allows re-use of existing
	www library connection - just curl at present.
	(raptor_www_get_connection): Added, returns current libwww library
	connection.
	(raptor_www_abort): Added to stop a www transaction.
	(raptor_www_file_fetch): Tidying of errors; handle abort.
	(raptor_www_fetch): Uses raptor_www_file_fetch for all files.

	* raptor/raptor_internal.h: Added W3C libwww prototypes.

	* raptor/raptor_parse.c:
	Throughout all SAX event handlers - if rdf_parser->failed set, return
	immediately, doing no work.

	* raptor/raptor_general.c (raptor_parse_uri_write_bytes):
	Use raptor_www_abort if parsing
	fails.
	(raptor_parse_uri): Return error status.  Pass on is_end empty chunk.
	(raptor_parser_abort): Added, setting failed flag.
	(raptor_print_statement_detailed): Typo

	* raptor/raptor.h: Added raptor_www_abort

	* raptor/raptor.h: Add raptor_www_init, raptor_www_finish.
	Add raptor_www_new_with_connection
	Add raptor_www_get_connection

	* raptor/configure.ac: Added w3c libwww configuring

2003-03-16  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.c: Use raptor_parse_uri

	* raptor/raptor_general.c (raptor_parse_uri_write_bytes):
	Added, to support:
	(raptor_parse_uri): Added, using raptor_www to get and deal with all
	the data in one go.

	* raptor/raptor.h: raptor_parse_uri takes optional base_uri

	* raptor/raptor_www_test.c: Use URI from context.
	Take www arg on handlers
	Use raptor_uri

	* raptor/raptor_www_libxml.c, raptor/raptor_www_ghttp.c, raptor/raptor_www_curl.c:
	Use URI from context.
	Take www arg on handlers

	* raptor/raptor_www.c (raptor_www_set_userdata): Gone
	(raptor_www_free): Free request uri
	(raptor_www_set_write_bytes_handler,
	raptor_www_set_content_type_handler): Added
	(raptor_www_file_fetch): pass www to write_bytes
	(raptor_www_fetch): Don't pass URI on.

	* raptor/raptor_internal.h: Store raptor_uri of request
	Use new declared write_bytes, content_type handlers
	*fetch methods don't take URI string

	* raptor/raptor.h:
	Declare handlers for raptor www write bytes, content type
	raptor_www_fetch now takes a raptor_uri

	* raptor/raptor_parse.c:
	Fix broken-fix for broken-expat UTF8 BOM crash.
	tokens_count is on the rdf_xml_parser not rdf_parser

	* raptor/configure.ac: tweak

	* raptor/configure.ac: tidy libcurl version

	* raptor/raptor_internal.h, raptor/configure.ac:
	No more gnome-xml/libxml.h

	* raptor/configure.ac: Min libxml2 now 2.4.0

	* raptor/configure.ac: Don't look for xml-config

	* raptor/raptor_general.c (raptor_parse_file):
	Tidy up if raptor_start_parse fails

	* raptor/raptor_general.c:
	Removed raptor_start_parse_file - merged into raptor_parse_file

	* raptor/raptor_www.c (raptor_www_file_fetch): Used for RAPTOR_WWW_NONE
	Report file open errors, correct file read eof handling.
	(raptor_www_fetch) Use only raptor_www_file_fetch for RAPTOR_WWW_NONE

	* raptor/raptor_general.c (raptor_start_parse_file):
	Improve file open error message

	* raptor/configure.ac:
	Added --with-www=none option and RAPTOR_WWW_NONE to indicate it

2003-03-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_www.c (raptor_www_error): Use RAPTOR_FREE

	* raptor/raptor_www_ghttp.c (raptor_www_ghttp_fetch):
	call raptor_www_error

	* raptor/raptor_www_test.c: Call raptor_uri_init

	* raptor/raptor_www_curl.c (raptor_www_curl_fetch):
	call raptor_www_error

	* raptor/raptor_internal.h: Use raptor_message_handler again

	* raptor/raptor.h:
	Use raptor_message_handler again in raptor_www_set_error_handler,
	raptor_www_error

	* raptor/raptor_www.c (raptor_www_free): Tidy locator URI
	(raptor_www_set_error_handler, raptor_www_error): Use
	raptor_message_handler again.
	(raptor_www_error): Use raptor_locator in output, error handler.
	(raptor_www_fetch): Store the URI string of request in the locator

	* raptor/raptor.h: Declare raptor_www_message_handler (no locator)

	* raptor/raptor_general.c: raptor_vsnprintf now internally visible.

	* raptor/raptor_internal.h:
	Use different error handler, no parser context.
	raptor_vsnprintf now internally visible.
	Added raptor_www_error prototype

	* raptor/raptor_www.c:
	Only enable raptor_www_file_fetch with libxml, libghttp
	(raptor_www_set_error_handler): Use different error handler, no
	parser context.

	* raptor/configure.ac, raptor/Makefile.am:
	Added raptor WWW enabling, configuring

	* raptor/raptor_internal.h:
	Added raptor WWW retrieval internal includes, structs, prototypes

	* raptor/raptor.h: Added raptor WWW retrieval prototypes

	* raptor/raptor_www.c, raptor/raptor_www_curl.c,
	raptor/raptor_www_ghttp.c, raptor/raptor_www_libxml.c,
	raptor/raptor_www_test.c: Raptor WWW retrieval

2003-03-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_parser_raptor.c (librdf_parser_raptor_parse_file_as_stream):
	Save copies of source
	and base URI - don't assume they will be around for the life of a
	parse.
	(librdf_parser_raptor_serialise_finished): Free the copied uris.

	* examples/example4.c: document parse-stream

	* python/RDF.py (Node.__init__):
	Add world arg to librdf_new_node_from_blank_identifier

2003-03-09  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Makefile.am: added demos dir

	* demos/rss-show.pl, demos/ntriples.pl: absolutize some doc urls

	* configure.ac, Makefile.am: added demos dir

	* demos/rss-show.pl, demos/parser-tests.pl, demos/ntriples.pl,
	demos/demo.pl: point to source code

	* demos/rss-show.pl, demos/parser-tests.pl, demos/ntriples.pl,
	demos/demo.pl: add source code

	* demos/Makefile.am, demos/demo.pl, demos/ntriples.pl,
	demos/parser-tests.pl, demos/rss-show.pl: Redland CGI demos

2003-03-06  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/serialize.pl (format_element_content,format_attribute):
	Added to escape &, <, > things  etc.

2003-03-05  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/serialize.pl: Whitespace tidying.
	(emit_literal_element): Added, for doing 1-line literals
	Fix datatypes.
	Emit multiple rdf:type properties too.

	* python/RDF.py:
	(Node.get_literal_value) Correct datatype URI method call to only
	operate if there is a datatype URI.
	(Stream._del_): Work when there is no redland Stream.
	(Parse.parse_as_stream): Return None if parsing fails, not an
	empty Stream.

	* perl/lib/RDF/Redland/Node.pm (literal_datatype):
	Added, return the URI of the literal datatype or
	undef if not one.

	* perl/lib/RDF/Redland/Parser.pm (parse_as_stream):
	Return undef if parsing fails, don't return a stream

	* perl/lib/RDF/Redland/Stream.pm (DESTROY):
	Handle wrapping an empty stream - might happen when a
	search fails straight away.

	* librdf/rdf_parser_raptor.c
	(librdf_parser_raptor_new_statement_handler): Handle XMLLiteral
	and datatyped literal URIs - pass them on as appropriate statements.

	* perl/serialize.pl (predicate_split): Fix
	(emit_empty_element) Pull out some more element
	building stuff - for now.
	main - declare an OWL namespace just for testing

	* perl/serialize.pl: added emit_start_element, emit_end_element
	now rdf:RDF declares its namespaces

	* perl/serialize.pl: Make blank node subject and objects work.

2003-03-04  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h: Export raptor_free_parser with RAPTOR_API

	* raptor/win32_config.h: add trailing #endif

	* raptor/win32_config.h: s/WIN32_LEAD_AND_MEAN/WIN32_LEAN_AND_MEAN/

	* raptor/raptor_parse.c (raptor_xml_comment_handler):
	Added - nop at present.
	(raptor_xml_parse_init): For expat, use raptor_xml_comment_handler

	* raptor/raptor_libxml.c:
	(raptor_libxml_init) Use raptor_xml_comment_handler

	* raptor/raptor_internal.h: Add raptor_xml_comment_handler prototype

	* raptor/configure.ac:
	Try to make maintainer mode flags match redland's defaults.

	* Makefile.am: expat sources not included

	* configure.ac: expat sources not included
	building against expat / libxml sources no longer supported.

	* librdf/rdf_model.c (librdf_model_add_typed_literal_statement):
	Document - nodes become owned by the model.

	* python/RDF.py: Bogus quotes

	* java/org/librdf/redland/Model.java:
	(add) (nodes, typed nodes versions) Copy redland nodes before adding.

	* perl/t/test2.t:
	Added test of Model method add_typed_literal_statement
	Added test of Serializer method serialize_model_to_file

	* python/test/test.py: Added test of model.add_typed_literal_statement

	* python/RDF.py (Model.add_typed_literal_statement):
	Copy redland nodes before adding.

2003-03-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* perl/Makefile.am: make clean targets run when perl isn't enabled.

	* perl/Makefile.PL: Grab version from redland-src-config

	* perl/Makefile.am: Use $(srcdir)/Makefile.PL

	* python/Makefile.am: Use $(srcdir)/setup.py

	* java/Makefile.am:
	clean-java - don't leap down subdirs to clean stuff too early -
	confuses autoconf.

	* configure.ac: More AC_CONFIG_FILES tidying

	* raptor/configure.ac, raptor/Makefile.am:
	Fix cflags/cppflags when in redland

	* librdf/rdf_serializer_raptor.c
	(librdf_serializer_print_statement_as_ntriple):
	No need for fprintf.

	* librdf/rdf_serializer_raptor.c: No need for ntriples.h
	(librdf_serializer_print_statement_as_ntriple): Handle datatyped
	literals.

	* perl/lib/RDF/Redland/Node.pm (new_from_typed_literal):
	Added, creating a datatyped literal.

	* perl/lib/RDF/Redland/Model.pm (add, add_typed_literal_statement):
	Copy the nodes before passing on
	- more natural for perl.

	* librdf/Makefile.am: s/AML_CFLAGS/AM_CFLAGS/
	Give rdf_uri_test the raptor includes

	* librdf/rdf_node.c (librdf_node_decode):
	Correct datatype URI/lang decoded in wrong
	order.
	(main): Add de/encode of datatyped node.

	* librdf/rdf_node.c (librdf_node_encode):
	Copy datatype URI string not junk from the URI
	object.
	(main): Add use of librdf_new_node_from_typed_literal

	* examples/example4.c:
	Added add-typed command to use librdf_new_node_from_typed_literal

	* autogen.sh: Require autoconf 2.53+

	* configure.ac: Require autoconf 2.53+
	Update AC_OUTPUT to new form

	* Redland.i: Added librdf_new_node_from_typed_literal

	* redland-src-config.in:
	use abs_top_srcdir for all refs to raptor-src-config

2003-03-02  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/ex-46.out, raptor/tests/ex-46.rdf:
	make rdf:li and rdf_2 property elements generate different triples

2003-02-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/ex-46.out: fix

	* raptor/tests/Makefile.am: Added ex-47

	* raptor/tests/ex-47.out, raptor/tests/ex-47.rdf:
	Exercise all rdfs vocab

	* raptor/tests/Makefile.am: Added ex-46

	* raptor/tests/ex-46.out, raptor/tests/ex-46.rdf:
	Exercise all rdf vocab

2003-02-26  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* python/RDF.py (RDF.set_message_handler): Added. Make a class method
	(message_handler): Moved out of RDF.World class.
	(RDF.World): Use RDF.message_handler
	(RDF.Stream): Docs.

	* librdf/rdf_storage_hashes.c
	(librdf_storage_hashes_node_iterator_create):
	Handle node being NULL,
	for example when dealing with a predicates index p2so.

2003-02-25  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* repat/Makefile.am: Make only librepat.la

	* Makefile.am: Add man pages to dist

2003-02-24  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/win32/raptor.plg, raptor/win32/Makefile.am: deleted raptor.plg

	* raptor/win32/raptor.dsp, raptor/win32/README:
	Updated win32 config - from contributed patches

2003-02-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_uri.c: Correct :'s turning into |'s in win32 file URIs

2003-02-20  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* autogen.sh: export autotools to configure

	* autogen.sh: Tidying

	* raptor/autogen.sh, autogen.sh:
	run libtoolize in each configure.ac dir

	* Makefile.am: acconfig.h not needed

	* utils/Makefile.am, repat/Makefile.am, librdf/Makefile.am,
	expat/xmltok/Makefile.am, expat/xmlparse/Makefile.am,
	examples/Makefile.am:
	Updates for automake 1.5-1.7

	* autogen.sh: Update to use automake 1.6, autoconf 2.52 min

	* configure.ac: configure.ac replacing configure.in

	* acconfig.h, configure.in:
	Replace configure.in acconfig.h with configure.ac

	* raptor/win32_config.h:
	win32 has C99 compatible vsnprintf called _vsnprintf

	* raptor/configure.ac:
	Check for vsnprintf and check for C99 compatible return value.

	* raptor/raptor_general.c (raptor_vsnprintf):
	Added for handling compatibilty with vsnprintf
	that doesn't match C99.

2003-02-19  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/tests/Makefile.am:
	Use $(ECHO) which may be different from sh's echo

	* raptor/Makefile.am:
	Changing, moved raptor_getopt.h to rapper_SOURCES after automake
	manual recommendation.

	* raptor/Makefile.am: Put raptor_getopt.h in noinst_HEADERS

	* raptor/Makefile.am: No need for EXTRA_libraptor_la_SOURCES

	* raptor/Makefile.am: Use LTLIBOBJS for extra libraptor objs

	* raptor/raptor.h: again

	* raptor/raptor.h: don't do deprecated on broken OSX gcc

	* raptor/raptor_getopt.h, raptor/getopt.c: No need for prefix

	* raptor/getopt.c: Duh - use raptor getopt header

	* raptor/Makefile.am:
	Use RAPPER_EXTRA_OBJS to optionally add getopt to rapper only

	* raptor/configure.ac: Add getopt object to RAPPER_EXTRA_OBJS

	* raptor/rdfdump.c: Add raptor_getopt.h for local version

	* raptor/raptor_getopt.h: Define rest of getopt externs

	* raptor/getopt.c: More prefixes

	* raptor/raptor_getopt.h, raptor/getopt.c: Public domain getopt

	* raptor/configure.ac: Check for getopt and add getopt.o if it missing.

	* raptor/autogen.sh:
	try asking the progs for their version - slower, but right

	* raptor/autogen.sh: tidying

	* raptor/acconfig.h: acconfig.h obsoleted

	* raptor/Makefile.am: Use AM_CPPFLAGS

	* raptor/autogen.sh:
	Min versions are bumpled - 2.52 for autoconf, 1.6 for automake
	Now hunts for newest, shinyest autoconf and automake/aclocal and uses
	them whatever is available.

2003-02-18  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/configure.ac: autoconf 2.5x configure.ac

	* raptor/configure.in: Replaced with autoconf 2.5x configure.ac

2003-02-15  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor.h: Move stuff around, consolidate deprecated bits

	* raptor/ntriples_parse.c, raptor/rdfdump.c: No need for ntriples.h

	* raptor/raptor.h: Moved all ntriples.h defines here.
	Defined RAPTOR_DEPRECATED (with gcc 3.1+) and used on old api calls.

	* raptor/ntriples.h:
	Moved all definitions to raptor.h - this file is now deprecated.

	* python/RDF.py: typo

2003-02-14  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* debian/changelog, debian/libredland0-dev.install:
	added redland-config.1

	* debian/libredland0-dev.manpages: Added redland-config.1

	* debian/Makefile.am: Added libredland0-dev.manpages

	* Makefile.am: added redland-config.1

	* redland-config.1: redland-config manual page

	* redland-config.in, raptor/raptor-config.in: No exec-prefix

	* raptor/raptor-config.1: Manual page for raptor-config

	* raptor/Makefile.am: Added raptor-config.1

	* debian/Makefile.am: missed a backslash

	* debian/redland1.files, debian/manpage.1.ex, debian/shlibs.local.ex:
	Not used

	* debian/rules.in, debian/copyright, debian/control,
	debian/changelog, debian/README.Debian, debian/Makefile.am:
	Updated for debian packaging 0.9.12

	* debian/libredland0-dev.docs, debian/libredland0-dev.files,
	debian/libredland0-dev.install, debian/libredland0.files,
	debian/libredland0.install, debian/libredland0.manpages:
	Redland debian packaging

	* debian/doc-base.package, debian/docs, debian/dirs: Not used

	* raptor/rapper.1: EXAMPLE

2003-02-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Redland.i: Added librdf_node_get_literal_value_datatype_uri

	* raptor/tests/Makefile.am: $name=>$$name

	* configure.in: Bumped version to 0.9.13

	* Snapshotted redland_0_9_12 for 0.9.12 release

	* raptor/raptor.spec.in: release is 1

2003-02-12  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/redland.spec.in, Makefile.am: No RELEASE, TODO

	* raptor/Makefile.am: Restore LICENSE.txt

	* raptor/Makefile.am: No LICENSE.txt in dist

	* java/org/librdf/redland/Makefile.am: Add Hash.ajva to dist

	* librdf/redland.spec.in:
	Move man pages around.  All headers now in devel.

	* librdf/redland.spec.in: Don't put raptor man files in this rpm

	* librdf/redland.spec.in: Require raptor 0.9.8

	* raptor/raptor_uri.c (main):
	Use lstat, reading a selection of dirs to try harder to test
	relative file URIs.

	* raptor/configure.in: check for sys/stat.h (for raptor_uri_test main)

	* raptor/tests/Makefile.am: There is no portable test == operator

	* raptor/raptor_parse.c (raptor_xml_start_element_handler):
	Updated call to
	raptor_namespaces_start_namespace to use error_handler, error_data
	parameters

	* raptor/raptor_general.c (raptor_start_parse):
	Updated call to raptor_namespaces_init to use
	error_handler, error_data parameters

	* raptor/raptor_internal.h:
	Updated raptor_namespaces_init, raptor_namespaces_start_namespace,
	raptor_namespace_new with error_handler, error_data parameters

	* raptor/raptor_namespace.c: (raptor_namespaces_init,
	raptor_namespaces_start_namespace, raptor_namespace_new) 
	Add error handler and data parameters and use to explain how
	xmlns:foo="" isn't allowed

	* utils/Makefile.am: Add update-api-0912.pl to dist

2003-02-11  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_parse.c:
	multiple objects of a property element (statement) is an error

	* raptor/tests/Makefile.am:
	Report warnings when they occur.  Print the warning text

	* raptor/tests/bad-11.rdf: now xml

	* raptor/tests/Makefile.am: typo - restult

	* raptor/tests/bad-11.rdf:
	Check xmlns with no namespace name (URI) fails

	* raptor/tests/Makefile.am: Added bad-11

	* raptor/raptor_libxml.c (raptor_libxml_update_document_locator):
	Handle empty parser context.

	* raptor/tests/Makefile.am: ex-46 now bad-10

	* raptor/tests/bad-10.rdf, raptor/tests/ex-46.rdf,
	raptor/tests/ex-46.out: Now bad-10

	* raptor/raptor_parse.c (raptor_xml_parse_chunk_):
	Make an empty rdf/xml bytestream an
	illegal doc (like expat) and return an error.

	* raptor/raptor_libxml.c (raptor_libxml_update_document_locator):
	Don't use loc if it is NULL
	such as errors before start of XML document.

	* raptor/raptor_libxml.c:
	Check for xmlSAXHandler externalSubset field, not present in old
	libxml v1.  Whether raptor works after this is unlikely and untested.

	* raptor/configure.in: Add old libxml V1 warning and suggestion

	* raptor/acconfig.h: Added RAPTOR_LIBXML_XMLSAXHANDLER_EXTERNALSUBSET

	* raptor/configure.in: Check for xmlSAXHandler externalSubset field

2003-02-10  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/raptor_locator.c (raptor_print_locator):
	Only print non-negative line numbers

	* raptor/raptor_parse.c (raptor_xml_parse_chunk_):
	For libxml, handle first chunk being
	empty, when the XML parser context, xc, is attempted to be initialiased

	* raptor/tests/Makefile.am:
	Added ex-46 (Empty file should give 0 triples)

	* raptor/tests/ex-46.out, raptor/tests/ex-46.rdf:
	Empty file should give 0 triples

2003-01-31  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* Makefile.am: utils now has buildable stuff

	* librdf/rdf_hash_bdb.c:
	Make BDB open failures return librdf_error even when debug set.

	* configure.in, Makefile.am: scripts now utils

	* utils/Makefile.am, utils/db_upgrade.c, utils/update-api-0912.pl:
	scripts now utils, added redland-db-upgrade util

	* librdf/redland.spec.in: Added redland-db-upgrade

2003-01-30  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* configure.in: don't destroy passed in LDFLAGS value

2003-01-29  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* tcl/Makefile.am:
	Fix so it doesn't compile over the input file.  Bad $? / $< again

	* java/org/librdf/redland/Model.java:
	(add/remove) Added contexts version of methods
	(context_remove): Added, remove all statements with given context.

	* java/Makefile.am: $< wrong for build jars, add all files

	* python/RDF.py:
	(Model) Added contexts methods, as optional args to add/remove
	statements methods.
	New method: context_remove_statements

	* perl/lib/RDF/Redland/Model.pm (add_statement, add_statements,
	remove_statement): Add optional context arg.
	(remove_context_statements): Added.

	* configure.in:
	Make xml parser prefer libxml, to match what raptor does.  Expat will
	be added if required/available anyway, such as for repat.
	If expat is needed for raptor, ensure it uses system expat first if
	available rather than always use source version.

	* php/php.ini: added safe_mode = Off, says Ron

	* autogen.sh: require swig 1.3.14 min

	* docs/api.sgml.in, docs/fix-python-xhtml,
	docs/fix-pod-xhtml: dates

	* configure.in, Makefile.am:
	Renamed aux dir to scripts

	* scripts/Makefile.am, scripts/update-api-0912.pl: Moved from aux dir

2003-01-27  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* tcl/Makefile.am, ruby/Makefile.am, raptor/Makefile.am,
	python/Makefile.am, perl/Makefile.am, java/Makefile.am,
	docs/Makefile.am, php/Makefile.am:
	Remove $? and replace with $< or full dependencies

2003-01-23  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_register):
	Return correct failure sense.

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_terminate):
	Protect from empty arrays

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_register):
	Don't leak full_name memory when
	librdf_new_hash fails.

	* librdf/rdf_storage_hashes.c (librdf_storage_hashes_init_common):
	On failure to create lower level
	hashes, don't free stuff here, librdf_storage_hashes_terminate will
	do it, eventually.

2003-01-21  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/ntriples_parse.c:
	Fix macro IS_ASCII_DIGIT which refused to allow '0'.

2003-01-13  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* raptor/rdfdump.1: rdfdump.1 renamed to rapper.1

	* librdf/redland.spec.in:
	Updated for redland 0.9.12 (requires installed raptor to build)

	* raptor/raptor.spec.in, raptor/tests/Makefile.am,
	raptor/Makefile.am, raptor/INSTALL.html: rdfdump now rapper

	* raptor/rapper.1: rdfdump.1 renamed to rapper.1

2003-01-07  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* librdf/rdf_parser_sirpac.c: No longer needed

2003-01-03  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* examples/Makefile.am: tweak clean

	* librdf/rdf_model.c:
	(librdf_model_get_source,librdf_model_get_target,librdf_model_get_arc):
	Return a new copy of the shared node, restoring the declared API
	behaviour.

	* Makefile.am: Added RELEASE.html RELEASE

	* librdf/rdf_parser_repat.c
	(librdf_parser_repat_parse_file_into_model): Tidy up internal stream
	after add_statements no longer frees it.

	* librdf/rdf_parser_raptor.c
	(librdf_parser_raptor_parse_file_into_model): Tidy up internal stream
	after add_statements no longer frees it.

2003-01-02  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* examples/example4.c: Revert temporary context dbd-store ifdef

	* librdf/rdf_node.c (librdf_new_node_from_typed_literal):
	Docucomment fix

2003-01-01  Dave Beckett  <Dave.Beckett@bristol.ac.uk>

	* docs/api.sgml.in: Added general query, serializer doc parts

	* aux/update-api-0912.pl: Add model add_statement changes checks

	* aux/update-api-0912.pl: Docs

	* aux/update-api-0912.pl: Try to catch stream/iterator api changes too