This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 2469 - [XSLT] Handling of Serialization Errors
Summary: [XSLT] Handling of Serialization Errors
Status: CLOSED FIXED
Alias: None
Product: XPath / XQuery / XSLT
Classification: Unclassified
Component: XSLT 2.0 (show other bugs)
Version: Candidate Recommendation
Hardware: PC Windows XP
: P2 normal
Target Milestone: ---
Assignee: Michael Kay
QA Contact: Mailing list for public feedback on specs from XSL and XML Query WGs
URL:
Whiteboard:
Keywords:
: 2540 (view as bug list)
Depends on:
Blocks:
 
Reported: 2005-11-04 17:19 UTC by Michael Kay
Modified: 2006-03-10 21:12 UTC (History)
1 user (show)

See Also:


Attachments

Description Michael Kay 2005-11-04 17:19:01 UTC
We have a statement in section 2.9 which seems surprising in the light of
recent decisions:

As with other aspects of serialization, the handling of serialization errors
is implementation-defined: see 20 Serialization.

Similarly, section 20 starts out:

Stylesheet authors can use the xsl:output declaration to specify how they
wish result trees to be serialized. If a processor serializes a final result
tree, it *should* do so as specified by these elements; however, it is not
*required* to do so.

(Concerning handling of serialization errors, section 20 passes the buck
back to section 2.9)

We may want to review whether we intended to be a bit more
prescriptive.

Michael Kay
previously raised as
http://lists.w3.org/Archives/Member/w3c-xsl-wg/2005Oct/0005.html
Comment 1 Colin Adams 2005-11-21 06:01:27 UTC
*** Bug 2540 has been marked as a duplicate of this bug. ***
Comment 2 Michael Kay 2005-12-13 11:41:41 UTC
The XSL WG discussed this comment at its meeting on 8 Dec 2005 and agreed with
the general sentiment: the specification should be more prescriptive about the
handling of serialization in general, and serialization errors in particular, to
reflect the decisions that had been made in the past on these topics. The editor
was asked to propose specific textual changes.

Proposal:

1. In section 2.9, change

As with other aspects of serialization, the handling of serialization errors is
implementation-defined: see 20 Serialization.

to

If the processor performs serialization, then it MUST do so as specified in 20
Serialization, and in particular it MUST signal any serialization errors that occur.

(This change will automatically cause item 5 in appendix F to disappear) 

2. In section 20, change

Stylesheet authors can use the xsl:output declaration to specify how they
wish result trees to be serialized. If a processor serializes a final result
tree, it *should* do so as specified by these elements; however, it is not
*required* to do so.

to

Stylesheet authors can use xsl:output declarations to specify how they
wish result trees to be serialized. If a processor serializes a final result
tree, it *must* do so as specified by these declarations.

3. Delete the note at the end of section 20, which refers back to section 2.9
for handling of serialization errors. Replace it with the following text:

"If the processor performs serialization, then it must signal any serialization
errors that occur. These have the same effect as non-recoverable dynamic errors:
that is, the processor must signal the error and must not finish as if the
transformation had been successful."
Comment 3 Michael Kay 2006-02-02 18:57:17 UTC
Need to change 

"If the processor performs serialization, then it must signal any serialization
errors that occur. These have the same effect as non-recoverable dynamic errors:
that is, the processor must signal the error and must not finish as if the
transformation had been successful."

to

"If the processor performs serialization, then it must signal any
*non-recoverable* serialization
errors that occur. These have the same effect as non-recoverable dynamic errors:
that is, the processor must signal the error and must not finish as if the
transformation had been successful."

With this change, the proposal was agreed.
Comment 4 Michael Kay 2006-03-10 21:12:28 UTC
This change has now been applied to the spec and the issue is therefore closed.