Hi All,

I have an existing XML file. After load(, modify) and save this file some mandatory formatting things lost from it. I need same indent as before, same text contents and not replace every special chars. With the below code I got the attached OUTPUT.xml from the attached INPUT.xml. Any idea how to solve this problem? (I use fixes_3_2.)

X:=TXMLDocument.Create;
ReadXMLFile(X,'INPUT.xml');
WriteXMLFile(X,'OUTPUT.xml');

Gabor

As far as I know you can't. I recently changed some things in xmlwriter so you can influence the formatting to some degree, but no attempt is made to respect the formatting of a previously read file. I believe the formatting info is discarded on read (I would need to veify) so this would require lots of rewriting.

Michael.

2019. 04. 27. 13:57 keltezéssel, Michael Van Canneyt írta:

The "indent" and "text content" problems solved on the reader side by ReadXMLFilePreserveWhitespace:

http://wiki.freepascal.org/XML_Tutorial#Whitespace_characters

Gabor

Nice. Again something learned.

In that case, disabling indent (set size=0) and CRLF (set lineending to 0-empty string) in the newly exposed XMLWriter will probably do the trick.

Michael.

2019. 04. 28. 9:35 keltezéssel, Michael Van Canneyt írta:

the newly exposed XMLWriter

Will be merged/backported into fixes_3_2?

Gabor

I just merged it.

Michael.

If you need the same indent or special chars, XML is not the right format for you.

Consider it

2019. 04. 28. 9:25 keltezéssel, Gabor Boros írta:

The "indent" and "text content" problems solved on the reader side by ReadXMLFilePreserveWhitespace:

With a sample application but not with the real life application. :-( (My real application just find node for every second FindNode call.)

With the attached example and test xml file I got the next result:

*a*
**

Is this not a bug? Lost of formatting is not disturb me but text between

and < is the data/text content of a node.

Gabor

2019. 04. 28. 21:24 keltezéssel, Santiago A. írta:

If you need the same indent or special chars, XML is not the right format for you. Consider it

XML is not my choice. ;-)

Gabor

Gabor Boros wrote

Is this not a bug? Lost of formatting is not disturb me but text between > and < is the data/text content of a node.

are you saying that you are trying to use fixed-width fields that are space-padded in XML files???

Am 30.04.2019 um 02:45 schrieb

There is nothing wrong with that,
in fact I am aware of a typesetting system which uses XML as input, and if you enclose your input text between certain tags (<pre> or <sourcecode>)
you expect that the blanks between those tags are preserved.

The W3C standards, IMO, don't tell anything about what a certain parser etc. should do to XML content (IMO, not much), there are not many restrictions
(<, of course, should be &lt;) ... attributes are different, of course.

Kind regards

Bernd

agreed...

in fact I am aware of a typesetting system which uses XML as input, and if you enclose your input text between certain tags (<pre> or <sourcecode>) you expect that the blanks between those tags are preserved.

true but the given examples do not show that...

[code]
<?xml version="1.0" encoding="utf-8"?>
<doc>
<text1> a</text1>
<text2> </text2>
</doc>
[/code]

2019. 04. 30. 2:45 keltezéssel, írta:

are you saying that you are trying to use fixed-width fields that are space-padded in XML files???

No. The XML files are exists already. The task is... Load the contents/file to a TObject descendant, modify the data in the object(s) then build up XMLs from it a save back to file(s). Without lost any "TDOMNode.TextContent".

Gabor