xpath basics: axes

--===============2059234322==
Content-Type: multipart/alternative; boundary=047d7b67239a52701e04e2a69f0a

--047d7b67239a52701e04e2a69f0a
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Dear list members,

some time ago I asked a simple question about using XPath queries in Perl.
In the meantime, I have decided that xslt stylesheets are most appropriate
for my task. As this question now, however, is an XPath basic, I would
still like to ask it on this list:

I have a simple ddt:

<?xml version=3D"1.0" encoding=3D"UTF-8"?>
<!-- anno.dtd -->
<!ELEMENT corpus (sentence+)>
<!ELEMENT sentence (token+)>
<!ELEMENT token (#PCDATA)>
<!ATTLIST corpus name CDATA #REQUIRED>
<!ATTLIST sentence sent_id CDATA #REQUIRED>
<!ATTLIST token
    id CDATA #REQUIRED
    lemma CDATA #REQUIRED
    isterm (true|false) #REQUIRED
    pos CDATA #REQUIRED
    morph CDATA #REQUIRED
    dep CDATA #REQUIRED
    sfun CDATA #REQUIRED
>

Normally, in each sentence there are many tokens. I would now like to find
just a linear sequence of tokens:

<xsl:when
test=3D"token[@isterm=3D'true']/following-sibling::token[1][@pos=3D',']/fol=
lowing-sibling::token[2][@lemma=3D'=D0=BA=D0=BE=D1=82=D0=BE=D1=80=D1=8B=D0=
=B9']>

That is, I would like to find the sequence "TERM, =D0=BA=D0=BE=D1=82=D0=BE=
=D1=80=D1=8B=D0=B9",

However, adressing specific elements in the token array via [1] or [2] does
not seem to work, that is, my query ignores these indexes and randomly
retrieves elements on the following-sibling axis. Is there any mistake in
the query?
NB: Yes, my document text is in Cyrillic.

I'm grateful for any hint.

Best,
anne

--047d7b67239a52701e04e2a69f0a
Content-Type: text/html; charset=KOI8-R
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><div><div><div><div><div>Dear list members,<br><br>so=
me time ago I asked a simple question about using XPath queries in Perl. In=
 the meantime, I have decided that xslt stylesheets are most appropriate fo=
r my task. As this question now, however, is an XPath basic, I would still =
like to ask it on this list:<br>
<br></div>I have a simple ddt:<br><br>&lt;?xml version=3D&quot;1.0&quot; en=
coding=3D&quot;UTF-8&quot;?&gt;<br>&lt;!-- anno.dtd --&gt;<br>&lt;!ELEMENT =
corpus (sentence+)&gt;<br>&lt;!ELEMENT sentence (token+)&gt;<br>&lt;!ELEMEN=
T token (#PCDATA)&gt;<br>
&lt;!ATTLIST corpus name CDATA #REQUIRED&gt;<br>&lt;!ATTLIST sentence sent_=
id CDATA #REQUIRED&gt;<br>&lt;!ATTLIST token <br>=9A=9A=9A id CDATA #REQUIR=
ED<br>=9A=9A=9A lemma CDATA #REQUIRED<br>=9A=9A=9A isterm (true|false) #REQ=
UIRED<br>=9A=9A=9A pos CDATA #REQUIRED<br>
=9A=9A=9A morph CDATA #REQUIRED<br>=9A=9A=9A dep CDATA #REQUIRED<br>=9A=9A=
=9A sfun CDATA #REQUIRED<br>&gt;<br><br></div>Normally, in each sentence th=
ere are many tokens. I would now like to find just a linear sequence of tok=
ens:<br><br>&lt;xsl:when test=3D&quot;token[@isterm=3D&#39;true&#39;]/follo=
wing-sibling::token[1][@pos=3D&#39;,&#39;]/following-sibling::token[2][@lem=
ma=3D&#39;=CB=CF=D4=CF=D2=D9=CA&#39;]&gt;<br>
<br></div><div>That is, I would like to find the sequence &quot;TERM, =CB=
=CF=D4=CF=D2=D9=CA&quot;,<br></div><div><br></div>However, adressing specif=
ic elements in the token array via [1] or [2] does not seem to work, that i=
s, my query ignores these indexes and randomly retrieves elements on the fo=
llowing-sibling axis. Is there any mistake in the query?<br>
</div>NB: Yes, my document text is in Cyrillic.<br><br></div>I&#39;m gratef=
ul for any hint.<br><br></div>Best,<br>anne<br></div>

--047d7b67239a52701e04e2a69f0a--

--===============2059234322==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Perl-XML mailing list
[email protected]
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs

--===============2059234322==--