Using xmldom.writetoclob

I am using the procedure xmldom.writetoclob with the PL/SQL
XDK.
I am running into speed concerns.
Below are some results of tests I have done.
As you can see the time taken to write the contents of the
DOM out to a clob can quickly become problem for a user waiting
for a reponse.
Can any one tell me of a way to speed up this process or a
work around?
Thanks very much
Rick Laird
clob size= 2913
start xmldom.writetoclob= 01:57:53
end xmldom.writetoclob= 01:57:54
clob size= 8909
start xmldom.writetoclob= 01:58:06
end xmldom.writetoclob= 01:58:12
clob size= 14905
start xmldom.writetoclob= 01:58:45
end xmldom.writetoclob= 01:58:55
clob size= 22400
start xmldom.writetoclob= 01:59:44
end xmldom.writetoclob= 01:59:59
clob size= 29895
start xmldom.writetoclob= 02:01:09
end xmldom.writetoclob= 02:01:29
clob size= 172300
start xmldom.writetoclob= 02:05:59
end xmldom.writetoclob= 02:07:54
clob size= 286224
start xmldom.writetoclob= 02:15:34
end xmldom.writetoclob= 02:18:42

what version of XDK are you using ?
How much memory do you have ?
I am getting this results :
     Writting to clob     01/09/2002 9:51:07 AM
     Done writting to clob     01/09/2002 9:51:13 AM
     Size of clob : 370433     01/09/2002 9:51:13 AM

Similar Messages

Using xmldom.writeToClob to update xml clob corrupts clob

We are storing an xml document in a clob field in the db. As part of a data check we need to parse the xml and possibly remove a node. I can do that, and check the results using 'xmldom.writeToBuffer'.
When I try to write the data back to the db using 'xmldom.writeToClob' I get strange results. Data seems to be appended to the end of the clob and contains chr(13) characters at the end of each line.
Also does anyone know if the existence of whitespace in the xml clob would cause a parse/search to be slower?
I am new to xml and might be missing something fairly basic in the following code. Thanks in advance for any help.
Here is some of the code:
doc xmldom.DOMDocument;
curNode xmldom.DOMNode;
parentNode xmldom.DOMNode;
theNodeList xmldom.DOMNodeList;
str varchar2(4000);
begin
-- open xml_rec cursor (xml is the clob field opened for update)
-- Loop through all item elements
theNodeList := xmldom.getElementsByTagName(doc, 'item');
for m in 0..xmldom.getLength(theNodeList)-1 loop
curNode := xmldom.item(theNodeList,m);
--perform a check and possibly delete the current node
parentNode := xmldom.getParentNode(curNode);
parentNode := xmldom.removechild(parentNode, curNode);
xmldom.writeToBuffer(doc, str);
--I check the results here and everything looks good
xmldom.writeToClob(doc, xml_rec.xml);
--When I check the clob the data is incorrect.
null

The next cod is a sample of the implementation xmldom.writeToClob used xmldom.DOMNode in the parameter of input.
DECLARE
XMLOut CLOB := EMPTY_CLOB();
XMLIn Varchar2(2000);
XSLPath Varchar2(2000);
BEGIN
DBMS_LOB.CREATETEMPORARY(XMLOut,TRUE); -- Give permit
DBMS_LOB.OPEN (XMLOut, DBMS_LOB.LOB_READWRITE); --open the file of read and write
XSLPath := '/users/gcardona/plantillas/report.xsl';
GE_BSXMLCONVERT.createElementName('COLUMN');
GE_BSXMLCONVERT.createElementAttribute('name', 'OUTXML');
GE_BSXMLCONVERT.addElement();
GE_BSXMLCONVERT.createElementName('table');
GE_BSXMLCONVERT.createElementAttribute('name', 'X1');
GE_BSXMLCONVERT.createElementAttribute('width', '100');
GE_BSXMLCONVERT.addElement();
XMLIn := GE_BSXMLCONVERT.toXML();
XSLTranform(XMLIn, XSLPath, XMLOut);--XMLOut is a variable in out in the procedure,
--here is where is return the file XML Process
insert into in_prueba (ID,XML_CLOB) values (200, XMLOut);
DBMS_LOB.CLOSE (XMLOut); -- Close File
DBMS_LOB.FREETEMPORARY(XMLOut); --free memory
GE_BSXMLCONVERT.clearMemory();
commit;
END;
null

Xmldom.writetoclob hanging forever when writing a domnode to a temp clob

Hi
Wondering has anyone come across anything similar (or know a work around)
1. Basically the the procedure below loads an XML file. (OK)
2. Removes reference to external dtd, ref was causing and error (OK)
3. Parses the XML (OK)
4. Gets a list of the "item" tags (OK) - should be up to 500 item elements in XML
5. Loops on all elements
a. writes node to temp clob *(PROBLEM HERE)*
b. Creates xmltype from clob (OK)
c. transforms xmltype and assigns to another xmltype (OK)
d. inserts new xmltype (OK)
This procedure is used by a threaded os java app to load thousands of files to db.
It works fine 99% of the time but for some files the procedure is hanging at (5.a : write domnode to clob using xmldom.writetoclob)
I am absolutely sure the problem is due to the characters in the domnode as all files that have hung on processing contain these bizarre characters,
the db character set is Unicode AL32UTF8 so it should be able to handle almost anything, saying that, if the it is a prob with the characters then why does
the loadclobfromfile procedure succeed.
Developed on Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit
I have no prob rejecting node and continue processing but procedure just hangs and i have to kill the thread on the os.
Is there a way to
1. Test for invalid AL32UTF8 chars and skip on error.
or
2. Get around the xmldom.writetoclob procedure and create an xmltype from the xmldom.domnode (i tried to convert domnode -> domdocument -> xmltype, no joy)
or
Open to suggestions
PROCEDURE
PROCEDURE process_blog_xml_file (
      p_file_name     IN   VARCHAR2,
      p_insert_date   IN   VARCHAR2,
      p_dir           IN   VARCHAR2 DEFAULT NULL
   IS
      newsurl        VARCHAR2 (80);
      parser         xmlparser.parser;
      newsxml        xmldom.domdocument;
      titles         xmldom.domnodelist;
      titles_found   NUMBER;
      curnode        xmldom.domnode;
      textchild      xmldom.domnode;
      dest_clob      CLOB;
      l_temp_clob    CLOB;
      src_clob       BFILE         := BFILENAME (g_zip_file_dir, p_file_name);
      dst_offset     NUMBER             := 1;
      src_offset     NUMBER             := 1;
      lang_ctx       NUMBER             := DBMS_LOB.default_lang_ctx;
      warning        NUMBER;
      l_xml          XMLTYPE;
      xsldata        XMLTYPE;
      xmldata        XMLTYPE;
      l_dir          VARCHAR2 (1000);
   BEGIN
      LOG (   'Start Processing file '
           || p_file_name
           || ' from zip '
           || p_insert_date
      IF p_dir IS NULL
      THEN
         SELECT directory_path
           INTO l_dir
           FROM all_directories
          WHERE directory_name = g_zip_file_dir;
      ELSE
         l_dir := p_dir;
      END IF;
      IF g_xsl_clob IS NULL
      THEN
         init_xsl_clob;
      END IF;
      com_polecat_xmldb_utils.set_load_date
                                  (TO_DATE (REPLACE (REPLACE (p_insert_date,
                                                     '.zip',
                                            'DD-MM-YYYY'
      LOG ('ITEM.Loaddate set : ' || com_polecat_xmldb_utils.get_load_date);
      xsldata := XMLTYPE.createxml (g_xsl_clob);
      LOG ('Read xsl file to clob');
      DBMS_LOB.OPEN (src_clob, DBMS_LOB.lob_readonly);
      DBMS_LOB.createtemporary (dest_clob, TRUE);
      DBMS_LOB.loadclobfromfile (dest_lob          => dest_clob,
                                 src_bfile         => src_clob,
                                 amount            => DBMS_LOB.getlength
                                                                     (src_clob),
                                 dest_offset       => dst_offset,
                                 src_offset        => src_offset,
                                 bfile_csid        => NLS_CHARSET_ID
                                                                   ('AL32UTF8'),
                                 lang_context      => lang_ctx,
                                 warning           => warning
      LOG ('Read xml file to clob');
      DBMS_LOB.CLOSE (src_clob);
      parser := xmlparser.newparser;
      dest_clob := REPLACE (dest_clob, g_blog_dtd_remove, '');
      LOG ('Parse xml ');
      xmlparser.parseclob (parser, dest_clob);
      DBMS_LOB.freetemporary (dest_clob);
      newsxml := xmlparser.getdocument (parser);
      xmlparser.freeparser (parser);
      titles := xmldom.getelementsbytagname (newsxml, 'item');
      LOG ('Load blogs items to ITEM table. ');
      FOR j IN 1 .. xmldom.getlength (titles)
      LOOP
         curnode := xmldom.item (titles, j - 1);
         DBMS_LOB.freetemporary (l_temp_clob);
         DBMS_LOB.createtemporary (l_temp_clob, TRUE);
         LOG ('write node to temp clob ' || j);
         xmldom.writetoclob (curnode, l_temp_clob);                                              <-- Hanging Here
         LOG ('create xml type from clob ' || j);
         l_xml := XMLTYPE.createxml (l_temp_clob);
         LOG ('apply xsl transform to xml ' || j);
         xmldata := l_xml.transform (xsldata);
         LOG ('Insert to item table ' || j);
         BEGIN
            INSERT INTO item
                 VALUES (xmldata);
         EXCEPTION
            WHEN OTHERS
            THEN
               LOG ('Error:: ' || SQLERRM);
         END;
      END LOOP;
      xmldom.freedocument (newsxml);
      LOG (   'Finished Processing file '
           || p_file_name
           || ' from zip '
           || p_insert_date
   EXCEPTION
      WHEN OTHERS
      THEN
         LOG (SQLERRM);
   END;note variables starting with g_ are defined in package spec
Cheers
Ian
Edited by: user3604054 on 01-Apr-2010 06:52
Edited by: user3604054 on 01-Apr-2010 14:57
Edited by: user3604054 on 01-Apr-2010 15:00
Edited by: user3604054 on 01-Apr-2010 15:00
Edited by: user3604054 on 01-Apr-2010 15:05
Edited by: user3604054 on 01-Apr-2010 15:06

2. Get around the xmldom.writetoclob procedure and create an xmltype from the xmldom.domnode (i tried to convert domnode -> domdocument -> xmltype, no joy) Which version of Oracle (4 digits)?
Also look in the FAQ in the upper right for how to use the tag to wrap PL/SQL to retain formatting to make it easier for all to read.

Problem with special characters and xmldom.writetoclob

Hi!
I have a problem with oracle parser in pl/sql.
My situation:
i have a clob with valid xml with for example decoded characters:
(spaces between & and # are for good representing characters by browser - in really there is no spaces)
<xml>
<any><![CDATA[ & #187; ]]> text <![CDATA[ & #187; ]]></any>
</xml>
i read this xml from clob, parse it and put into another clob by using
xmldom.writetoclob procedure.
And in second clob i have"
<xml>
<any>& #38;#187; text & #38;#187; </any>
</xml>
(ampersand is representing as & #38; !) Why there is not the cdata sections ?
Why the value of these sections changed?
Any ideas? I'm using newset xdk (9202) in 8.1.7.3 database.
Please help!

Hi John,
According to your description, my understanding is that the Author showed incorrect character in SharePoint 2013 search result page.
I tested the same scenario per your post in my environment, and the Müller showed correctly in SharePoint search result page.
I recommend to reset the index in Search Service Application and run a full crawl to see if the issue still occurs.
Best regards.
Thanks
Victoria Xia
TechNet Community Support

To : Oracle Team.. xmldom.writetoclob is too slow

HI;
I had Oracle 8.1.6, Used the xmldom package a lot to manipulate and create xml documents ( all sizes ). then I used xmldom.writetoclob which it was giving a GREAT performance.
I UPGRADED ti Oracle 8.1.7 and xmldom.writetoclob is TOO SLOW now, I have the same settings ( java_pool_size, share_pool, etc)
What changed from 8.1.6 to 8.1.7 ??
Why xmldom.writetoclob is so much slower in 8.1.7 ??
when tranfering a 300 KB document with 8.1.6 it used to take 30 seconds, no with 8.1.7 IT TAKES 2-1/2 minutes ?????
WHY ? ? ? ? ?
Tahnks for any information..

My guess is that you are creating a clob with cached set to false
e.g
dbms_lob.createtemporary(v_myclob,false);
When set to false any time the clob is written to the characters are written to the disk. causing it to be slow.
try
dbms_lob.createtemporary(v_myclob,true);
I had the same problem. This speeds it up significantly
Rick Laird

Steve, why is xmldom.writetoclob so slow ??

I am using xmldom.writetoclob, which is OK when transfering small ( < 1MG ) documents to a CLOB, but as the documetns get bigger it is very slow..
It takes 15 minutes to transfer a 5MB document to a clob using xmldom.writetoclob.
Anything I should be doing to see better results ?
Thanks.
I am using the PL/SQL parser

what is your java_pool_size?

Error While Writing DOM to a CLOB i.e. xmldom.writeToClob

I am running into error while writng and reading from a clob. The Sample Code is as follows :
declare
p xmlparser.Parser;
doc xmldom.DOMDocument;
i_tmp CLOB;
i_tmp1 CLOB;
i_buffer varchar2(32767);
i_amount pls_integer;
begin
p := xmlParser.NewParser;
xmlparser.setbaseDir(p,'/sqlcom/inbound');
xmlparser.setValidationMode(p, TRUE);
xmlparser.showwarnings(p, TRUE);
xmlparser.seterrorlog(p, '/sqlcom/inbound/veshaal.err');
xmlparser.parse(p,'/sqlcom/inbound/mpoi.xml');
doc := xmlparser.getDocument(p);
dbms_lob.createtemporary(i_tmp1,true,dbms_lob.session);
xmldom.writeToClob(doc,i_tmp1);
xmldom.writeToFile(doc,'/sqlcom/inbound/veshaal_out.xml');
dbms_lob.createtemporary(i_tmp,true,dbms_lob.session);
xmldom.writeTobuffer(doc,i_buffer);
dbms_lob.write(i_tmp,length(i_buffer),1,i_buffer);
i_amount := dbms_lob.getLength(i_tmp);
dbms_lob.read(i_tmp,i_amount,1,i_buffer);
dbms_lob.freetemporary(i_tmp1);
commit;
exception
when others then
dbms_lob.freetemporary(i_tmp1);
dbms_output.put_line(SQLERRM);
end;
The oracle Error is as follows:
declare
ERROR at line 1:
ORA-03113: end-of-file on communication channel
ORA-24323: value not allowed
Error accessing package DBMS_APPLICATION_INFO
ERROR:
ORA-03114: not connected to ORACLE
The above code works while I try to write to a buffer or a file using xmldom.writeToFile or xmldom.writeToBuffer.
The xmlparser v2 is running inside the 8i ( 8.1.5.0. ) database.
The alert file has the following entry :
Errors in file /home/applrt/db/superbwl/log/superbwl_ora_2371.trc:
ORA-07445: exception encountered: core dump [EE1F05DC] [SIGSEGV] [Address not mapped to object] [0] [] []
Any help will be appreciated.
Thanks
null

UTL_FILE comes as part of the database, you shouldn't have to install it, and you most certainly shouldn't install it as SCOTT. System packages should be installed using a system user such as sys.
Firstly, I would suggest dropping the UTL_FILE package from your scott schema and then logging on as sys and granting execute permission on UTL_FILE to the scott user.
Secondly...
The UTL_FILE_DIR parameter has been deprecated by oracle in favour of direcory objects because of it's security problems.
The correct thing to do is to create a directory object e.g.:
CREATE OR REPLACE DIRECTORY mydir AS 'c:\myfiles';Note: This does not create the directory on the file system. You have to do that yourself and ensure that oracle has permission to read/write to that file system directory.
Then, grant permission to the users who require access e.g....
GRANT READ,WRITE ON DIRECTORY mydir TO myuser;Then use that directory object inside your FOPEN statement e.g.
fh := UTL_FILE.FOPEN('MYDIR', 'myfile.txt', 'r');Note: You MUST specify the directory object name in quotes and in UPPER case for this to work as it is a string that is referring to a database object name which will have been stored in uppercase by default.

Potential problem with XMLDOM.writeToClob()

I'm really hoping that someone can help me with this. Last fall, wrote some code that used XMLDOM to build a document and send it to a web service. I remember having problems with WriteToBuffer and WriteToClob during my development efforts, which I thought were successfully resolved. Now it looks like the problem may be occurring at a client who recently upgraded to use this code. I can't get this to fail in-house, so I can't do much debugging on it. So I'm hoping that someone else has come across a problem like this and remembers what they did to fix it.
Basically, what happened was, WriteToClob (or sometimes WriteTuBuffer) didn't return anything, even though the DOM document was not empty. I remember going back and forth between WriteToBuffer and WriteToClob, trying to get it resolved. But I don't remember specifically finding any "magic bullet". Here's what I currently have in the code:
DBMS_LOB.createtemporary (v_source_data, TRUE);
DBMS_LOB.OPEN (v_source_data, DBMS_LOB.lob_readwrite);
xmldom.writetoclob (v_main_doc, v_source_data);
DBMS_LOB.CLOSE (v_source_data);
Has anyone had a similar problem, and remember what the fix was? I'm also not sure that the Open/Close calls are necessary. Or what TRUE means on CreateTemporary (aside from some generic comment about "buffer cache" without telling me what a buffer cache is).

I'm really hoping that someone can help me with this. Last fall, wrote some code that used XMLDOM to build a document and send it to a web service. I remember having problems with WriteToBuffer and WriteToClob during my development efforts, which I thought were successfully resolved. Now it looks like the problem may be occurring at a client who recently upgraded to use this code. I can't get this to fail in-house, so I can't do much debugging on it. So I'm hoping that someone else has come across a problem like this and remembers what they did to fix it.
Basically, what happened was, WriteToClob (or sometimes WriteTuBuffer) didn't return anything, even though the DOM document was not empty. I remember going back and forth between WriteToBuffer and WriteToClob, trying to get it resolved. But I don't remember specifically finding any "magic bullet". Here's what I currently have in the code:
DBMS_LOB.createtemporary (v_source_data, TRUE);
DBMS_LOB.OPEN (v_source_data, DBMS_LOB.lob_readwrite);
xmldom.writetoclob (v_main_doc, v_source_data);
DBMS_LOB.CLOSE (v_source_data);
Has anyone had a similar problem, and remember what the fix was? I'm also not sure that the Open/Close calls are necessary. Or what TRUE means on CreateTemporary (aside from some generic comment about "buffer cache" without telling me what a buffer cache is).

Xmldom.writetoclob give wrong character set

I use xmldom to create a xml clob. I create document for xml as follow:
doc := xmldom.NewDomDocument;
xmldom.setVersion(doc, '1.0');
xmldom.setStandalone(doc, 'no');
xmldom.setCharSet(doc, 'ISO-8859-1'); -- Should be character set for danish.
Create xml clob as follow:
xmldom.writeToClob(root, out_xml);
Problem: Now is danish 'x, f, e' replace with ??.
When a use 'xmldom.WriteToFile(root, 'd:\test_document');' I get correctly danish 'x, f, e'!
I have seen you can use procedure/function 'SetEncoding', but I can not find this function in my Oracle installation (8.1.7.1.1).

The character set you specify via setCharset() procedure is ignored unless you use writeToFile() later.
http://docs.oracle.com/cd/E11882_01/appdev.112/e25788/d_xmldom.htm#CHDCGDDB
Usage Notes
This is used for WRITETOFILE Procedures if not explicitly specified at that time.You can also use something like this :
SQL> set serveroutput on
SQL>
SQL> declare
2
3   export_file clob;
4   prolog       clob := '<?xml version="1.0" encoding="UTF-8"?>';
5
6 begin
7
8    select prolog || chr(10) ||
9           xmlserialize(document
10             xmlelement("TextTranslation"
11             , xmlattributes(
12                 '1.0' as "version"
13               , 'ja' as "language"
14               , 'DEMOAND' as "module"
15               , 'VC' as "type"
16               )
17             )
18             indent
19           )
20    into export_file
21    from dual ;
22
23    dbms_output.put_line ( export_file );
24
25 end;
26 /
<?xml version="1.0" encoding="UTF-8"?>
<TextTranslation version="1.0" language="ja" module="DEMOAND" type="VC"/>
PL/SQL procedure successfully completed

Xmldom.writeToClob error. [URGENT]

Any idea why I get the following error when I call:
xmldom.writeToClob(root, xml_out);
ORA-20000: An internal error has occurred: CLOB to write to can not be null ORA-06512: at "XML.XMLDOM", line 37 ORA-06512: at "XML.XMLDOM", line 358 ORA-06512: at line 131
'xml_out' is initialised with null, but 'root' have some nodes and elements, but no text nodes. I do not get this error if I use
xmldom.writeToBuffer(root, v_xml);
Thanks
null

Thx for your response
the program looks like this:
I executed it from a Form;
PROCEDURE PRUEBA IS
xml clob;
vxml varchar2(100);
len integer := 1;
pos integer := 1;
f utl_file.file_type;
doc xmldom.DOMDocument;
ppal_node xmldom.DOMNode;
main_node xmldom.DOMNode;
root_node xmldom.DOMNode;
user_node xmldom.DOMNode;
tercer_node xmldom.DOMNode;
item_node xmldom.DOMNode;
ppal_elmt xmldom.DOMElement;
root_elmt xmldom.DOMElement;
item_elmt xmldom.DOMElement;
item_text xmldom.DOMText;
--Elementos de la cabecera
header_pi xmldom.DOMProcessingInstruction;
header_node xmldom.DOMNode;
CURSOR MODULOS IS
SELECT T_COAPI AS COAPI,
T_DNAPI AS DNAPI
FROM     T
WHERE     T_COAPI = 'PES';
BEGIN
--Creamos un nuevo DOM document
doc := xmldom.newDOMDocument();
ppal_node := xmldom.makeNode(doc);
FOR Modulo IN Modulos LOOP
-- El nodo es cada modulo
root_elmt := xmldom.createElement(doc, 'Modulo');
root_node := xmldom.appendChild(main_node, xmldom.makeNode(root_elmt));
--COAPI
item_elmt := xmldom.createElement(doc, lit_COAPI);
item_node := xmldom.appendChild(root_node, xmldom.makeNode(item_elmt));
item_text := xmldom.createTextNode(doc, Modulo.COAPI);
item_node := xmldom.appendChild(item_node, xmldom.makeNode(item_text));
--DNAPI
item_elmt := xmldom.createElement(doc, lit_DNAPI);
item_node := xmldom.appendChild(root_node, xmldom.makeNode(item_elmt));
item_text := xmldom.createTextNode(doc, Modulo.DNAPI);
item_node := xmldom.appendChild(item_node, xmldom.makeNode(item_text));
END LOOP
f := utl_file.fopen('TEMP_DIR','PRUEBA.xml','w');
--Obtenemos la longitud del xml lob
len := dbms_lob.getlength(xml);
DBMS_LOB.CreateTemporary(xml,TRUE);
xmldom.writeToClob(doc,xml); -- HERE I GET THE ERROR
dbms_xmldom.WriteToFile(f,xml);
WHILE (pos < len) LOOP
vxml := dbms_lob.substr(xml,100,pos);
utl_file.put(f,vxml);
pos := pos+100;
END LOOP;
utl_file.fclose(f);
END;

Xmldom.writetoClob()

Hi ,
I've some problem saving back my xml document in Clob:
If I performe the procedure
xmldom.writetobuffer(DomDocument,variable_of_output);
the variable_of_output in varchar2 is well valorized.
Instead
If I performe the procedure
xmldom.writetoClob(DomDocument,variable_of_output);
where the variable_of_output is Clob
THE CODE STOP. The problem is present Even if i inizialize variable_of_output:=empty_clob();
Answer my question,Please.
Thanks you

Hi,
It probably depends on your database version. If your using 9i, you should use dbms_xmldom instead of xmldom. Furthermore, you might consider to use the following code (instead of the empty clob code):
DECLARE
t_doc DBMS_XMLDOM.DOMDocument;
t_clob CLOB;
BEGIN
... do stuff to fill t_doc
DBMS_LOB.createTemporary(t_clob, FALSE);
DBMS_XMLDOM.writeToClob(t_doc, t_clob);
DBMS_XMLDOM.freeDocument(t_doc);
... do stuff with t_clob
DBMS_LOB.freeTemporary(t_clob);
END;
I'm certain that dbms_xmldom replaces xmldom (v9.2+), I only can't remember the URL which says so...
Kind regards,
Michiel

Using XMLDOM to process document in XML DB

I am new to XML and especially XML DB which I however see as an excellent platform on which to base XML based solutions. I have a specific question and apologise beforehand should any of my assumptions be incorrect.
I wish to create an application which from a number of XML fragments, together with rules governing the behaviour of the fragments (insertion points, overwrite rules etc) are merged together to finally leave me with a complete and valid XML document in XML DB.
I thought at first that I could use DBMS_XMLDOM to perform this low level processing required. I have however not managed to understand how this can be used together with XMLDB. I am reluctant to use XMLDOM without XMLDB as I am afraid that I will surpass the limits in XML document size apparently imposed by XMLDOM.
I see a solution where I create a schema based resource in XML DB containing the initial XML structure and then processing any number of the above mentioned fragments to finaly be left with a complete XML document.
Can I use XMLDOM to do this or am I barking up the wrong tree?
I would be grateful for any input on techniques that I can apply to perform elementary search/replace/insert on an XML document which I am, from PLSQL, compiling.
Thanks
Hans Christiansen

Hi
The size of the documents I want to compose are expected vary from a few Kb to 1 or 2 Gb! In a FAQ found somewhere on OTN there was an entry which mentioned that somebody was having trouble with a 50Mb XML document. The recommendation was to use SAX but I am under the impression that this is more useful when the processing is of a more sequential nature.
The main thread of my question is to gain clarity in the most suitable technique to use when actualy compsing the XML.
What I want to do is:
Create resource (XML document) in XML DB.
Loop through XML fragments together with rules:
For each fragment process according to rules and insert
fragment or update existing XML based on fragment contents.
At the end of the loop the XML document is complete and ready for further processing.
XMLDOM seems to give me the tools to do this but I am unsure of how to do this and have XMLDOM work directly against the contents in the resource controlled by XML DB (making use of possible indexes and the scalable nature of XMLDB). I get the feeling that if I simply read the XML from XMLDB resource into a DOM document I am still using the memory based representation of the XML until I write the XML back to XMLDB - is this a correct assumption.
Alternatively is it the functionality offered by XPath, extract(), updatexml() etc. that I should be looking at?
Does this make my question more understandable?
Hans

Load XLM into pl sql using xmldom

Hi ,
I have problems when I use xmldom to retrieve infotmation from xml to storethe information into a table. For the first xml , I can do it , for the second one I have problems ( I can't retrieve the values I need ) and I don't know why. Please don't reply me by providing links , I just would like to have an idea of what is going wrong with the second xml. Thanks in advance.
It works fine with this one :
<?xml version="1.0" encoding="windows-1252" ?>
- <GraydonBeDialogue>
<TransactionCode>RTB</TransactionCode>
- <Table ClassTable="Country">
- <TableEntry>
<TableLanguage>N</TableLanguage>
<TableCode>AD</TableCode>
<TableValue>Andorra</TableValue>
</TableEntry>
- <TableEntry>
<TableLanguage>N</TableLanguage>
<TableCode>AE</TableCode>
<TableValue>Verenigde Arabische Emiraten</TableValue>
</TableEntry>
- <TableEntry>
<TableLanguage>N</TableLanguage>
<TableCode>AF</TableCode>
<TableValue>Afghanistan</TableValue>
</TableEntry>
- <TableEntry>
<TableLanguage>N</TableLanguage>
<TableCode>AG</TableCode>
<TableValue>Antigua en Barbuda</TableValue>
</TableEntry>
But it does not work with this one :
<?xml version="1.0" encoding="windows-1252" ?>
- <GraydonBeDialogue xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.w3schools.com Z:\PROGRAMMATIE\AS400\projecten\XML\GraydonBe\InfoReview\Review.xsd" xmlns="http://www.w3schools.com">
<Transaction>PDT</Transaction>
- <GraydonBeInformation>
- <Default>
<DateFormat>YYYYMMDD</DateFormat>
<TimeFormat>HHMMSS</TimeFormat>
<Currency>EUR</Currency>
<CountryCode>BE</CountryCode>
<DecimalSign>.</DecimalSign>
</Default>
- <Header>
<Name>GRAYDON BELGIUM NV /SA</Name>
<ContentGenerationDate>20050722</ContentGenerationDate>
<ContentGenerationTime>101110</ContentGenerationTime>
<Product>REV</Product>
<ClientNumber>156</ClientNumber>
</Header>
- <Body>
- <Msg48>
- <MsgHeader>
<MsgDate>20050712</MsgDate>
<MsgCode Table="RevDsoMsg">48</MsgCode>
- <Business>
<BusinessNumber>0480316383</BusinessNumber>
<Name>STANLEYBET BELGIQUE SA</Name>
- <Address>
<Street>PEGASUSLAAN 5</Street>
<PostCode>1831</PostCode>
<City>MACHELEN</City>
</Address>
</Business>
</MsgHeader>
<SinceDate Format="YYYYMM">200506</SinceDate>
</Msg48>
- <Msg123>
- <MsgHeader>
<MsgDate>20050711</MsgDate>
<MsgCode Table="RevDsoMsg">123</MsgCode>
- <Business>
<BusinessNumber>0403157932</BusinessNumber>
<Name>CELANESE NV</Name>
- <Address>
<Street>INDUSTRIEWEG 80</Street>
<PostCode>3620</PostCode>
<City>LANAKEN</City>
</Address>
</Business>
</MsgHeader>
- <PreviousCreditAssessement>
<CreditAssessementAmount>10691000</CreditAssessementAmount>
</PreviousCreditAssessement>
- <NewCreditAssessement>
<CreditAssessementAmount>9706000</CreditAssessementAmount>
</NewCreditAssessement>
</Msg123>
Could you explain me what could be wrong ?
I am using the same logic for both xml :
declare
bnb xmldom.domnodelist;
l_node xmldom.domnode;
l_doc xmldom.domdocument;
begin
parse_document (l_file_name, doc_in);
bnb := selected_nodes (doc_in, '/GraydonBeDialogue/Table[@ClassTable="Review"]/TableEntry');
FOR entry_index IN 0 .. xmldom.getlength (bnb) - 1
LOOP
l_node := xmldom.item (bnb, entry_index);
TableLanguage := trim(xslprocessor.valueof (l_node, 'TableLanguage'));
TableCode := trim(xslprocessor.valueof (l_node, 'TableCode'));
TableValue := trim(xslprocessor.valueof (l_node, 'TableValue'));
-- Treat the data
.....

Couple of friendly tips.
When pasting XML, don't copy it directly from the browser's normal view as that usually inserts a "-" on some lines, which makes it invalid XML when you copy/paste. Either open the XML in something besides a browser or do a View Page Source (or similar) to copy the true XML.
Use the tag to wrap XML and code to retain formatting as shown in the FAQ (under your sign-in name in the upper right).
In your first sample XML, the root node is defined as
<GraydonBeDialogue>
and in the second it is defined as (ignoring attributes that don't matter to us now)
<GraydonBeDialogue xmlns="http://www.w3schools.com">
This means the second XML resides in a default namespace and so to access any/all nodes that reside in this namespace, you must provide that namespace to the extract method.
For many Oracle based functions that take an Xpath as a parm, Oracle has the next parm defined as the namespace. This allows you to pass in the namespace associated to the XML nodes. With default namespaces, it does add a twist as you have to make up a namespace prefix for the Xpath/namespace declaration so that the engine knows the nodes reside in a namespace (be it default) instead of not residing in a namespace. Here's a little something I wrote up on this
http://anononxml.blogspot.com/2010/06/xml-parsing-with-default-namespaces-via.html
I prefer now to use XMLType when parsing XML in PL/SQL as a bit more flexible/easier to code.
Without seeing what you hid in
selected_nodes
it is hard to say exactly where your problem is but hopefully that gives you a start.

INSERTION OF XML DATA INTO THE TABLE USING XMLDOM

hello,
i am using XMLDOM to insert the data into the table
i am using different function of it.
but i am facing the problem to retrive the the multiple entry.
like in my example i have two entry of the ' po number '
& i am using the function
dbms_xmldom.item(l_nodelist, 0)
i which i have to pass index no.
& through this i am getting only single entry according to the index no.
Example on which i am working is
declare
l_xml_data CLOB;
l_xml_doc dbms_xmldom.domdocument;
l_nodelist dbms_xmldom.DOMNodeList;
l_node dbms_xmldom.domnode;
l_xmltype XMLTYPE;
l_po_num VARCHAR2(30);
l_cust_ord VARCHAR2(30);
l_item_code VARCHAR2(30);
begin
l_xml_data := '<?xml version="1.0" encoding="UTF-8"?>

<MobileInventoryResponse>
     <message>
          <message-header>
               <message-id>16244182</message-id>
               <transaction-name>ship-advice</transaction-name>
               <partner-name>cbeyond</partner-name>
               <source-url>http://www.brightpoint.com</source-url>
               <create-timestamp>20080826150709</create-timestamp>
               <response-request>1</response-request>
          </message-header>
          <ship-advice>
               <header>
                    <customer-id>297859</customer-id>
                    <shipment-information>
                         <ship-first-name>RA_13Aug_1</ship-first-name>
                         <ship-last-name>MIND</ship-last-name>
                         <ship-address1>test</ship-address1>
                         <ship-city>test</ship-city>
                         <ship-state>VA</ship-state>
                         <ship-post-code>22102-4931</ship-post-code>
                         <ship-country-code>US</ship-country-code>
<ship-phone1>0040726335068</ship-phone1>
<ship-email>[email protected]</ship-email>
<ship-via>FX01</ship-via>
<ship-request-date>20080826</ship-request-date>
<ship-request-warehouse>CBY1</ship-request-warehouse>
</shipment-information>
<purchase-order-information>
<purchase-order-number>380928</purchase-order-number>
<purchase-order-number>380929</purchase-order-number> ----modi by Ananda Dubey
<account-description/>
<purchase-order-amount>0.0</purchase-order-amount>
<currency-code>USD</currency-code>
</purchase-order-information>
<order-header>
<customer-order-number>0002759</customer-order-number>
<customer-order-date>20080826</customer-order-date>
<order-sub-total>19.0</order-sub-total>
<order-discount>0.0</order-discount>
<order-tax1>0.0</order-tax1>
<order-tax2>0.0</order-tax2>
<order-tax3>0.0</order-tax3>
<order-shipment-charge>18.0</order-shipment-charge>
<order-total-net>0.0</order-total-net>
<order-status>Completed</order-status>
<order-type/>
<brightpoint-order-number>35028788</brightpoint-order-number>
<warehouse-id>CBY1</warehouse-id>
<ship-date>20080826</ship-date>
</order-header>
</header>
<detail>
<line-item>
<line-no>1</line-no>
<item-code>SKU1</item-code>
<universal-product-code>0</universal-product-code>
<ship-quantity>1.0</ship-quantity>
<unit-of-measure>EA</unit-of-measure>
<serial-list>
<serial-numbers>
<esn>TIMI000013</esn>
</serial-numbers>
</serial-list>
<line-status/>
<base-price>0.0</base-price>
<line-discount>0.0</line-discount>
<line-tax1>0.0</line-tax1>
<line-tax2>0.0</line-tax2>
<line-tax3>0.0</line-tax3>
<bill-of-lading>929406733828</bill-of-lading>
<scac>FX01</scac>
</line-item>
</detail>
</ship-advice>
<transactionInfo>
               <eventID>16244182</eventID>
          </transactionInfo>
     </message>
</MobileInventoryResponse>';
l_xml_doc := dbms_xmldom.newDomDocument(l_xml_data);
-- Method 1 to get data
l_nodelist := dbms_xmldom.getelementsbytagname(l_xml_doc, 'purchase-order-number');
l_node := dbms_xmldom.item(l_nodelist, 0); -- gets first item from list
l_po_num := dbms_xmldom.getnodevalue(dbms_xmldom.getfirstchild(l_node));
dbms_output.put_line(l_po_num);
l_nodelist := dbms_xslprocessor.selectnodes(dbms_xmldom.makenode(l_xml_doc),
'/MobileInventoryResponse/message/ship-advice/detail/line-item/item-code');
l_node := dbms_xmldom.item(l_nodelist, 0); -- gets first item from list
l_item_code := dbms_xmldom.getnodevalue(dbms_xmldom.getfirstchild(l_node));
dbms_output.put_line(l_item_code);
l_xmltype := XMLTYPE(l_xml_data);
l_cust_ord := l_xmltype.extract('/MobileInventoryResponse/message/ship-advice/header/order-header/customer-order-number/text()').getStringVal();
dbms_output.put_line(l_cust_ord);
dbms_xmldom.freeDocument(l_xml_doc);
end;
/

In the following code
l_nodelist := dbms_xmldom.getelementsbytagname(l_xml_doc, 'purchase-order-number');
l_node := dbms_xmldom.item(l_nodelist, 0); -- gets first item from listYou need to understand what the second parm on the .item call does. See [dbms_xmldom.item|http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14258/d_xmldom.htm#i1126138]
The nodelist is a 0 based array of information and you are only requesting to pull the node info in the first array position. So to get the info in the second array position, you need to use ", 1)". You can also use .getLength and a loop to parse through everything in the node list.

ERROR WITH XMLDOM.WRITETOCLOB

Hi,
I have written the following procedure...............
procedure Prc_UpdateNodes(clString in out clob, nodename varchar2, nodeValue varchar2) is
     l_doc_node xmldom.DOMNode;
     l_xml_out varchar2(2000);
     begin
               domDocument      := Fn_GetXmlDocument(clString);
               nodeList           := xmldom.getElementsByTagName(domDocument, nodename);
               intNoOfNode      := xmldom.getLength(nodeList);
               l_doc_node := xmldom.item(nodeList, 0);
               for loopCount_int in 0.. intNoOfNode - 1 loop
     node := XmlDom.item(nodeList, loopCount_int);
                    valueNode := XmlDom.getFirstChild(node);
                    XmlDom.setNodeValue(valueNode,nodeValue);
               end loop;
               XmlDom.writeToClob(domDocument,clString);
end;
This below XML Is passed as the first parameter
The output after this operation for an xml should be..
<ROWSET>
<ROW>
<SITE_NO Key="Yes">1</SITE_NO>
<PARAMETER_VALUE_NO Key="Yes">0</PARAMETER_VALUE_NO>
<PARENT_PARAMETER_SPEC_NO>0</PARENT_PARAMETER_SPEC_NO>
<PARAMETER_SPEC_NO>1</PARAMETER_SPEC_NO>
<REF_TABLE_KEY_NO>0</REF_TABLE_KEY_NO>
<REF_TABLE_KEY_TYPE>S</REF_TABLE_KEY_TYPE>
<SEQ_NO>1</SEQ_NO>
<SUB_SEQN_NO>0</SUB_SEQN_NO>
<DATA_TYPE_NO>1</DATA_TYPE_NO>
<INTEGER_VALUE>0</INTEGER_VALUE>
<STRING_VALUE>00039</STRING_VALUE>
<NUMERIC_VALUE>0</NUMERIC_VALUE>
<DATE_VALUE>01/01/1600</DATE_VALUE>
<CREATED_ON>0</CREATED_ON>
<MODIFIED_ON>0</MODIFIED_ON></ROW>
<ROW>
<SITE_NO Key="Yes">1</SITE_NO>
<PARAMETER_VALUE_NO Key="Yes">0</PARAMETER_VALUE_NO>
<PARENT_PARAMETER_SPEC_NO>0</PARENT_PARAMETER_SPEC_NO>
<PARAMETER_SPEC_NO>2</PARAMETER_SPEC_NO>
<REF_TABLE_KEY_NO>0</REF_TABLE_KEY_NO>
<REF_TABLE_KEY_TYPE>S</REF_TABLE_KEY_TYPE>
<SEQ_NO>2</SEQ_NO><SUB_SEQN_NO>0</SUB_SEQN_NO>
<DATA_TYPE_NO>2</DATA_TYPE_NO>
<INTEGER_VALUE>0</INTEGER_VALUE>
<STRING_VALUE/>
<NUMERIC_VALUE>0</NUMERIC_VALUE>
<DATE_VALUE/>
<CREATED_ON>0</CREATED_ON>
<MODIFIED_ON>0</MODIFIED_ON>
</ROW>
</ROWSET>
But the output I get is ……….
<ROWSET>
<ROW>
<SITE_NO Key="Yes">1</SITE_NO>
<PARAMETER_VALUE_NO Key="Yes">0</PARAMETER_VALUE_NO>
<PARENT_PARAMETER_SPEC_NO>0</PARENT_PARAMETER_SPEC_NO>
<PARAMETER_SPEC_NO>1</PARAMETER_SPEC_NO>
<REF_TABLE_KEY_NO>0</REF_TABLE_KEY_NO>
<REF_TABLE_KEY_TYPE>S</REF_TABLE_KEY_TYPE>
<SEQ_NO>1</SEQ_NO>
<SUB_SEQN_NO>0</SUB_SEQN_NO>
<DATA_TYPE_NO>1</DATA_TYPE_NO>
<INTEGER_VALUE>0</INTEGER_VALUE>
<STRING_VALUE>00039</STRING_VALUE>
<NUMERIC_VALUE>0</NUMERIC_VALUE>
<DATE_VALUE>01/01/1600</DATE_VALUE>
<CREATED_ON>0</CREATED_ON>
<MODIFIED_ON>0</MODIFIED_ON></ROW>
<ROW>
<SITE_NO Key="Yes">1</SITE_NO>
<PARAMETER_VALUE_NO Key="Yes">0</PARAMETER_VALUE_NO>
<PARENT_PARAMETER_SPEC_NO>0</PARENT_PARAMETER_SPEC_NO>
<PARAMETER_SPEC_NO>2</PARAMETER_SPEC_NO>
<REF_TABLE_KEY_NO>0</REF_TABLE_KEY_NO>
<REF_TABLE_KEY_TYPE>S</REF_TABLE_KEY_TYPE>
<SEQ_NO>2</SEQ_NO><SUB_SEQN_NO>0</SUB_SEQN_NO>
<DATA_TYPE_NO>2</DATA_TYPE_NO>
<INTEGER_VALUE>0</INTEGER_VALUE>
<STRING_VALUE/>
<NUMERIC_VALUE>0</NUMERIC_VALUE>
<DATE_VALUE/>
<CREATED_ON>0</CREATED_ON>
<MODIFIED_ON>0</MODIFIED_ON>
</ROW>
</ROWSET>
0</MODIFIED_ON>
</ROW>
</ROWSET>
The letter in bold above are the newly introduced junk letters
I think this problem is due to the buffer isssues………….
Do we have any solution for clearing the buffer….
Also this error doesnt occur all the time.........
Sometimes it comes, Sometimes it works fine
The error is introduced during execution of the statement
XmlDom.writeToClob(domDocument,clString);
Can anyone help me on this front?????
Thanks
Message was edited by:
user609257

Thx for your response
the program looks like this:
I executed it from a Form;
PROCEDURE PRUEBA IS
xml clob;
vxml varchar2(100);
len integer := 1;
pos integer := 1;
f utl_file.file_type;
doc xmldom.DOMDocument;
ppal_node xmldom.DOMNode;
main_node xmldom.DOMNode;
root_node xmldom.DOMNode;
user_node xmldom.DOMNode;
tercer_node xmldom.DOMNode;
item_node xmldom.DOMNode;
ppal_elmt xmldom.DOMElement;
root_elmt xmldom.DOMElement;
item_elmt xmldom.DOMElement;
item_text xmldom.DOMText;
--Elementos de la cabecera
header_pi xmldom.DOMProcessingInstruction;
header_node xmldom.DOMNode;
CURSOR MODULOS IS
SELECT T_COAPI AS COAPI,
T_DNAPI AS DNAPI
FROM     T
WHERE     T_COAPI = 'PES';
BEGIN
--Creamos un nuevo DOM document
doc := xmldom.newDOMDocument();
ppal_node := xmldom.makeNode(doc);
FOR Modulo IN Modulos LOOP
-- El nodo es cada modulo
root_elmt := xmldom.createElement(doc, 'Modulo');
root_node := xmldom.appendChild(main_node, xmldom.makeNode(root_elmt));
--COAPI
item_elmt := xmldom.createElement(doc, lit_COAPI);
item_node := xmldom.appendChild(root_node, xmldom.makeNode(item_elmt));
item_text := xmldom.createTextNode(doc, Modulo.COAPI);
item_node := xmldom.appendChild(item_node, xmldom.makeNode(item_text));
--DNAPI
item_elmt := xmldom.createElement(doc, lit_DNAPI);
item_node := xmldom.appendChild(root_node, xmldom.makeNode(item_elmt));
item_text := xmldom.createTextNode(doc, Modulo.DNAPI);
item_node := xmldom.appendChild(item_node, xmldom.makeNode(item_text));
END LOOP
f := utl_file.fopen('TEMP_DIR','PRUEBA.xml','w');
--Obtenemos la longitud del xml lob
len := dbms_lob.getlength(xml);
DBMS_LOB.CreateTemporary(xml,TRUE);
xmldom.writeToClob(doc,xml); -- HERE I GET THE ERROR
dbms_xmldom.WriteToFile(f,xml);
WHILE (pos < len) LOOP
vxml := dbms_lob.substr(xml,100,pos);
utl_file.put(f,vxml);
pos := pos+100;
END LOOP;
utl_file.fclose(f);
END;

Using xmldom.writetoclob

Similar Messages

Maybe you are looking for