DO NOT REPLY [Bug 10009] - indentation not working with Node.appendChild()

[bu...@apache.org]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10009>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10009

indentation not working with Node.appendChild()

minchau@ca.ibm.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID



------- Additional Comments From minchau@ca.ibm.com  2004-02-18 17:21 -------
===== THIS WAS POSTED TO xalan-dev@xml.apache.org ON JUNE 24, 2002 ===========
===== BY Henry Zongaro =======================================================
Hi Pavani,

     Sorry I didn't follow up on this on xalan-j-users right away.  I've 
read section 16.1 of XSLT 1.0 ("XML Output Method") [1] a little more 
carefully, and there's one passage that I'm a little concerned about; it 
reads as follows:

[[
The xml output method should use an algorithm to output additional whitespace 
that ensures that the result if whitespace were to be stripped from the 
output using the process described in [3.4 Whitespace Stripping] with the set 
of \
whitespace-preserving elements consisting of just xsl:text would be the same 
when \
additional whitespace is output as when additional  whitespace is not output.
]]

     If the resulting document (before the effect of indent and 
indent-amount) contained a fragment like the following, does the passage 
that I've quoted from 16.1 mean that no additional whitespace should be 
introduced before the xsl:text end-tag?  If so, I believe the current 
behaviour of SerializerToXML would meet this requirement, but the change 
you're recommending would have to account for this case.  Similarly, if 
xml:space attributes appeared with the value "preserve".

<!--
   There are several of blanks and line-feed characters
   between the following tags
  -->
<xsl:text xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 
 

<xsl:text/>


     I'm also unsure whether the behaviour with your patch would be 
expected for a result like the following, where all blank lines between 
the tags contain only linefeed characters.

<abc>



xyz


</abc>

I believe your patch would strip out any linefeed characters that appeared 
in element content before any non-whitespace text, and leave alone any 
that came after, so producing the following result.

<abc>xyz


</abc>

Thanks,

Henry 
========================================================================

I have this to add.

The original input xml document was this
<Inventory_Attributes>
..<Products>
....<Product name="my_product">
......<Version>A</Version>
......<Version>B</Version>
......<Version>C</Version>
....</Product>
..</Products>
</Inventory_Attributes>

I've replaced the leading whitespace (spaces) in your document. Each group 
of leading whitespace characters (actually the new line character at
the end of the previous line plus the leading whitespace characters) is 
effectively considered to be  a <xsl:text> node
and is significant.  

Your new element that you append to be a child
of the <Product> node is inserted after the last child node of <Product> and 
the last child node is not <Version>C</Version> but the text node that is just 
before the closing </Product> tag. So you essentially have this in your DOM 
after the extra child is added:
<Inventory_Attributes>
..<Products>
....<Product name="my_product">
......<Version>A</Version>
......<Version>B</Version>
......<Version>C</Version>
....<Version>new_added_version</Version></Product>
..</Products>
</Inventory_Attributes>


During serialization an opening element tag or closing element tag will not be 
indented if a text node was serialized just before it, otherwise we are adding
characters to an existing text node and changing it. The only opening or 
closing tag that has no text before it is the </Product> tag so the
serializer is free to indent that one thing (I've put semicolons to indicate
the spaces inserted due to indentation being on:
<Inventory_Attributes>
..<Products>
....<Product name="my_product">
......<Version>A</Version>
......<Version>B</Version>
......<Version>C</Version>
....<Version>new_added_version</Version>
;;;;</Product>
..</Products>
</Inventory_Attributes>

Xalan is working correctly. This is not a bug.
Regards,
Brian Minchau