TODO revision 38ae7e4efe803ea78b6499cd05a394db32623e41
5066N/A124907 HTML parse buffer problem when parsing larse in-memory docs
5066N/A124110 DTD validation && wrong namespace
5066N/A123564 xmllint --html --format
5066N/A
5066N/A TODO for the XML parser and stuff:
5066N/A ==================================
5066N/A
5066N/A $Id: TODO,v 1.44 2005/01/07 13:56:19 veillard Exp $
5066N/A
5066N/A this tend to be outdated :-\ ...
5066N/A
5066N/ADOCS:
5066N/A=====
5066N/A
5066N/A- use case of using XInclude to load for example a description.
5066N/A order document + product base -(XSLT)-> quote with XIncludes
5066N/A |
5066N/A HTML output with description of parts <---(XSLT)--
5066N/A
5066N/ATODO:
5066N/A=====
5066N/A- XInclude at the SAX level (libSRVG)
5066N/A- fix the C code prototype to bring back doc/libxml-undocumented.txt
5066N/A to a reasonable level
5066N/A- Computation of base when HTTP redirect occurs, might affect HTTP
5488N/A interfaces.
5066N/A- Computation of base in XInclude. Relativization of URIs.
5066N/A- listing all attributes in a node.
5066N/A- Better checking of external parsed entities TAG 1234
5066N/A- Go through erratas and do the cleanup.
5066N/A http://www.w3.org/XML/xml-19980210-errata ... started ...
5066N/A- jamesh suggestion: SAX like functions to save a document ie. call a
5066N/A function to open a new element with given attributes, write character
5066N/A data, close last element, etc
5066N/A + inversted SAX, initial patch in April 2002 archives.
5066N/A- htmlParseDoc has parameter encoding which is not used.
5066N/A Function htmlCreateDocParserCtxt ignore it.
5066N/A- fix realloc() usage.
5492N/A- Stricten the UTF8 conformance (Martin Duerst):
5066N/A http://www.w3.org/2001/06/utf-8-test/.
5066N/A The bad files are in http://www.w3.org/2001/06/utf-8-wrong/.
5066N/A- xml:id normalized value
5066N/A
5066N/ATODO:
5066N/A=====
5066N/A
5066N/A- move all string manipulation functions (xmlStrdup, xmlStrlen, etc.) to
5066N/A global.c. Bjorn noted that the following files depends on parser.o solely
5066N/A because of these string functions: entities.o, global.o, hash.o, tree.o,
5066N/A xmlIO.o, and xpath.o.
5066N/A
5066N/A- Optimization of tag strings allocation ?
5066N/A
5066N/A- maintain coherency of namespace when doing cut'n paste operations
5066N/A => the functions are coded, but need testing
5066N/A
5066N/A- function to rebuild the ID table
5066N/A- functions to rebuild the DTD hash tables (after DTD changes).
5066N/A
5066N/A
5066N/AEXTENSIONS:
5066N/A===========
5066N/A
5066N/A- Tools to produce man pages from the SGML docs.
5066N/A
5066N/A- Add Xpointer recognition/API
5066N/A
5066N/A- Add Xlink recognition/API
5066N/A => started adding an xlink.[ch] with a unified API for XML and HTML.
5066N/A it's crap :-(
5066N/A
5066N/A- Implement XSchemas
5066N/A => Really need to be done <grin/>
5066N/A - datatype are complete, but structure support is very limited.
5066N/A
5066N/A- extend the shell with:
5066N/A - edit
5066N/A - load/save
5066N/A - mv (yum, yum, but it's harder because directories are ordered in
5066N/A our case, mvup and mvdown would be required)
5066N/A
5066N/A
5066N/ADone:
5066N/A=====
5066N/A
5066N/A- Add HTML validation using the XHTML DTD
5066N/A - problem: do we want to keep and maintain the code for handling
5066N/A DTD/System ID cache directly in libxml ?
5066N/A => not really done that way, but there are new APIs to check elements
5066N/A or attributes. Otherwise XHTML validation directly ...
5066N/A
5066N/A- XML Schemas datatypes except Base64 and BinHex
5066N/A
5066N/A- Relax NG validation
5066N/A
5066N/A- XmlTextReader streaming API + validation
5066N/A
5066N/A- Add a DTD cache prefilled with xhtml DTDs and entities and a program to
5066N/A manage them -> like the /usr/bin/install-catalog from SGML
5066N/A right place seems $datadir/xmldtds
5066N/A Maybe this is better left to user apps
5066N/A => use a catalog instead , and xhtml1-dtd package
5066N/A
5066N/A- Add output to XHTML
5066N/A => XML serializer automatically recognize the DTd and apply the specific
5066N/A rules.
5066N/A
5066N/A- Fix output of <tst val="x&#xA;y"/>
5066N/A
5066N/A- compliance to XML-Namespace checking, see section 6 of
5066N/A http://www.w3.org/TR/REC-xml-names/
5066N/A
5066N/A- Correct standalone checking/emitting (hard)
5066N/A 2.9 Standalone Document Declaration
5066N/A
5066N/A- Implement OASIS XML Catalog support
5066N/A http://www.oasis-open.org/committees/entity/
5066N/A
5066N/A- Get OASIS testsuite to a more friendly result, check all the results
5066N/A once stable. the check-xml-test-suite.py script does this
5066N/A
5066N/A- Implement XSLT
5066N/A => libxslt
5066N/A
5066N/A- Finish XPath
5066N/A => attributes addressing troubles
5066N/A => defaulted attributes handling
5066N/A => namespace axis ?
5066N/A done as XSLT got debugged
5066N/A
5066N/A- bug reported by Michael Meallin on validation problems
5066N/A => Actually means I need to add support (and warn) for non-deterministic
5066N/A content model.
5066N/A- Handle undefined namespaces in entity contents better ... at least
5066N/A issue a warning
5066N/A- DOM needs
5066N/A int xmlPruneProp(xmlNodePtr node, xmlAtttrPtr attr);
5066N/A => done it's actually xmlRemoveProp xmlUnsetProp xmlUnsetNsProp
5066N/A
5066N/A- HTML: handling of Script and style data elements, need special code in
5066N/A the parser and saving functions (handling of < > " ' ...):
5066N/A http://www.w3.org/TR/html4/types.html#type-script
5066N/A Attributes are no problems since entities are accepted.
5066N/A- DOM needs
5066N/A xmlAttrPtr xmlNewDocProp(xmlDocPtr doc, const xmlChar *name, const xmlChar *value)
5066N/A- problem when parsing hrefs with & with the HTML parser (IRC ac)
5066N/A- If the internal encoding is not UTF8 saving to a given encoding doesn't
5066N/A work => fix to force UTF8 encoding ...
5066N/A done, added documentation too
5066N/A- Add an ASCII I/O encoder (asciiToUTF8 and UTF8Toascii)
5066N/A- Issue warning when using non-absolute namespaces URI.
5066N/A- the html parser should add <head> and <body> if they don't exist
5066N/A started, not finished.
5066N/A Done, the automatic closing is added and 3 testcases were inserted
5066N/A- Command to force the parser to stop parsing and ignore the rest of the file.
5066N/A xmlStopParser() should allow this, mostly untested
5066N/A- support for HTML empty attributes like <hr noshade>
5066N/A- plugged iconv() in for support of a large set of encodings.
5066N/A- xmlSwitchToEncoding() rewrite done
5066N/A- URI checkings (no fragments) rfc2396.txt
5066N/A- Added a clean mechanism for overload or added input methods:
5066N/A xmlRegisterInputCallbacks()
5066N/A- dynamically adapt the alloc entry point to use g_alloc()/g_free()
5066N/A if the programmer wants it:
5066N/A - use xmlMemSetup() to reset the routines used.
5066N/A- Check attribute normalization especially xmlGetProp()
5066N/A- Validity checking problems for NOTATIONS attributes
5066N/A- Validity checking problems for ENTITY ENTITIES attributes
5066N/A- Parsing of a well balanced chunk xmlParseBalancedChunkMemory()
5066N/A- URI module: validation, base, etc ... see uri.[ch]
5066N/A- turn tester into a generic program xmllint installed with libxml
5066N/A- extend validity checks to go through entities content instead of
5066N/A just labelling them PCDATA
5066N/A- Save Dtds using the children list instead of dumping the tables,
5066N/A order is preserved as well as comments and PIs
5066N/A- Wrote a notice of changes requires to go from 1.x to 2.x
5066N/A- make sure that all SAX callbacks are disabled if a WF error is detected
5066N/A- checking/handling of newline normalization
5066N/A http://localhost/www.xml.com/axml/target.html#sec-line-ends
5066N/A- correct checking of '&' '%' on entities content.
5066N/A- checking of PE/Nesting on entities declaration
5066N/A- checking/handling of xml:space
5066N/A - checking done.
5066N/A - handling done, not well tested
5066N/A- Language identification code, productions [33] to [38]
5066N/A => done, the check has been added and report WFness errors
5066N/A- Conditional sections in DTDs [61] to [65]
5066N/A => should this crap be really implemented ???
5066N/A => Yep OASIS testsuite uses them
5066N/A- Allow parsed entities defined in the internal subset to override
5066N/A the ones defined in the external subset (DtD customization).
5066N/A => This mean that the entity content should be computed only at
5066N/A use time, i.e. keep the orig string only at parse time and expand
5066N/A only when referenced from the external subset :-(
5066N/A Needed for complete use of most DTD from Eve Maler
5066N/A- Add regression tests for all WFC errors
5066N/A => did some in test/WFC
5066N/A => added OASIS testsuite routines
5066N/A http://xmlsoft.org/conf/result.html
5066N/A
5066N/A- I18N: http://wap.trondheim.com/vaer/index.phtml is not XML and accepted
5066N/A by the XML parser, UTF-8 should be checked when there is no "encoding"
5066N/A declared !
5066N/A- Support for UTF-8 and UTF-16 encoding
5066N/A => added some convertion routines provided by Martin Durst
5066N/A patched them, got fixes from @@@
5066N/A I plan to keep everything internally as UTF-8 (or ISO-Latin-X)
5066N/A this is slightly more costly but more compact, and recent processors
5066N/A efficiency is cache related. The key for good performances is keeping
5066N/A the data set small, so will I.
5066N/A => the new progressive reading routines call the detection code
5066N/A is enabled, tested the ISO->UTF-8 stuff
5066N/A- External entities loading:
5066N/A - allow override by client code
5066N/A - make sure it is alled for all external entities referenced
5066N/A Done, client code should use xmlSetExternalEntityLoader() to set
5066N/A the default loading routine. It will be called each time an external
5066N/A entity entity resolution is triggered.
5066N/A- maintain ID coherency when removing/changing attributes
5066N/A The function used to deallocate attributes now check for it being an
5066N/A ID and removes it from the table.
5066N/A- push mode parsing i.e. non-blocking state based parser
5066N/A done, both for XML and HTML parsers. Use xmlCreatePushParserCtxt()
5066N/A and xmlParseChunk() and html counterparts.
5066N/A The tester program now has a --push option to select that parser
5066N/A front-end. Douplicated tests to use both and check results are similar.
5066N/A
5066N/A- Most of XPath, still see some troubles and occasionnal memleaks.
5066N/A- an XML shell, allowing to traverse/manipulate an XML document with
5066N/A a shell like interface, and using XPath for the anming syntax
5066N/A - use of readline and history added when available
5066N/A - the shell interface has been cleanly separated and moved to debugXML.c
5066N/A- HTML parser, should be fairly stable now
5066N/A- API to search the lang of an attribute
5066N/A- Collect IDs at parsing and maintain a table.
5066N/A PBM: maintain the table coherency
5066N/A PBM: how to detect ID types in absence of DtD !
5066N/A- Use it for XPath ID support
5066N/A- Add validity checking
5066N/A Should be finished now !
5066N/A- Add regression tests with entity substitutions
5066N/A
5066N/A- External Parsed entities, either XML or external Subset [78] and [79]
5066N/A parsing the xmllang DtD now works, so it should be sufficient for
5066N/A most cases !
5066N/A
5066N/A- progressive reading. The entity support is a first step toward
5066N/A asbtraction of an input stream. A large part of the context is still
5066N/A located on the stack, moving to a state machine and putting everyting
5066N/A in the parsing context should provide an adequate solution.
5066N/A => Rather than progressive parsing, give more power to the SAX-like
5066N/A interface. Currently the DOM-like representation is built but
5066N/A => it should be possible to define that only as a set of SAX callbacks
5066N/A and remove the tree creation from the parser code.
5066N/A DONE
5066N/A
5066N/A- DOM support, instead of using a proprietary in memory
5066N/A format for the document representation, the parser should
5066N/A call a DOM API to actually build the resulting document.
5066N/A Then the parser becomes independent of the in-memory
5066N/A representation of the document. Even better using RPC's
5066N/A the parser can actually build the document in another
5066N/A program.
5066N/A => Work started, now the internal representation is by default
5066N/A very near a direct DOM implementation. The DOM glue is implemented
5066N/A as a separate module. See the GNOME gdome module.
5066N/A
5066N/A- C++ support : John Ehresman <jehresma@dsg.harvard.edu>
5066N/A- Updated code to follow more recent specs, added compatibility flag
5066N/A- Better error handling, use a dedicated, overridable error
5066N/A handling function.
5066N/A- Support for CDATA.
5066N/A- Keep track of line numbers for better error reporting.
5066N/A- Support for PI (SAX one).
5066N/A- Support for Comments (bad, should be in ASAP, they are parsed
5066N/A but not stored), should be configurable.
5066N/A- Improve the support of entities on save (+SAX).
5066N/A
5066N/A