RE: [xsl] using xsl:message with UTF-8 characters

["Michael Kay" <mike@xxxxxxxxxxxx>]

Take this stylesheet:

<xsl:template name="main">
  <xsl:value-of select="'V: &#xaa;&#xba;&#x20ac;'"/>
  <xsl:message>M: &#xaa;&#xba;&#x20ac;</xsl:message>
</xsl:template>

If I run it from the command prompt in UltraEdit under WinXP, the message
(M:) displays correctly and the principal output (V:) does not.

If I run it from the command console (obtained using "cmd") both outputs are
garbage.

So it all depends on the environment; I'm afraid there's not much I can do
about that.

Michael Kay
http://www.saxonica.com/


> -----Original Message-----
> From: Manfred Staudinger [mailto:manfred.staudinger@xxxxxxxxx]
> Sent: 20 April 2007 23:22
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject: Re: [xsl] using xsl:message with UTF-8 characters
>
> On 20/04/07, Michael Kay <mike@xxxxxxxxxxxx> wrote:
> > By default the output from xsl:message in Saxon goes to System.err,
> > and the handling of special characters therefore depends
> partly on the
> > Java IO library and partly on the operating system (perhaps even on
> > which command shell you are using within the operating
> system). I've
> > generally found that a number of shells are unable to display even
> > iso-8859-1 correctly, which is why I reverted to use of ASCII
> > characters only in Saxon error messages. A bit
> disappointing in this day and age.
> I'm using win2000 and Java version 1.5.0_10. In my example I
> did not specify an output file so both, the output from
> xsl:message and the output doc are displayed in the command
> shell. Isn't this proof that at least the command shell can
> handle the characters as expected ?
>
> > It's easy enough to tell Saxon to send the message somewhere else
> > where it won't be mangled, but the question is, where do
> you actually want it?
> I would like the error msg stream to be sent to the shell in
> the same way as the output document was sent there. In
> testing, my output doc is normally set to go to a disc file
> so it would not be mixed up with the error messages.
>
> I've seen the -m classname parameter in your documentation,
> but after reading the description I still have no idea what
> to specify - most likely my ignorance in java.
>
> Manfred
>
> > > -----Original Message-----
> > > From: Manfred Staudinger [mailto:manfred.staudinger@xxxxxxxxx]
> > > Sent: 20 April 2007 21:12
> > > To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> > > Subject: [xsl] using xsl:message with UTF-8 characters
> > >
> > > Using XSLT 2.0 with Saxon 8.9.0.3J I would like to use
> xsl:message
> > > much more, but I don't know how to get
> > > UTF-8 chars displayed correctly. Here is what I get when I invoke
> > > test.xsl with saxon from the command line:
> > >
> > > >java net.sf.saxon.Transform dummy.xml test.xsl
> > > a or ?that's the question  <---- I replaced a square by a true ?
> > > <doc>a or d, that's the question</doc>>Exit code: 0
> > >
> > > The square stands for x'E4 AC A0' (which would triggers
> > > base64 encoding in gmail by the way) and the text has not
> only lost
> > > the "d"
> > > (x'C3 A4') but also the next 2 bytes ", " (x'C2 20').
> > >
> > > Any idea how to get xsl:message working ?
> > >
> > > Manfred
> > >
> > > test.xsl
> > > <?xml version="1.0" encoding="UTF-8" ?> <xsl:stylesheet
> > > version="2.0"
> > >   xmlns:xsl="http://www.w3.org/1999/XSL/Transform";>
> > > <xsl:output omit-xml-declaration="yes"/> <xsl:template match="/">
> > > <doc>
> > >   <xsl:message select="'a or d, that''s the question'"/>
> > >   <xsl:value-of select="'a or d, that''s the question'"/> </doc>
> > > </xsl:template> </xsl:stylesheet>