W3C home > Mailing lists > Public > public-qt-comments@w3.org > August 2003

Comments on http://www.w3.org/TR/xquery/#N401458

From: Ian Davis <ijdavis@softbase.math.uwaterloo.ca>
Date: Tue, 12 Aug 2003 11:38:26 -0400
Message-Id: <200308121538.LAA14051@softbase.math.uwaterloo.ca>
To: public-qt-comments@w3.org
Cc: ijdavis@softbase.math.uwaterloo.ca


I don't think XQuery should allow the EscapeQuot and EscapeApos productions.
This invites users to embed XML text in XQuery which does not conform to XML.
A naive method of parsing constructors is to have an XML parser parse the
material outside the {..}, and to produce as a string the concatenation of
all such fragments so parsed with the result of the values computed within
the {...}.  If one allows the EscapeQuot and EscapeApos escaping mechanism
then the XML parser has to be conscious that these escaping mechanisms exist,
but only in very explicit contexts.  Such quotes and double quotes have to be
translated in to the correct XML escape sequence of &apos; and &quot;.  Why
introduce yet one more escaping mechanism into a standard needlessly.

If the EscapeQuot and EscapeApos productions are to remain valid then
production [106] AttributeValueContent needs to be split into two
productions.  This is because [28] QuoteAttributeContentChar allows any
char but ["{}<&] while [29] AposAttributeContentChar allows any char but
['{}<&].  Since the [106] production allows either production without
qualification only the characters [{}<&] are excluded by [28] | [29].
This clearly is not the intent.



[128]  Missing (..) defining scope of *

[135]  Missing (..) before ? likewise [36]

Hope my comments are viewed as constructive.

Regards, Ian Davis.
Received on Tuesday, 12 August 2003 11:50:43 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:56:49 UTC