Bug #4134: XQuery: default collation and starts-with causes fatal error cannot be cast to net.sf.saxon.lib.SubstringMatcher - Saxon - Saxonica Developer Community

Actions

Send by e-mail Copy link

Bug #4134

closed

XQuery: default collation and starts-with causes fatal error cannot be cast to net.sf.saxon.lib.SubstringMatcher

Added by Boris Lehečka about 5 years ago. Updated about 5 years ago.

Status:

Closed

Priority:

Normal

Assignee:

Michael Kay

Category:

Internals

Sprint/Milestone:

Start date:

2019-02-13

Due date:

% Done:

100%

Estimated time:

Legacy ID:

Applies to branch:

9.9, trunk

Fix Committed on Branch:

9.9, trunk

Fixed in Maintenance Release:

9.9.1.2

Platforms:

Description

When default collation is declared, using starts-with function with two parameters, i.e. starts-with($arg1, $arg2), causes the error net.sf.saxon.expr.sort.SimpleCollation cannot be cast to net.sf.saxon.lib.SubstringMatcher.

There are two solutions:

removing default collation declaration
adding collation argument to starts-with function, for example starts-with($arg1, $arg2, "http://saxon.sf.net/collation?lang=cs-CZ")

I would expect that default collation is used for comparison within starts-with function without "redefining" it.

Am I wrong?

Example file with commented code is attached.

Files

collation-error.xquery (414 Bytes) collation-error.xquery

default collation and starts-with error example

Boris Lehečka, 2019-02-13 11:14

Actions

Copy link

Updated by Michael Kay about 5 years ago

Thanks for reporting it. There's logic for handling substring matching using such collations when explicitly specified in the argument to substring(), but the logic is missing for the case where it's the default collation.

By the way, I would suggest that you switch to using the standardized collation URIs such as

http://www.w3.org/2013/collation/UCA?lang=cs-CZ;fallback=yes

in place of the Saxon-specific collation URIs.

I have to say that I have found the effect of using functions like starts-with() and contains() with collations that ignore some characters like spaces and hyphens pretty hard to comprehend.

Actions

Copy link