W3C home > Mailing lists > Public > www-forms@w3.org > July 2001

Re: positional predicates in canonical binding-expressions?

From: A Rafael D Teixeira <rafaelteixeirabr@hotmail.com>
Date: Wed, 11 Jul 2001 11:35:36 -0300
To: mimasa@w3.org, joern@webman.de
Cc: www-forms@w3.org
Message-ID: <LAW2-F255gZfUEJM39T0000f2dd@hotmail.com>
In my implementation Iīve decided to post the data as an XML string inside a 
textarea, so I get it the way I need it. In my thin (3 lines) ASP/PHP layer 
I put it through the XML parser to feed a DOM to the business objects that 
have to process it.

Doing a field-by-field post probably was the way you found to reuse your old 
page-processing code or your skills, but I can tell you XML IS THE WAY.

For example, some of my application objects donīt even bother to 
desserialize the XML: they work over the in-memory DOM and store the 
processed version in the database as a string.

----Original Message Follows----
From: "joern turner" <joern@webman.de> (by way of "Masayasu Ishikawa" 
To: www-forms@w3.org
Subject: positional predicates in canonical binding-expressions?
Date: Wed, 11 Jul 2001 04:57:12 -0400 (EDT)

in my effort to implement a web-based implementation i've come over some
pratical problem with the current definition of canonical binding
expressions in conjunction with repeat-elements.

i'm not an expert in xpath, but from my current understanding i need a
'positional' predicate ( like e.g. [1] ) to completely handle
(please correct me, if i've missed something!...)

although there's no problem while processing the form, it occurs when
the collected instance-data have to be submitted via http as key/value
pairs. as http makes no guarantee about the order of posted parameters,
each single instance-value must be referenced uniquely by a canonical
binding-expression, so the instance can be 're-assembled' on the server.

the repeat example clarifies the  problem:
(i hope this gets not too scrambled)


to submit these instance-data i see no other way than specifying a
parameter with a positional predicate in the ref-attribute like this:


then the above data could be transferred in the following form (but
arriving the server in no specific order)


the same problem applies for simple lists.

please excuse the lengthy mail and tell me about your thoughts in that area.

Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.
Received on Wednesday, 11 July 2001 10:36:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:36:04 UTC