|
Difference between
current version
and
version 1:
| Line 21 was replaced by line 21 |
| - * Developed a new algorithm for importing XML Schema files into the__internal schema trees in SCIA, based on Apache__XML Schema APIs and DOM 3 APIs, in order to acquire constraint__information and more useful type information for matching and mapping, and__also to support more__complicated XML Schema files, including EML. |
| + * Developed a new algorithm for importing XML Schema files into the internal schema trees in SCIA, based on Apache XML Schema APIs and DOM 3 APIs, in order to acquire constraint information and more useful type information for matching and mapping, and also to support more complicated XML Schema files, including EML. |
| Lines 24-25 were replaced by lines 24-25 |
| - * Solve the memory issues with large size schemas which possibly contain recursive references. For example, currently, loading more than 1/3 of EML would cause OutOfMemoryError or StackOverFlowError, though 128MB is allocated for heap and 512MB for Java Virtual Machine (default only__64MB), 200MB for stack (default only 400k). The StackOverFlowError is caused by the recursive references in EML, e.g., A DataSet element has a subelement Methods, while a Methods element has a subelement DataSource of type DataSetType. __This issue reinforces that SCIA should be able to represent an input schema __ as a main tree and multiple sub trees, instead of a single schema tree. |
| - * Look for new free XQuery engine. There are two main reasons for this. __Firstly, the Quilt engine, Kweelt, which we are using now, uses a very old __version of xerces package that is not compatible with the new version that covers XML Schema __APIs. In addition, SCIA'd better generate view in standard XQuery, in order to easily work with other applications together |
| + * Solve the memory issues with large size schemas which possibly contain recursive references. For example, currently, loading more than 1/3 of EML would cause OutOfMemoryError or StackOverFlowError, though 128MB is allocated for heap and 512MB for Java Virtual Machine (default only 64MB), 200MB for stack (default only 400k). The StackOverFlowError is caused by the recursive references in EML, e.g., A DataSet element has a subelement Methods, while a Methods element has a subelement DataSource of type DataSetType. This issue reinforces that SCIA should be able to represent an input schema as a main tree and multiple sub trees, instead of a single schema tree. |
| + * Look for new free XQuery engine. There are two main reasons for this. Firstly, the Quilt engine, Kweelt, which we are using now, uses a very old version of xerces package that is not compatible with the new version that covers XML Schema APIs. In addition, SCIA'd better generate view in standard XQuery, in order to easily work with other applications together |
| Removed lines 29-30 |
| - Significant issues |
| - * |
| Removed line 32 |
| - |
Back to Status Report 4 April 2005,
or to the Page History.
|
