MK@MSITStore contains an invalid HTML tag

When I compile my master help project which contains many
merged projects I get the following message after compiling:
The file mk:@MSITStore:C:\[path of
project]\master.chm::/master.hhc contains an invalid HTML tag.
I have looked at the .hhc file and it looks fine. I do not
get this message when compiling and running any of the
sub-projects. Any ideas?
I am running Windows Vista and RoboHelp 7

Hi,
I have tried to recompile everything and readd it. I have
tried adding the sub-projects at different levels within the book
(as seen in a post back in September). I have tried deleting the
.cpd and .hhc files and recreating them. Nothing works.
I am not sure if this is a clue, but when opening the
compiled project, I get the error message and then the main topic
displays. The TOC entries do not work, and when right-cicking on
the main topc in the right pane and selecting properties the
address displays as:
mk:@MSITStore:C:\HTML%20Help\WW1\Compiled\Master.chm::/Help_Intro.htm
Thanks in advance for any insight!

Similar Messages

Invalid HTML tag - Is this true? Child .chm files cannot have underscores in name?

I’ve been having the following problem:
file mk:@MSITStore: … xxx.chm::/xxx.hhc contains an invalid HTML tag
In more detail what happens:
You click on the .chm file to open it and you get a dialog with an error message:
The file mk:@MSITStore:E:\p4_work\1663\isentris\isentris\end-user\trunk\src\Symyx_Isentris\!SSL!\M icrosoft_HTML_Help\Symyx_Isentris.chm::/Symyx_Isentris.hhc contains an invalid HTML tag
There is an OK button and when you click that, then the .chm file opens just fine.
All child .chms open fine when I click on them.
This merged help project worked fine until I upgraded from RH X5 to RH7. I hadn’t changed anything. Then I started reading the forum (RoboHelp HTML Help).
What I’ve done so far:·         Upgraded all help projects from RH X5 to HR7.
·         Regenerated all .chm files, added them again to parent help project’s TOC. Child .chm files open fine.
·         Regenerated a new parent .chm.
So, I see that one of Colum's answers in another thread was to make sure that none of the child .chms had underscores in the title.
>>>>>>>>>> !!!!!!!!   you mean that RoboHelp now does not accept underscores in .chm titles? Please don’t let it be so.
Also, several people said that they solved the problem like this:
·         “You will get this message if your merged CHMs are not in a book in your merged TOC. Check that. “ You mean I have to associate a book icon with each child .chm in the parent’s TOC?
·         My guess is that the merged project should be placed inside a book in the main project's TOC. You'll just have to rearrange the subprojects TOCs so you don't have books inside books.
What does “in a book” mean? See my attached jpegs. The child project’s TOC is in book form … within the child project. However, in the parent project’s TOC, the child projects are shown with that merged symbol … kinda looks like a very angular G. Should I create books that link to destinations? Is that what I should do in the parent project’s TOC? Somehow that doesn’t look like it’s going to work – I looked at both dialogs – link and no link – neither one looked like it would work.
Thanks so much for your help!!

When you have a directory path in the TOC, that means that the directory structure should match. So for example:
TOC (in parent.chm)
!SSL!\Microsoft_HTML_Help\symyx_client.chm
symyx_for_excel.chm
End folder structure: (+ for folder, - for file)
+ Help
   - parent.chm
   - symyx_for_excel.chm
     + !SSL!
        + Microsoft_HTML_Help
           - symyx_client.chm
Now if the symyx_client file shows up okay after the error message, possibly the parent chm file is smart enough to correct for the directory structure not being there, but a file with the correct name existing in the same directory as the parent.
We've always tried to ensure out merge references don't have a path - RH has the annoying habit of requiring the chm file to be copied into the root, but we just delete it after the first build, as it shouldn't need to be there for the actual compile process. In fact, we've had problems with the chm being physically built into our parent chm, drastically increasing the file size and resulting in duplicate and outdated topics in search. We leave the file there for the first build as sometimes the merge doesn't work first go without it; I think it has to do with how it gets added to the hhp file.
To test that our merge works, we just copy a couple of the files into the same folder (e.g. c:\temp\test) to double-check. We need this process as our clients get different combinations of the total chm files built into our parent.
Something else I've just thought of is maybe there is something strange happening in the actual hhc file. Do you have any topics in that TOC with characters other than a-z, 0-9 and underscore in the filename (in the topic title should be okay, but maybe some particularly odd ones might cause an issue)?
Amber

Handle invalid html tag in the htmlText

Hello,
I"ve a <mx:text control which takes html text, it is a chat application, so, the client will simply concatinate the text comming through the network. But if the html text has invalid tags, it does not show the rest of the chat messages.
I get something like
<a href="http://google.com"">http://google.com<a>
here the red text is creating problems.
P:S: you might want to say, handle it when sending from other client, but, we are dealing with legacy data, which has a lot of html format issues.

Thanks for responding
I understand that adding more HTML tags will change the font size. What I don't get is why it changed from one size to another by adding a HTML tag in the first place. I figured that it should have retained the same size it was without the tag when I added the tag. If I had added a font tag I would have excpected the font to change size but I added a href.
Sorry for the confusion.

When Exported in PDF reports displays HTML Tags

Hi All,
Business Object XI R2
WebI Report
I am Reporting on Oracle CLOBs. This CLOB contains Data with HTML Tags, Clearly this CLOB is the combination of DATA formatted in HTML tags.
When I create a WebI Reports on this (CLOB) WebI displays correctly in the sence data only.
When I export this report to PDF; The PDF display contains both data with HTML tags as well.
Is there any solution/workaround or any setting in the BO Server to overcome/solve this issue.
Appreciated you prompt help.
Thanks,
Ashok

Christine,
My question is wether Adobe support CLOBs or not.
I was reported on oracle CLOBs (This CLOBs contains data wrapped in HTML tags), by using Business Object Web Intelligence. This WebI displaying the reports as I desire, there is no problem, in WebI. WebI displaying data with no HTML Tags.
My problem is that when I export this WebI report, this Adobe is getting HTML tags, why? Since this WebI is not displaying the HTML tags, since adobe does not know wether there is a HTML tags or not.
Apprecited your prompt reply.
Thanks,
Ash

Smartform with text containing html tags

I am reading from a table that contains html tags like
<br>
<menu>
<LI type="disc">This is a line which contains html tags
that goes over to 2nd line
<LI type="disc"> This is also another line which contains html
tags.
and each line is stored in a different row. The customer has agreed to use only 3 of the above html tags (
<br>, <menu> and <LI type="disc">
When the output is previewed in smartforms, the html tags needs to be formatted and not print in raw format as follows:
<LI type="disc">This is a line which contains html tags
that goes over to 2nd line
<LI type="disc"> This is also another line which contains html
tags.
How can this be done ?

ok understood...
you can try this
Loop at i_html into wa_html.
tmp = strlen( wa_html-line).
do tmp times.
if wa_html-line+0(1) = <
l_start = 1.
__elseif wa_html-line+0(1) = >
l_start = 0.
concatenate l_char wa_html-line+0(1) into l_char.
endif.
if l_start = 1.
concatenate l_char wa_html-line+0(1) into l_char.
wa_html-line = wa_html-line+1(l_tmp - sy-index) OR use the keyword SHIFT LEFT...plz chk the syntax
elseif l_start = 0 and l_char = <LItype=disc>
concatenate <BULLET> wa_html-line into wa_html-line separated by space.
endif.
enddo.
modify i_html form wa_html.
endloop.
now u define two text element one with bullet and another without bullet. this is how u insert bullet:
goto to the texteditor of a text element in smartform. Its a small button on the left hand side of.
Then goto insert>characters>sap symbols.
for ur requirement u can use the symbol SYM_FILLED_CIRCLE.
Remember these icons or symbols whatever u will not be able to see in print preview....u have to take a print
and see
Now at the time of display check if the line contains the
tag <BULLET> ( if wa_html-line+0(8) = '<BULLET>')
if the line has the tag then display ur text in the text element with bullet
otherwise
display it in the non bullet text element..
remeber to remove the tag before dispalying
hope this solves ur prob...

Contains HTML tags & print

Hi,
Reports6i patch10
Scenario is: I have a column (database colummn text with embedded HTML ) where I set property 'Contains HTML tags'
I get the column incl html rendered find where desformat=html&destype=file&desname=myfile.html '
but when setting
destype=printer&desformat=html&desname=myqueue and sending output straight to printerqueue the html is not being rendered in fact the column is not rendered at all
HAs anyone similar experience ? solutions ?
TIA
Per

Hello Per Elmgreen,
When DESTYPE is specified as PRINTER, Reports formats output in PS or PCL format (depending on the printer language understood by the destination printer). Since these formats do not understand HTML tags, any such markup in the output will remain unformatted.
In order to get properly formatted output using HTML tags, you could generate the report to a HTML file (with the 'Contains HTML tags' property set) and then print the file using a browser.
Thanks,
The Oracle Reports Team.

Reports does not show data which contains HTML tags

Dear Gurus
Im trying to resolve this problem but im really stuck in it. I have the requirment to customize PO reports in R12 and at the end of every report we have to show its Terms and Conditions.I have the following query wich brings me the data i need.
SELECT
otu.document_number po_segment1,
otu.document_id,
otta.template_id,
otta.template_name,
otta.description,
otta.intent,
otta.status_code,
otta.start_date,
otta.end_date,
otta.instruction_text,
otu.document_type,
otu.document_number po_segment1,
otu.document_id po_header_id,
osb.HEADING titulo_seccion,
oav.display_name,
oav.article_text article_text
FROM
apps.okc_terms_templates_all otta,
apps.okc_template_usages otu,
apps.okc_k_articles_b okab,
apps.okc_sections_b osb,
apps.okc_articles_all oaa,
apps.okc_article_versions oav
WHERE otta.template_id = otu.template_id
AND otu.document_id = okab.document_id
AND osb.scn_id(+) = okab.scn_id
AND okab.article_version_id = oav.article_version_id
AND oaa.article_id = oav.article_id
AND otu.document_number='21000000111' -- Purchase Document
AND otu.DOCUMENT_TYPE NOT IN ('RFQ_RESPONSE', 'RFQ','RFI_RESPONSE','RFI','AUCTION_RESPONSE','AUCTION')
The problem is the field okc_article_versions.article_text storage data with HTML tags as an example : ''<P>COMPANY will make it’s best effort to provide an accurate forecast to CONTRACTOR, <B><I>at the end of every quarter,</I></B> however, a forecast in any form, shall not represent a firm commitment by COMPANY.,</P>"
So when i tried to show the information my .rtf template nothing comes out. Does anybody knows whats happenig and how to resolve it??
Please, resolve this issue is more than urgent an i guess somebody have the answe Thanks a lot
Mily

Hi ,
Currently image url have value http://serverip:port/reports/rwservlet
I don't know is this correct? Suggest What value it should have?
Thanks
Vinod

Want text input containing HTML tags to appear as HTML in output format

Hi,
We have a table in Oracle database that has a column named detail,one of its values is like this: <bold><italics>Good Morning</italics></bold>. What our client wants is that the output format should show: <b><i>Good Morning</b></i>. That is,Bi Publisher should be able to parse the HTML tags and provide the desired output. Please tell me how to achieve this. Any help is much appreciated.
Thanks and regards,
Debarati,
[email protected]

Hi,
have a look here (http://blogs.oracle.com/xmlpublisher/2007/01/formatting_html_with_templates.html) to get an idea.
regards
Rainer

Pulling Data Containing HTML TAGS

We have a DB application that allows users to freely enter data. Our problem is that if a users enters HTML tags as their data and submits it to our DB, we are able to encode the response and store the data in it's original context (a requirement).
Now, here is the problem. When the data is pulled out from the DB and displayed on the HTML page, if the user had submitted HTML tags, our page doesn't display the HTML tags but rather their browser interprets the tag.
We do not want the browser to interpret the tag, just want the actual characters displayed.
so if a user enteres <B> HELLO WORLD</B>
we want the less than and greater than signs to be displayed..not the phrase HELLO WORLD to be displayed in bold.
Any help is appreicated.

Darn,
When you pull the string from the DB, you would be able to pass it through a simple 'filter' method that converts important HTML characters to the display element equivelents.
For example, you would want to change all < tags to < all & to & and quotes to " This might just about do everything you would need.
It might be as easy as doing a few
dbString = dbString.replaceAll("<", "<");
etc...

Method to Format a string column containing HTML tags as simple string.

Hello,
I am working with formating a string column which holds Html tags.
I want to remove these tags from the actual data which has to be shown on the BI Publisher report.
Can you suggest how can we format this in the DataSet designer of BI publisher so that my data is recognised on the Layout designer.
I have been trying to create an expression using the "Add Element by Expression" option to format this html tag data as normal string without the tags.
Can you suggest if this is the correct method to do this.
I found this below code being used in an existing DateSet but i am not able to recreate a similar formating on this kind of data column.
<![CDATA' || '['|| TO_CLOB(SUCCESS_CRITERIA) || ']' || ']>
Kindly suggest if you have any idea on the above mentioned issue.
Thanks,
Shweta

And read this:
Navigate yourself around pitfalls related to the Runtime.exec() method
http://www.javaworld.com/javaworld/jw-12-2000/jw-1229-traps.html

Proprietary tags - invalid HTML

Does anyone know why RoboHelp adds <spaces> tags and what are they used for? Is there a way to prevent RoboHelp from adding it? Since they're not valid HTML tags, the topics created using RoboHelp really aren't compliant. Also, I've never seen the CSS attribute "x-right" and "x-cell" either. Unfortunately Google doesn't have any useful information on these tags.
Any information would be helpful.
Thanks!

Hi,
These attributes and tags are just RoboHelps way of maintaining files, and don't appear in the output.
If I recall correctly, RoboHelp inserts the spaces tag whenever you add a couple of spaces in the WYSIWYG editor. If you don't want that, insert some non-breakable spaces (& nbsp;) in html mode. (without the space though, but if I add the html character, the forum turns it into a non-breakable space :-( )
The attributes x-right, x-cell, x-save-position, x-use-null-cells, x-next-tag, and so on help RoboHelp position elements and control how RoboHelp maintains them. Example:
H1 {
x-next-class: Body-Text;
x-next-type: P;
in the definition of a style, makes that the next paragraph after a header is a standard paragraph with the class Body-Text. You can remove these from your style sheet, but RoboHelp may put them back whenever you edit the css trough the built-in css editor.
I wouldn't worry about the x- commands in topics, as there removed when you generate your output, and RoboHelp just adds them again when you remove them in html mode. The only way to get rid of them is to edit the html files in an external editor.
Greets,
Willam

HTML Tags in sql

Hi guys
Im executing the below sql with the html tags, as it gone mess and returning error like invalid table name.
SELECT '<p style="font-family: Script MT Bold;color: #800080;font-size:35px;text-align: center;">' || 'Welcome' || PAPF.FULL_NAME|| '</p>
<p style="font-family: Times New Roman;color: #000000;font-size:15px;text-align: center;"> The Employees Information of</p>
<p style="font-family: Times New Roman;color: #000000;font-size:20px;text-align: center;"> <strong>Central Bank of Kuwait</strong>'
FROM FROM PER_ALL_PEOPLE_f PAPF, FND_USER FU
WHERE PAPF.PERSON_ID = FU.EMPLOYEE_ID AND
TRUNC(SYSDATE) BETWEEN PAPF.EFFECTIVE_START_DATE AND
PAPF.EFFECTIVE_END_DATE
Im not sure whether i missed any || or quotes in html tags while using with the column name in the select statement
Thanks in advance.
Regards,
Vel

Don't do this via the SQL selection as hardcoded HTML??
I think the following 2 approaches far superior.
If you need the actual SQL projection to contain HTML, implement user functions for data transformation. E.g.
SQL> create or replace function MakeDiv(
2          divBody varchar2,
3          fontColor varchar2 default 'black',
4          tooltip varchar2 default null )
5 return varchar2 is
6          DIV_TEMPLATE constant varchar2(4000) :=
7          '<div style="color:$COLOR" title="$TOOLTIP"> $BODY </div>';
8
9          htmlDiv varchar2(4000);
10 begin
11          htmlDiv := replace( DIV_TEMPLATE, '$BODY', divBody );
12          htmlDiv := replace( htmlDiv, '$COLOR', fontColor );
13          htmlDiv := replace( htmlDiv, '$TOOLTIP', tooltip );
14          return( htmlDiv );
15 end;
16 /
Function created.
SQL>
SQL> select MakeDiv( ename, 'blue', 'Employee '||empno ) from emp;
MAKEDIV(ENAME,'BLUE','EMPLOYEE'||EMPNO)
<div style="color:blue" title="Employee 7369"> SMITH </div>
<div style="color:blue" title="Employee 7499"> ALLEN </div>
<div style="color:blue" title="Employee 7521"> WARD </div>
<div style="color:blue" title="Employee 7566"> JONES </div>
<div style="color:blue" title="Employee 7654"> MARTIN </div>
<div style="color:blue" title="Employee 7698"> BLAKE </div>
<div style="color:blue" title="Employee 7782"> CLARK </div>
<div style="color:blue" title="Employee 7788"> SCOTT </div>
<div style="color:blue" title="Employee 7839"> KING </div>
<div style="color:blue" title="Employee 7844"> TURNER </div>
<div style="color:blue" title="Employee 7876"> ADAMS </div>
<div style="color:blue" title="Employee 7900"> JAMES </div>
<div style="color:blue" title="Employee 7902"> FORD </div>
<div style="color:blue" title="Employee 7934"> MILLER </div>
14 rows selected.
SQL> An even better approach is not to do this in the SQL projection - but instead have a PL/SQL framework that accepts a SQL, and turns the SQL projection of that SQL into a HTML report. Where this framework can use different templates to generate different types of HTML reports.
This approach is very successfully used by Oracle Apex.

Filting html tags, css, and javascript with regex

hi everyone,
im writing a small application where a user types in a url, and the text of the webpage is displayed in a text area.
ive got it to work, however it takes some time, and also alot of content i dont want is displayed - tags, scripts and sometimes css.
initally i filtered out the html tags with a regular expression, but i still get alot of unwanted content.
im not a confident java programmer, and the idea of parsing html, css and javascript is the scariest idea ever to me, so my next idea is to keep only everything between the <body> tags - everything above and below it is deleted - hopefully that should leave me only with the visible content on the site.
ive messed around with regular expressions but i cant get it to work, can anyone help out?
thanks alot,
Torre

Darryl.Burke wrote:
I tried out the regexes I posted on the source of a forum page, which is not valid html (contains two each opening and closing body tags). With a bit of trial and error I was able to remove everything upto the first, and not the second, opening tag by using a reluctant qualifier, ^.*?, but couldn't for the life of me achieve removal of only the last closing tag, leaving the other, invalid one intact. How would you do that?Regexes always try to match the first occurrence of whatever they're looking for (the sentinel), and there's no way to change that behavior (but it would be handy if you could). What you have to do instead is make sure the rest of the regex can't match the sentinel. For that you need lookahead, and the simplest way to use it is to scan the rest of the text looking for the sentinel and, if it doesn't find one, go ahead and gobble up the remaining text: "(?is)</body>(?!.*</body).*$" However, if there are many occurrences of the sentinel, you could take a serious performance hit. Here's a much more efficient way: "(?is)</body>(?:[^<]++|<(?!/body>))*+$" After matching the sentinel, this regex gobbles up anything that's not the first character of the sentinel, or the first character as long as it isn't followed by the remaining characters of the sentinel. The advantages of this regex are that it never has to backtrack, and the lookahead is only applied when it's necessary, where the first regex applies it every time.

Query to extract HTML tag with data

Hi All,
I have a string.
'<HTML><HEAD>THIS IS HEAD.</HEAD><BODY>THIS IS BODY.<P>THIS IS P1.</P>NIMISH<P>THIS IS P2.</P></BODY></HTML>'
I want to extract a html tag including its opening & closing tab with data as
if i say P1
then the output should be
'<P>THIS IS P1.</P>'
for P2
then the output should be
<P>THIS IS P2.</P>
please help me in writing this query with regular expression
i have tried it as following but it is not giving desired result:
WITH T AS
SELECT
    '<HTML><HEAD>THIS IS HEAD.</HEAD><BODY>THIS IS BODY.<P>THIS IS P1.</P>NIMISH<P>THIS IS P2.</P></BODY></HTML>' STR
FROM
    DUAL
SELECT REGEXP_SUBSTR(STR, '<P>.+P2.+</P>') FROM T
Thanks & Regards
Nimish GargEdited by: Nimish Garg on May 7, 2012 5:49 PM

Nimish Garg wrote:
My requirement is to extract a <tag>data</tag> from a HTML/XML string
where data contains any specified value.HTML is not XML.
And that is a critical distinction to make. HTML parsing is horribly complex. XML is quite easy. For HTML you have to code your own parser in PL/SQL. XML can be parsed using the XMLTYPE class/data type in PL/SQL.
So if you need to find a single specific tag in HTML - I would not try to treat it as XML. I may not even try to use regular expressions.
I would do a basic substring search for the start of the tag. Read the data following the tag. Ensure that there are no nested or embedded tags in the data. Until the end tag is read. Because HTML is that much abused - and because that is an accepted norm as parsers used by browsers deals with that abuse without complaining.
Proper HTML is mostly a myth in my experience of "screen scraping" web servers for data extraction as they do not have web services supplying the data.

How can I eliminate HTML tags from Oracle Text Snippet?

I perform a search on many tables and on many columns of those tables.
Some of those columns are VARCHAR2 and some CLOB.
Also, some of the searchable data are HTML and some are plain text.
My problem is that ctx_doc.snippet fetches the HTML tags.
For example I get this, as a snippet result in one of my searches: Qual Germany n1 <p>Test Qual Germany n1</p>
I want the result to be fetched without the HTML tags.
In my index configuration I have used NULL FILTER and HTML_SECTION_GROUP.With that configuration I managed to eliminate the HTML tags but not in all cases!
For example:
I search table CONTENTS columns TITLE(VARCHAR2) and MAIN_TEXT(CLOB)
I created the following procedure that concatenates the two columns:
CREATE OR REPLACE PROCEDURE CONTENTS_PROC( p_id in rowid, p_lob IN OUT clob)
IS
BEGIN
FOR c1 IN (SELECT main_text||' '||title data FROM contents WHERE ROWID = p_id)
LOOP
dbms_lob.copy( p_lob, c1.data,
dbms_lob.getlength( c1.data ));
END LOOP;
END;
I created a user Datastore:
BEGIN
ctx_ddl.create_preference( 'content_trans_datastore', 'user_datastore' );
ctx_ddl.set_attribute( 'content_trans_datastore', 'procedure', 'CONTENTS_PROC' );
END;
and finally I create the index:
CREATE INDEX content_trans_ot_idx ON contents(ORACLE_TEXT_COLUMN)
INDEXTYPE IS ctxsys.CONTEXT PARAMETERS ('datastore content_trans_datastore SYNC(ON COMMIT) STORAGE INDEX_STORAGE filter ctxsys.null_filter section group ctxsys.html_section_group');
When I perform the search on those data: <p> <strong>Test Doc-Test </strong> </p> the snippet I get is: Test Doc-Test.
That's fine, the html tags are removed!
In another case I search table NCP columns NAME(VARCHAR2) and BODY(VARCHAR2)
I created the following procedure that concatenates the two columns:
CREATE OR REPLACE PROCEDURE NCP_PROC( p_id in rowid, p_lob IN OUT clob)
IS
BEGIN
FOR c1 IN (SELECT name||' '||body data FROM ncp WHERE ROWID = p_id)
LOOP
dbms_lob.copy( p_lob, c1.data,
dbms_lob.getlength( c1.data ));
END LOOP;
END;
I created a user Datastore:
BEGIN
ctx_ddl.create_preference( 'ncp_trans_datastore', 'user_datastore' );
ctx_ddl.set_attribute( 'ncp_trans_datastore', 'procedure', 'NCP_PROC' );
END;
and finally I create the index:
CREATE INDEX ncp_trans_ot_idx ON ncp(ORACLE_TEXT_COLUMN)
INDEXTYPE IS ctxsys.CONTEXT PARAMETERS('datastore ncp_trans_datastore SYNC(ON COMMIT) STORAGE INDEX_STORAGE filter ctxsys.null_filter section group ctxsys.html_section_group');
When I perform the search on those data: test <strong> </strong>http://deleteme.com the snippet I get is: test <strong> </strong>http://deleteme.com!!!!!!!!!!
How is this possible? Why in the first case the HTML tags are eliminated and in the second case they are not?
Thanks,
Margarita
Edited by: user13312701 on 07-Sep-2010 08:51

Doing various tests I found out that the problem is when I need to search in multiple columns of a table.
That is when I create a user_datastore that uses a procedure that concatenates the columns.
And especially when the data with the html tags is in a VARCHAR2 column.
e.g
--create the table*
CREATE TABLE CONTENT_TRANS (content_trans_id NUMBER,
main_text CLOB,
title vARCHAR2(2000),
oracle_text_column VARCHAR2(1));
alter table "CONTENT_TRANS" add constraint CONTENT_PK primary key("CONTENT_TRANS_ID") ;
--Insert dummy data*
Insert into CONTENT_TRANS
(CONTENT_TRANS_ID,MAIN_TEXT,TITLE)
values
(1,'lorem','lorem <p>qualification</p> 2.1 ');
Insert into CONTENT_TRANS
(CONTENT_TRANS_ID,MAIN_TEXT,TITLE)
values
(2,'lorem','lorem <br>qualification</br> 2.1 ');
--CREATE THE procedure that concatenates main_text(CLOB) and title(VARCHAR2)*
CREATE OR REPLACE PROCEDURE CONTENT_TRANS_PROC( p_id in rowid, p_lob IN OUT clob)
IS
BEGIN
FOR c1 IN (SELECT main_text||' '||title data FROM content_trans WHERE ROWID = p_id)
LOOP
dbms_lob.copy( p_lob, c1.data,
dbms_lob.getlength( c1.data ));
END LOOP;
END;
--Create the user datastore*
BEGIN
ctx_ddl.create_preference( 'content_trans_datastore', 'user_datastore' );
ctx_ddl.set_attribute( 'content_trans_datastore', 'procedure', 'CONTENT_TRANS_PROC' );
END;
--Create the index*
CREATE INDEX content_trans_ot_idx ON content_trans(ORACLE_TEXT_COLUMN)
INDEXTYPE IS ctxsys.CONTEXT PARAMETERS ('datastore content_trans_datastore SYNC(ON COMMIT) filter ctxsys.null_filter section group ctxsys.html_section_group');
exec ctx_doc.set_key_type('PRIMARY_KEY');
--Perform the query
SELECT SCORE(1),ct.content_trans_id, ctx_doc.snippet('content_trans_ot_idx', ct.content_trans_id, 'lorem') as snippet
from content_trans ct
where contains(ct.ORACLE_TEXT_COLUMN, 'lorem', 1) > 1;
Results WITH NOT WANTED HTML TAGS:
6 1 <b>lorem</b> <b>lorem</b> <p>qualification</p> 2.1
6 2 <b>lorem</b> <b>lorem</b> <br>qualification</br> 2.1
Edited by: user13312701 on 13-Oct-2010 01:18

MK@MSITStore contains an invalid HTML tag

Similar Messages

Maybe you are looking for