xml-xalan-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ari Miller (JIRA)" <xalan-...@xml.apache.org>
Subject [jira] Commented: (XALANJ-1974) Doubles newline character
Date Thu, 03 Feb 2005 17:44:25 GMT
     [ http://issues.apache.org/jira/browse/XALANJ-1974?page=comments#action_58567 ]
Ari Miller commented on XALANJ-1974:

I've divorced this from the HTML issues described in the original bug by the individual who
entered that bug, and I'm about to attach a reproducible testcase without external dependencies.

To explain the original HTML issues, as I understand them:
If you have a textarea on an HTML page, and you submit the text in it to an application server,
you end up persisting any line breaks as crlf.  If you then use that in your source xml document,
the crlf gets inappropriately transformed into &#13; and then a line feed, rather than
just a line feed.
The documents I am about to attach should clarify and allow for reproduction.
Because internet explorer treats the &#13 character in HTML as another line break, you
end up doubling the number of displayed line breaks when you read back the submitted text
into the textarea on the page.

Take a very simple xml input file, attached (transform.xml).  Put it in the directory /tmp/stylesheet_utilities.
Take the attached simplistic stylesheet, attached (paragraph.xsl).
Transform it using the attached code (UtilityStylesheetTester).  The attached code requires
the stylesheet to be in the same directory as you are running UtilityStylesheetTester from,
but it shouldn't have any other dependencies.
Output will be generated and saved to:

I think XERCESJ-92 has some useful information on this.  Specifically, this might be an issue
with org.apache.xerces.parsers.DOMParser, rather than xalan.  I think the &#xd;&#xa;
should be normalized to just &#xa; by xerces, when the Element is created (http://www.w3.org/TR/2000/REC-xml-20001006#sec-line-ends).

Note that running the same code with Saxon (saxon-6.5.3) does not show this problem (I will
attach the output as saxontransformed.html).

I'm using the libraries from xalan-j_2_6_0 in this example.

> Doubles newline character
> -------------------------
>          Key: XALANJ-1974
>          URL: http://issues.apache.org/jira/browse/XALANJ-1974
>      Project: XalanJ2
>         Type: Bug
>   Components: Serialization
>     Versions: 2.6
>  Environment: Windows Xp Professional O.S., Jrun 4, netBeans 3.6, sql server, jdk 1.4.2_03
>     Reporter: vaibhav kakadia
>     Priority: Critical

> Hi,
>    It seems to be a strage problem. I put a text like 
> 1
> 2
> and saves data. When I come back it is 
> 1
> 2
> I again saved data and came back it is
> 1
> 2
> It seems that it is unable to parse #xd;#xa; (CR-LF) into #xa; (LF). So it increases
newline character everytime. I don't know why this problem arises but when I was using older
version of Xalan set it was 2.2.0 this was not taking place.
> Please do the needful.
> Vaibhav

This message is automatically generated by JIRA.
If you think it was sent incorrectly contact one of the administrators:
If you want more information on JIRA, or have a bug to report see:

To unsubscribe, e-mail: xalan-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xalan-dev-help@xml.apache.org

View raw message