W3C home > Mailing lists > Public > www-voice@w3.org > October to December 2014

Re: Discrepancy in normative XPath references (XPath 1 and 2)?

From: Elvis Stansvik <elvstone@gmail.com>
Date: Tue, 21 Oct 2014 00:31:55 +0200
Message-ID: <CAHms=ea3oo4qH5aAMjs4agf_4hn-rtoY67MuRwYJqeib6+qAnw@mail.gmail.com>
To: Jim Barnett <jim.barnett@genesys.com>
Cc: "www-voice@w3.org" <www-voice@w3.org>
2014-10-21 0:24 GMT+02:00 Jim Barnett <jim.barnett@genesys.com>:

>  Elvis,
>   There are three implementations of the XPath data model that I am aware
> of, though none of them have yet submitted implementation reports. From
> comments from the developers of those implementations, I gather that
> finding a suitable XPath implementation requires some thought.
>

Ah, right. I completely misread that e-mail as saying there weren't any, my
bad. And yes I can imagine it requires some thought. But no way around that
really, it's not like In() can be dropped from the spec.

Elvis

Jim
>
>
>
> On Oct 20, 2014, at 5:15 PM, Elvis Stansvik <elvstone@gmail.com> wrote:
>
>   (Sorry. Forgot to send my answer to list)
>
> 2014-10-20 23:40 GMT+02:00 Jim Barnett <1jhbarnett@gmail.com>:
>
>> Elvis,
>>   I will have to check with other group members before giving a
>> definitive reply, but we do intend to specify XPath 1.0, while we like
>> 2.0's definition of effective Boolean value to describe what to do when you
>> pull a value out of XPath into SCXML in a Boolean context.  I will have to
>> make sure we think that this is acceptably clear spec language.
>>
>
>  Ah. Then I see the reason. Maybe it could be made more explicit what
> role the different references has. But good that you'll bring it up.
>
>
>> Note that you can build your data model on XPath 2 if you want. You just
>> have to give it a different name than the one we define. Data models are
>> intended to be pluggable and extensible, and you aren't limited to the ones
>> that we define.
>>
>
>  Yep. Upon reading further I realized that, and also saw in some earlier
> mail that there are no actual implementations of the XPath data model yet.
> Is that still the case?
>
>  One thing about implementing that is a little problematic is that not
> all XPath implementations allow you to define custom functions, which is
> needed to add the In(), and I think that some which do allow it will only
> allow functions with some "prefix:" to be added. Not sure though. This is
> of course assuming that you use a pre-built XPath library, and is no
> problem if you implement it yourself as part of implementing SCXML. I see
> no way around that though, just one thing that limits the selection of
> XPath libraries to pick from when implementing.
>
>  Thanks for your answers.
>
>  Elvis
>
>
>>
>> Jim Barnett
>> Genesys
>>
>> > On Oct 18, 2014, at 5:55 AM, Elvis Stansvik <elvstone@gmail.com> wrote:
>> >
>> > Hi folks,
>> >
>> > I only just heard about SCXML.. I got interested and started reading
>> the spec.
>>  >
>> > In "B.3 The XPath Data Model", the section starts out by referencing
>> XPath 1.0:
>> >
>> >    "Implementations that support this data model must support [XPath
>> 1.0]."
>> >
>> > But then in "B.3.2 Conditional Expressions" goes on to require XPath
>> 2.0:
>> >
>> >     "The SCXML Processor must accept any XPath expression as
>> >       a conditional expression and must convert it into its effective
>> >       boolean value as described in section 2.4.3 of the [XPath 2.0]
>> >       specification."
>> >
>> > Could someone clarify? Should the first reference be to XPath 2.0 or is
>> the discrepancy intended?
>> >
>> > Cheers,
>> > Elvis Stansvik
>>
>
>
Received on Monday, 20 October 2014 22:32:22 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:07:45 UTC