gmane.lisp.scheme.ssax-sxml

Re: XML parsing bug?

Oleg Parashchenko — 2011-08-29T11:40:30

Hello Oleg,

thank you for the explanation.

On Fri, 26 Aug 2011 20:18:58 -0700 (PDT)
oleg-gU9taWJb54bYtjvyW6yDsg< at >public.gmane.org wrote:

...

As space elimination is often a good thing, I decided to implement a
conditional approach.

In a private copy of code, I changed

(let ((seed (ssax:reverse-collect-str-drop-ws seed))

to

(let ((seed ((if (tag-drops-xml-space? elem-gi)
  ssax:reverse-collect-str-drop-ws ssax:reverse-collect-str) seed))

where "tag-drops-xml-space?" is something like

(define tag-drops-xml-space?
  (let ((drop-list (map add-my-ns (list "chapter" "subchapter" ...))))
    (lambda (elem-gi) (let ((full-gi (if (pair? elem-gi)
                        (string->symbol (string-append
                                        (symbol->string (car elem-gi)) ":"
                                        (symbol->string (cdr elem-gi))))
                         elem-gi)))
        (memq full-gi drop-list)))))

...

Re: XML parsing bug?

Andy Wingo — 2011-08-27T11:55:17

Hi,

On Sat 27 Aug 2011 05:18, oleg-gU9taWJb54bYtjvyW6yDsg< at >public.gmane.org writes:


FWIW this seemed reasonable to me when I first started with SSAX, but it
surprised me later on in another context where I needed the whitespace.
By that time I had forgotten about this procedure and so spent some time
in figuring out what went wrong.

I would suggest that packagers of SSAX provide this whitespace-trimming
behavior as an option that is easy to enable, and furthermore that it be
off by default.  That's the only way to preserve all information in the
original document.

Regards,

Andy

Re: XML parsing bug?

2011-08-27T03:18:58

Hello!


SSAX is generally a framework for XML parsing. The SSAX library
includes useful instances of the framework, such as the function
ssax:xml->sxml, which you must be using. That function invokes the
function
ssax:reverse-collect-str-drop-ws

(see line 2771 of SSAX.scm), whose comments describe it
as follows:

;     ssax:reverse-collect-str-drop-ws LIST-OF-FRAGS -> LIST-OF-FRAGS
; given the list of fragments (some of which are text strings)
; reverse the list and concatenate adjacent text strings.
; We also drop "unsignificant" whitespace, that is, whitespace
; in front, behind and between elements. The whitespace that
; is included in character data is not affected.
; We use this procedure to "intelligently" drop "insignificant"
; whitespace in the parsed SXML. If the strict compliance with
; the XML Recommendation regarding the whitespace is desired, please
; use the ssax:reverse-collect-str procedure instead.

So, you my wish to re-define ssax:xml->sxml in your project to call
ssax:reverse-collect-

XML parsing bug?

Oleg Parashchenko — 2011-08-26T12:28:44

Hello,

using the "ssax-sxml-gambit-060529.tgz", parsing the XML document

<p><b>so,</b> <b>so</b> and <b>so</b></p>

I get

(*TOP* (p
  (b "so,")
  (b "so")
  " and "
  (b "so")))

The space between the first "b"-tags is lost. Right now I can't check
if the latest version from the sorceforge repository produces the same
error, but I'm afraid yes, why not. Can you reproduce the bug? If yes,
is there a chance that the bug would be fixed?

Re: how to unwrap a tag in pre-post-order

Oleg Parashchenko — 2011-08-18T11:11:27

Hello Peter,

On Wed, 17 Aug 2011 14:34:25 +0200
Peter Bex <Peter.Bex-qWit8jRvyhVmR6Xm/wNWPw< at >public.gmane.org> wrote:

...

Thanks a lot!

Re: how to unwrap a tag in pre-post-order

Peter Bex — 2011-08-17T12:34:25

Hi!

[snip]

I think what you're looking for is "pre-post-order-splice":

(pre-post-order-splice doc
  `((*default* . ,(lambda args args))
    (proxy     . ,(lambda (tag . kids) kids)))))
=>
(*TOP* (chapter (section) (section)))

See the notes near the end of lib/SXML-tree-trans.scm in the SSAX
distribution.

HTH,
Peter Bex

how to unwrap a tag in pre-post-order

Oleg Parashchenko — 2011-08-17T11:47:57

Hello,

I try to "unwrap" an XML tag: delete the tag, but retain the subtrees.
For example, the source document:

<chapter>
  <proxy><section></section></proxy>
  <proxy><section></section></proxy>
</chapter>

after unwrapping the tag "proxy" gives:

<chapter>
  <section/>
  <section/>
</chapter>

A first attempt with SXSLT:

(define doc '(*TOP* (chapter (proxy (section)) (proxy (section)))))
(define unwrapped (pre-post-order doc `(
  (*default* . ,(lambda args args))
  (proxy     . ,(lambda (tag . kids) kids))
  )))
(display unwrapped)(newline)

The result is:

(*TOP* (chapter ((section)) ((section))))

But I want:

(*TOP* (chapter (section) (section)))

Any trick to get the desired result?

The idea is to apply several "pre-post-order" in sequence, and mix them
with sxpath expressions. But right now the result of a transformation is
not SXML anymore. Any suggestions are welcome.

sxml

Dan Pimentel — 2009-12-20T21:24:02

Hello all! 

I am trying to learn SXML and attempting to follow to follow a tutorial at the following URL:

http://modis.ispras.ru/Lizorkin/sxml-tutorial.html

Right away I am receiving an error on example 1 hindering my progress:

here's the code (In Modular language and using PTL scheme 4.2.3 plus the latest version of SXML)



#lang scheme

(require (planet lizorkin/sxml:2:1/sxml))

(sxml:document "http://modis.ispras.ru/Lizorkin/XML/poem.xml")




The error i receive is as Follows

XLink API error: resource doesn't exist: http://modis.ispras.ru/Lizorkin/XML/poem.xml
#f


Could I be following old syntax from this website? The SXML source code (http://planet.plt-scheme.org/package-source/lizorkin/sxml.plt/2/1/xpath-context_xlink.ss) states 
"(define (sxml:document req-uri . namespace-prefix-assig)

and "NAMESPACE-PREFIX-ASSIG is an optional argument"


any help would be greatly appreciated,

Thanks!
Dan




      
------------------------------------------------------------------------------
This SF.Net em