W3C home > Mailing lists > Public > public-expath@w3.org > November 2013

Re: [expath] File Module

From: John Lumley <john@saxonica.com>
Date: Fri, 29 Nov 2013 17:29:38 +0000
Message-Id: <D5F4286B-2774-42D5-B6BD-3A7C5F35AD46@saxonica.com>
Cc: EXPath <expath@googlegroups.com>, EXPath <public-expath@w3.org>
To: Christian Grün <christian.gruen@gmail.com>
Having gone through the exercises of implementing dir-name() and base-name() to meet the current spec. and trying to make any rationality about them, I'm all for changing to name() and parent(). Let's face it: when you're doing file system manipulation in XSLT, you're going to be taking a lot of care, and unlike cavalier methods in the *sh/perl etc worlds, you really want simple, easily comprehendable primitives to work with and then you'll do lots of checking.

(It strikes me that a good example test would be to write a simple make/Ant type of XSLT processor, entirely in XSLT, using the file module and perhaps transform invokers, such as saxon:transform())

John

Sent from my iPad

On 29 Nov 2013, at 17:10, Christian Grün <christian.gruen@gmail.com> wrote:

>> I'm happy to go with these incompatible changes, but I really would like to see real focus now on getting specs frozen at 1.0 level where implementors and users can have confidence in their stability.
> 
> Michael, thanks for the immediate response.
> 
> I completely agree that there cannot be more changes to existing
> features in this specification, so if someone wants to give feedback,
> please give it now.
> 
> Christian
> 
> 
> 
>>> On 29 Nov 2013, at 15:55, Christian Grün <christian.gruen@gmail.com> wrote:
>>> 
>>> Dear all,
>>> 
>>> after more than two years, the EXPath File Module is about to be
>>> finished. I have just put the latest versions of the spec online
>>> (including the diff):
>>> 
>>> http://files.basex.org/publications/expath.org/spec/file.html
>>> http://files.basex.org/publications/expath.org/spec/file-diff.html
>>> 
>>> Thanks to John Lumley, and other people on our own mailing list, who
>>> have given us helpful feedback. Our plan was to avoid major changes to
>>> the spec, due to its advanced state, but I think the feedback should
>>> not be simply discarded:
>>> 
>>> 1) The vintage Linux commands "dirname" and "basename" have also been
>>> adopted to the spec, but it has turned out it’s not only the wording
>>> that’s unfortunate (in most cases, dir-name() does not return a name,
>>> but rather the path to a directory). We would thus like to drop these
>>> two functions and instead introduce the following alternatives:
>>> 
>>> SIGNATURE
>>> file:parent($path as xs:string) as xs:string?
>>> DESCRIPTION
>>> Returns the parent directory of a file or directory. – An empty
>>> sequence is returned if the path points to a root directory, or if it
>>> contains no directory separators. – This function is ·deterministic·
>>> (no path existence check is made).
>>> 
>>> SIGNATURE
>>> file:name($path as xs:string) as xs:string
>>> DESCRIPTION
>>> Returns the name of a file or directory. – An empty string is
>>> returned if the path points to the root directory, or if it contains
>>> no directory separators. – This function is ·deterministic· (no path
>>> existence check is made).
>>> 
>>> 2) Florent suggested to replace the existing error codes with his new
>>> EXPath naming conventions:
>>> 
>>> http://www.w3.org/community/expath/wiki/Naming
>>> 
>>> I first resisted, as the File Module is already used in the wild, but
>>> if we decide to replace the functions mentioned above, it would be
>>> just consequent to also make the codes more accessible.
>>> 
>>> If everyone agrees to the two changes, I will once more update the
>>> spec and hope that we can release it as official EXPath Proposed
>>> Module.
>>> 
>>> Thanks for your attention,
>>> Christian
>>> 
>>> --
>>> You received this message because you are subscribed to the Google Groups "EXPath" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an email to expath+unsubscribe@googlegroups.com.
>>> To post to this group, send email to expath@googlegroups.com.
>>> Visit this group at http://groups.google.com/group/expath.
>>> For more options, visit https://groups.google.com/groups/opt_out.
>> 
>> --
>> You received this message because you are subscribed to the Google Groups "EXPath" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to expath+unsubscribe@googlegroups.com.
>> To post to this group, send email to expath@googlegroups.com.
>> Visit this group at http://groups.google.com/group/expath.
>> For more options, visit https://groups.google.com/groups/opt_out.
> 
Received on Friday, 29 November 2013 17:30:01 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:47:37 UTC