I am pleased to announce that the revamped text table support is now complete and committed to CVS hsqldb-dev HEAD. This will be tested and refined further in the next couple of weeks.
Text tables now support BINARY and OTHER columns and can have primary keys just like other types of table. As a result of the changes, the contents of the csv file (the source for the text table) are read and indexed at startup time, instead of on first use. As we approach the remaining tasks leading to the release of the next version, more developer involvement is needed. Some of the tasks involve rewriting existing tokeniser and parser code and can be done without deep involvement in the engine code: 1. Parser for DDL statements: Currently there is a lot of hand-written code that uses conditions, switches and loops for parsing DDL statements. This code can be rewritten more systematically to produce a binary result in a standard format, and thus reduce the overall code size. 2. Parser for lists: A simple parser to translate property lists into property pairs is needed. A basic example of this already exists in HsqlProperties.java to translate command line arguments into properties. An enhanced version is required for use with property lists for text tables. 3. The tokeniser for text tables needs to be rewritten to work directly on a byte stream, rather than using an intermediate String object to build rows of text tables (see TextDatabaseRowInput.java and TextCache.java). Volunteers please get in touch. Fred Toussi ------------------------------------------------------- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power & Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en _______________________________________________ hsqldb-developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/hsqldb-developers
