Bug #2715: XPath parsing edge cases - Saxon - Saxonica Developer Community

Actions

Send by e-mail Copy link

Bug #2715

closed

XPath parsing edge cases

Added by Michael Kay over 8 years ago. Updated over 7 years ago.

Status:

Closed

Priority:

Low

Assignee:

Category:

Sprint/Milestone:

9.7.0.5

Start date:

2016-04-19

Due date:

% Done:

100%

Estimated time:

Legacy ID:

Applies to branch:

9.7, trunk

Fix Committed on Branch:

9.7, trunk

Fixed in Maintenance Release:

9.7.0.5

Platforms:

Description

See https://lists.w3.org/Archives/Public/public-xsl-query/2016Mar/0037.html

The WG believes that the XPath grammar allows expressions such as

12!(12 div.)

 12 div-3

 3!(12 div-.)

which Saxon cannot parse. The key point is that Saxon is reading the "." or "-" as a continuation of the token starting "div", which it should not do, because the tokenization rule is to accept the longest token consistent with the grammar, and in these contexts, an NCName containing a "." or "-" is not consistent with the grammar.

Actions

Copy link

Updated by Michael Kay over 8 years ago

Status changed from New to Resolved
Applies to branch 9.7, 9.8 added
Fix Committed on Branch 9.7, 9.8 added

I've tweaked the tokenizer to handle these cases, though the solution might not be completely general. Essentially the fix is as follows:

if we encounter "." or "-" when reading (what appears to be) a name, then we check whether the following conditions are true:

(a) the content of the token so far is a valid operator/keyword

(b) the preceding token is not classified as an operator symbol (token number > LAST_OPERATOR) and is not "?" or "*:"

if all these conditions are true then we stop the tokenizing at the "." or "-" and return the operator token.

Condition (c) was found to be necessary because when we are parsing a FLWOR expression, keywords such as "in" and "return" are returned as NAME tokens rather than as specific operator tokens.

I'm committing a patch on the 9.7 and 9.8 branches.

Actions

Copy link