- Email: firstname.lastname@example.org
- Registered on: 2014-03-13
- Last connection: 2019-01-17
In fact (as Mike pointed out) since 9.9 only provides XPath 3.1, it is not worth keeping the links to older specs. Updates committed in the relevant generating stylesheets on the 9.9 branch: xsl/functions-data.xsl, functions-functions.xs... Bug #4080 (In Progress): Incorrect links to XPath 2.0 spec for fn functionsIn the documentation Function Library, links are provided to the W3C specifications. But the links to XPath 2.0 Functions and Operators actually just direct to the most recent spec (http://www.w3.org/TR/xpath-functions/).
Perhaps we s...
Configuration file schemas in 9.9 samples have now also been updated (to add global/@thresholdForHotspotByteCode). Bug #4016: Setting the threshold for generating bytecodeAt the level of the EnterpriseConfiguration, it is possible to set the threshold for hotspot bytecode generation by passing an integer (or a string in the form of an integer) to setConfigurationProperty with the Feature key GENERATE_BYTE...
Still seems to be some issue. Try the more complicated stylesheet taskBug2.xsl (again `-it:ex1 -xsl:taskBug2.xsl`). Without setting the threshold, it works eventually, but takes a looooong time. Bug #2321 (In Progress): Function inliningWe should do some experiments with function inlining. (a) the limit on expression size for inlining is fixed at 15, which seems rather low. BaseX defaults the limit to 100, and makes it configurable. (b) we don't inline a function if it ...
See attached stylesheet taskBug.xsl. Transform with initia...
Thanks for pointing this out!
Links corrected to direct to the relevant version of the Feature Matrix for 9.9, 9.8, and 9.7 documentation, from about/packages and about (i.e. about/index.html). Fixes committed, and online documentatio... Bug #4085 (Closed): Current documentation (i.e. of Saxon 9.9) links to feature matrix of Saxon 9.6The current documentation http://saxonica.com/html/documentation/about/packages.html which in general is about Saxon 9.9 has a link "feature matrix" linking to http://www.saxonica.com/products/feature-matrix-9-6.xml which then in the bre...
Patch committed for the added `FormatInteger.copy` method to use `System.Function.makeFunction()` rather than `new FormatInteger();` so that the details are actually copied.
The problem was causing a number of QT3 converted tests to f... Feature #4039: Optimizing format-integer() when the picture/language are known staticallyUnlike other functions such as format-number() and format-date(), the format-integer() function does not optimize for the case where the picture and language are known statically.
This has been remedied. (Moreover, the mechanism now t...
Bug #4080 (In Progress): Incorrect links to XPath 2.0 spec for fn functionsIn the documentation Function Library, links are provided to the W3C specifications. But the links to XPath 2.0 Functions and Operators actually just direct to the most recent spec (http://www.w3.org/TR/xpath-functions/).
Perhaps we s...
In fact, we shouldn't need code changes after all!
The `ixsl:schedule-action/@wait` attribute allows an *expression*, which must evaluate to an integer, but does not need to be an integer literal. So you should be able to supply a var... Feature #4065: Saxon-JS: Allow AVTs in ixsl:schedule-action/waitI’m currently experimenting with visualizing a transformation using Saxon-JS. To this end, I injected code into an existing transformation that calls, by means of `ixsl:schedule-action`, a template that will attach classes to preformatte...
Patch committed on 9.9 and 9.8 branches.
For the `maxCapture` limit in `REMatcher.replace()`, use `program.maxParens` (from the REMatcher's associated REProgram), which gives the count of parenthesized subexpressions in the pattern; r... Bug #4076 (Resolved): fn:replace() with unmatched double digit capturing groupThe wrong result is returned from fn:replace() when the supplied replacement string contains reference `$N` with N > 9, but the Nth parenthesized subexpression in the pattern is not matched.
This is apparently a longstanding bug - I h...
Also available in: Atom