Hi,
What is the more simpler way to get all text in each paragraph in the XML below, using fcl-xml?
For example:

Page 1
-------------------
Para 1:
-1524
-888
-14/06/06
Para 2
-TEXT 2

Page 2
-------------------
Para 1
--TEXT page 2

I need each paragraph in a TStringList.

<?xml version="1.0" encoding="UTF-8"?>
<Document filename="127.pdf" pageCount="7" filesize="141859" linearized="false" pdfVersion="1.5">
<Pages>
<Page number="1" width="612" height="1008">
<Options>granularity=line</Options>
<Content granularity="line" dehyphenation="false" dropcap="false" font="false" geometry="false" shadow="false" sub="false" sup="false"> <Para>
<Line>
<Text>1524</Text>
</Line>
<Line>
<Text>888</Text>
</Line>
<Line>
<Text>14/06/06</Text>
</Line>
</Para>
<Para>
<Line>
<Text>TEXT 2</Text>
</Line>
</Para>
</Content>
</Page>
<Page number="2" width="612" height="1008">
<Options>granularity=line</Options>
<Content granularity="line" dehyphenation="false" dropcap="false" font="false" geometry="false" shadow="false" sub="false" sup="false"> <Para>
<Line>
<Text>TEXT page 2</Text>
</Line>
</Para>
</Content>
</Page>
</Pages>
</Document>

Thanks,
Marcos Douglas

There is no "simple" way. You must walk the nodes and append the text of all <Text> nodes.

Or you can try XPath, this should give you all the <Text> nodes, but I am not familiar with the exact structure of XPath.

Michael.

OK Michael, thanks.

Marcos Douglas

I have some code that reads/writes an addressbook in XML format. I can attach if if you think it may be usefull to you.

Bart

Bart wrote

I have some code that reads/writes an addressbook in XML format. I can attach if if you think it may be usefull to you.

OK, if you can I appreciate that.
Some functions of fcl-web returns a new object... I have to be careful with memory leaks.

Marcos Douglas

Hi Marcos,

OK, if you can I appreciate that. Some functions of fcl-web returns a new object... I have to be careful with memory leaks.

See: addressbook.pp.
Mind you, my code isn't web related, it just handles an xml-file.

Bart

Attached a sample xml file created with TAddressBook.

Bart

Bart wrote

Attached a sample xml file created with TAddressBook.

Hi Bart,
Sorry for late.
Thank you for addressbook.pp and test.xml files. I will study them.

Marcos Douglas

The Wiki has some XML tutorials:

http://wiki.lazarus.freepascal.org/XML_Tutorial

Felipe Monteiro de Carvalho wrote

The Wiki has some XML tutorials: http://wiki.lazarus.freepascal.org/XML_Tutorial

Yes, I saw that.
My doubt is how use a safe algorithm to parse my XML. See the first
example "Printing the names of nodes and attributes" in link that you
posted:
with Child.ChildNodes do
try
//..
finally
  Free;
end;

I have to know which functions create a new object. That is not clear if we use Child.ChildNodes, right?

Marcos Douglas

My XmlToList procedure is based upon the XmlToTree example in the wiki.

There are no memory leaks in my code (checked it with heaptrace).

Bart

Bart wrote

My XmlToList procedure is based upon the XmlToTree example in the wiki. There are no memory leaks in my code (checked it with heaptrace).

Hi Bart,
Below is my "simple way" to get all <Text> for each <Page>:

----------------------------------------------------------------------------
program t_xml;
{$mode objfpc}{$H+}
uses
  Classes, sysutils, dom, XMLRead;
procedure ProcessPage(ANode: TDOMNode; APageNo: Integer);
var
  Node: TDOMNode;
begin
  if not Assigned(ANode) then
    Exit;
  Node := ANode.FirstChild;
  while Assigned(Node) do
  begin
    if Node.CompareName('Text') = 0 then
    begin
      writeln(Node.TextContent); // << do anything...
    end;
    Node := Node.NextSibling;
    ProcessPage(Node, APageNo);
  end;
end;
procedure ProcessNodes(ANode: TDOMNode);
var
  Node: TDOMNode;
  PageNo: Integer;
begin
  if not Assigned(ANode) then
    Exit;
  Node := ANode.FirstChild;
  while Assigned(Node) do
  begin
    if Node.CompareName('Page') = 0 then
    begin
      PageNo := StrToInt(Node.Attributes.GetNamedItem('number').NodeValue);
      ProcessPage(Node, PageNo);
    end;
    Node := Node.NextSibling;
    ProcessNodes(Node);
  end;
end;
var
  Doc: TXMLDocument;
begin
  ReadXMLFile(Doc, 'cc.xml');
  try
    ProcessNodes(Doc);
  finally
    Doc.Free;
  end;
end.
----------------------------------------------------------------------------

Thanks for you example!
BTW, I could not use FindNode('Pages'). Looks like is broke, isn't?

Marcos Douglas