Daniel Veillard | 01791d5 | 1998-07-24 19:24:09 +0000 | [diff] [blame^] | 1 | |
| 2 | TODO for the XML parser: |
| 3 | |
| 4 | - Support for UTF-8 encoding |
| 5 | - progressive parsing. Currently the parser uses a single |
| 6 | string containing the full document. The good point is |
| 7 | that there is no context associated with the parser, the |
| 8 | full state is in the stack. The bad point is that such a |
| 9 | recursive design is hard to make progressive ... |
| 10 | - Better error handling, use a dedicated, overridable error |
| 11 | handling function. |
| 12 | - Keep track of line numbers for better error reporting. |
| 13 | - DOM support, instead of using a proprietary in memory |
| 14 | format for the document representation, the parser should |
| 15 | call a DOM API to actually build the resulting document. |
| 16 | Then the parser becomes independent of the in-memory |
| 17 | representation of the document. Even better using RPC's |
| 18 | the parser can actually build the document in another |
| 19 | program. |
| 20 | - finish the support for Entities. |
| 21 | - Support for Comments (bad, should be in ASAP, they are parsed |
| 22 | but not stored). |
| 23 | - Support for PI. |
| 24 | - Support for CDATA. |
| 25 | |
| 26 | Done: |
| 27 | - C++ support : John Ehresman <jehresma@dsg.harvard.edu> |
| 28 | - Updated code to follow more recent specs, added compatibility flag |
| 29 | |
| 30 | $Id$ |