Conversion of several lines (with 1 column) to 1 row of concatenated values

Similar Questions

• SQL to convert a row with 4 columns to 4 rows with 1 column

  Sorry, I know this question has been asked hundreds of times before!

  This is my starting SQL:

  set linesize 500
  
  with tbl_data AS
   (select 'N' argument1, 'Y' argument2, NULL argument3, 'Y' argument4 from dual)
  select * from tbl_data;
  
  
  ARGUMENT1 ARGUMENT2 ARGUMENT3 ARGUMENT4
  --------- --------- --------- ---------
  N         Y                   Y        
  

  Is it possible to enter the data in this format:

                  COL_HEADING
  ARGUMENT1       N
  ARGUMENT2       Y
  ARGUMENT3       
  ARGUMENT4       Y
  

  I have read comments UNPIVOT but can't really work on the syntax to get a single line with 4 columns to appear as 4 rows with 1 column.

  Sorry - I realize that I'm probably be lazy and stupid.

  Any advice much appreciated, thank you.

  Hello

  Here's a way to do it with UNPIVOT:

  SELECT *.

  OF tbl_data

  MUST INCLUDE NULL VALUES

  (col_heading

  FOR column-name IN (argument1

  argument2

  argument3

  argument4

  )

  )

  ;

  Output:

  COLUMN COL_HEADING

  --------- -----------

  ARGUMENT1 N

  ARGUMENT2 Y

  ARGUMENT3

  ARGUMENT4 Y

• Get several lines with the request - please help

  Hello
  
  I have query that gives me the output below.

  select distinct a.*,cu1.usr_key,cu1.first_nm,cu1.last_nm
  from(
        select ng.grp_key, ng.grp_nm , gt.grp_typ_nm, stts.nm as grp_sts, ng.expiry_date, ng.updt_dttm, stts.stts_key
        from new_group ng, user_group ug, group_type gt, status stts
        where ug.grp_key = ng.grp_key and ng.grp_typ_key = gt.grp_typ_key and 
        ng.stts_key = stts.stts_key and
        ug.usr_grp_rl_typ_key = 1 and 
        lower(ng.grp_nm) like 'test foe%'
        union 
        select ng.grp_key, ng.grp_nm , gt.grp_typ_nm, sts.nm as grp_sts, ng.expiry_date, ng.updt_dttm, sts.stts_key
        from new_group ng, group_type gt, status sts
        where ng.grp_typ_key = gt.grp_typ_key and ng.stts_key = sts.stts_key and 
        lower(ng.grp_nm) like 'test foe%')a, common_user cu1, user_group ug1
  where cu1.stts_key = a.stts_key and cu1.usr_key = ug1.usr_key and ug1.usr_grp_rl_typ_key = 1 and ug1.grp_key(+) = a.grp_key;

  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 09.38.34 29742 Sam Saed
  18345 enemy test remove Group organization DELETED 9 February 12 09.38.34 AM 3 29643 dummyName514 dummy
  18345 enemy test remove Group organization DELETED February 9, 12 09.38.34 AM 3 28917 TestMObileUser Gujral
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 27284 Rocky jegou 09.38.34
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 28920 Bhavani Gujral 09.38.34
  18345 enemy test remove Group organization DELETED 9 February 12 09.38.34 AM 3 29645 dummyFName516 dummy
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 27316 Karthik Gilani 09.38.34
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 28147 Bowl David Mathews 09.38.34
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 09.38.34 29731 Sam Saed
  18345 enemy test remove Group organization DELETED February 9, 12 AM 3 28919 Madhu mango 09.38.34
  
  For a group 18345, there are 10 users and this is the reason why I get a multiple user records.
  
  I can't avoid these last three columns, because I have to show these values in the application.
  
  In my opinion, the UNION is at the origin of the problem. Can I replace the UNION here? Please share your ideas on this.
  
  Thank you
  Rambeau
  
  Published by: ouali on February 9, 2012 22:35

  Hello

  because I have to show these values in the application.

  What is your problem then? If you simply issue a SELECT statement to application code or send the data back to REF CURSOR, it will be quite OK to have several lines. That's what the SELECT statements are for: to return sets of results with several lines.
  So, could not explain your problem more in detail please? If you get an error somewhere, provide the code causing the error and the error message (it is probably ORA-001422 exact extraction returns several lines somewhere in PL: / SQL, isn't it? then post here).

  In my opinion, the UNION is at the origin of the problem. Can I replace the UNION here? Please share your ideas on this.

  Us don't know not the answer to this question, because no one here knows your needs (what you need).

• Failed to insert a new line with a column generated sequence id

  Hello
  
  I'm trying to insert a new row in a table containing a column of sequence according to db, used as the primary key.
  
  Here is the code for the custom method of the corresponding AM I use to insert the new line:
  
  Line row2 = getCaAppointments1 () .createRow ();
  row2.SetAttribute ("Member ID", Member ID);
  row2.SetAttribute ("Status", "Planned");
  row2.SetAttribute ("CreatedBy", coachId);
  row2.SetAttribute ("CreationDate", date);
  AppointmentId will be generated by the sequence
  DBSequence newId = row2.getAttribute("AppointmentId") (DBSequence);
  SequenceImpl s = new SequenceImpl ("XX_CA_APPOINTMENT_ID", getDBTransaction());
  Long next = newId.getSequenceNumber () .longValue ();
  System.out.println ("Got sequence as id:" + newId);
  row2.SetAttribute ("AppointmentId", newId);
  row2.SetAttribute ("CoachId", coachId);
  row2.SetAttribute ("Type", "0");
  row2.SetAttribute ("AppointmentNumber", 0);
  row2.SetAttribute ("PlannedDate", Department of justice);
  getCaAppointments1 (.insertRow (row2));
  
  I put the attribute EO 'AppointmentId' to use a DBSequence. I tried different combinations to get the sequence Id and insert a line. Nothing worked well and I'm unable to insert a new line.
  
  If I avoid using the attribute 'AppointmentId' and that you try to insert a row, it will fail with the error below:
  
  ORA-01400: cannot insert NULL into ('APPS'. "" "" XX_CA_APPOINTMENTS '. "" APPOINTMENT_ID") ORA-06512: at line 1
  
  Tried to insert the new row using BCTester but I am struck by the same error it (I was unable to change the value of the "AppointmentId" column when inserting the line because this column is read-only and it showed as - 2 value).
  
  I guess, since the attribute of column/is already marked to use a DBSequence, the value should be assigned by default and I shouldn't hit "cannot insert the value NULL to the column.
  Did I miss something necessary to trigger the sequence to automatically generate new Id and assign it to the new line before it is created?
  
  Please let me know the fix.
  
  Thanks in advance
  Rouhaud

  Hello

  Oracle changes this behavior each version, so I hope that its still OK.
  If you have a trigger on the table that inserts the value of the id if none is preset, then mark it as DBSequence, or be it the number.

  Now for your entity, just generate the impl class and override the method create, kind of old school, I know, but it never fails.
  In the create just add the code to your sequence you had:

  SequenceImpl seqImpl = new SequenceImpl("", thisTransaction);
  Number idValue= s.getSequenceNumber();
  setAttribute("Id", idValue);
  

  -Anton

• Insert several lines with dynamic data

  Hello
  When I do an insert into a table that has several lines of dynamic information. on
  a page, it inserts all rows that are dynamically filled. I want only what he
  to insert the lines that I choose the initials on. How can I do? Let's say I
  have 5 items that are displayed on a page and I want to insert only original
  the first 2 because the last 3 are left blank, how should I do this? I tried
  implementation of an if statement that says:
  
  < cfif "form. "Help of # PNR_Approval_Initials #" NEQ "" > "".
  
  Then make the insertion in the table
  
  < cfelse >
  < / cfif >
  
  I have this game were also up in a loop that says:
  
  < cfloop index = "help" list = "" #form.listofids # "delimiters =",">"
  
  Here is the code for just the insertion. I also have a code here
  Update another table. I got this to work, so I just try to get the insertion
  works fine for the moment. I can not use this where the query statement:
  Where ItemID = #id # because I'm trying to insert into the table of approval, not
  the items table. The table is just updated with some other info. in
  a few other input boxes.
  
  < cfloop index = "help" list = "" #form.listofids # "delimiters =",">"
  < cfif "form. "Help of # PNR_Approval_Initials #" NEQ "" > "".
  
  < cfquery Datasource = '#application. DataSource #">"
  Insert into accreditation (APV_ItemID,
  APV_ECID,
  Document_Type,
  Approval_Initials)
  
  Values (' #Evaluate ("form.) (' ItemID # help # ") #',
  ' #Evaluate ("form. ("ECID # help #") #',
  < cfif isDefined (form '. ') ("Help of # PNR_Doc_Type # ') and
  "form. "Help of # PNR_Doc_Type #" NEQ "" > "".
  ' #Evaluate ("form. ("Help of # PNR_Doc_Type # ') #',
  < cfelse >
  NULL,
  < / cfif >
  
  < cfif isDefined (form '. ') ("Help of # PNR_Approval_Initials # ') and
  "form. "Help of # PNR_Approval_Initials #" NEQ "" > "".
  ' #Evaluate ("form. (Aide de # PNR_Approval_Initials #») #
  < cfelse >
  Null value
  (< / cfif >)
  
  < / cfquery >
  
  < / cfif >
  < / cfloop >
  
  On the page display, just a table of dynamic info. Here are the 4
  things that I have on this page that needs to be inserted to the database.
  
  < input type = "hidden" name = "' ECID #ItemID #" value = "#ECID #" > "
  < input type = "hidden" name = "' ItemID #ItemID #" value = "#ItemID #" > "
  < input type = "hidden" name = "" PNR_Doc_Type #ItemID # ' value = "PNR Req" > "
  
  < td align = "center" >
  < cfif Approval_Initials is not "" > "".
  #Approval_Initials #.
  < cfelse >
  
  < select name = "PNR_Approval_Initials #ItemID #" >
  < option value = "" > select initials < / option >
  < cfloop query = "ShowInitials" >
  < Cfif Engineer_Initials EQ 1 >
  < option value = "#Initials #" > #Initials # < / option >
  < / cfif >
  < / cfloop >
  < / select >
  
  < / cfif >
  < table >
  
  Can someone help me please on just the insertion in the database so that it
  will insert on the lines that I chose the initials on and not others? Thank you.
  
  Andy

  You almost managed, but I see that you have commented.

  (" The help of # PNR_Approval_Initials # ') AND form ["PNR_Approval_Initials # help #" ""] NEQ "" >

  Do things here

• Updated several lines with different values

  Hello!
  I have a problem. I need to update more than 1000 lines with different values. How can I do?
  For exsample I have table:
  ID; color, date,
  1 red
  2 green
  3 white
  
  I need to update the date field.
  
  Update table
  Set date = '01.02.03'
  where id = 1
  
  Update table
  Set date = '01.03.03'
  where id = 2
  
  
  Maybe it's how to update multiple rows in a single request?
  
  Sorry for my bad English.
  Thank you!

  Hello

  You can try this

  UPDATE TABLE SET DATE = CASE
                      WHEN ID = 1 THEN TO_DATE('01-02-03','DD-MM-RR')
                      WHEN ID = 2 THEN TO_DATE('01-03-03','DD-MM-RR')
                      END
  

  see you soon

  VT

• Assignment of sequential numbers to several lines with a common attribute

  Oracle 10i, seek to alter a query only select a total of 96 unique compoundnames and number each batchids Group 1-96.
  
  
  Entry
  Table EX
  
  Unique_ID, CompoundName, BatchID, quantity
  -------------------------------------------------------------------
  1, 1000 EX, EX 1000 100, 100
  2, 1000 EX, EX 101 1000, 100
  3, 1000 EX, EX 1000 102, 100
  4, EX 1001, EX 1001 100, 100
  5, EX 1001, EX 1001 101, 100
  6, EX 1001, EX 1001 102, 100
  7, EX 1001, EX 1001 103, 100
  8, EX 1001, EX 1001 104, 100
  9, EX 1001, EX 1001 105, 100
  10, EX 1002, 1002 100 EX, 100
  11, EX 1002, 1002 101 EX, 100
  12, EX 1002, 1002 102 EX, 100
  .
  .
  .
  100 +, EX 1095, EX-1095 100, 100
  
  
  -----
  Using
  
  Select distinct * from
  (
  Select Unique_id, CompoundName, batchid, quantity, Sum (Quantity) on sumqty (CompoundName partition)
  EX
  )
  where < 500 sumqty
  Order by 2, 3;
  
  Provides data collected by compoundnames and the total amount of all sharing the same name composed data (and only the results with the total amount < 500).
  
  I am looking for a way to recover only 96 compoundnames unique (not total the data lines) at a time and the numbering of resemble them this desired output;
  
  
  Unique_ID, CompoundName, BatchID, quantity, Sum (Quantity)?
  -------------------------------------------------------------------
  1 EX 1000, 1000, 100, 100, 300, 1 EX
  2, 1000 EX, EX 101 1000, 100, 300, 1
  3, 1000 EX, EX 1000 102, 100, 300, 1
  
  10, EX 1002, 1002, 100, 100, 300, 2 EX
  11, EX 1002, 1002, 101, 100, 300, 2 EX
  12, EX 1002, 1002, 102, 100, 300, 2 EX
  .
  .
  .
  100 + EX-1095, EX 96, 100, 100, 100 1095
  
  Thanks in advance.
  
  Published by: user13112928 on June 26, 2010 11:49
  
  Published by: user13112928 on June 26, 2010 11:50
  
  Published by: user13112928 on June 26, 2010 12:01

  Then place the DENSE_RANK of the outer query when you filter the data, not the inner query where you have it now.

• How to get several lines with series C 9501 running with SoftMotion on a cRIO 9074 modules?

  Hello

  I'll have set up many of the variable axes defined by the user with two modules in series c 9501 running on a Compact RIO 9074 stepper.

  I started with the example of "9501 Stepper Drive (Getting Started)", which worked fine with a simple variable user defined axis (Test-interactive panel, SoftMotion functionality on the host).

  I need to control 2 axes and I could not find any help on adding a second axis, so I came up with the following approach. My approach was to expand the FPGA VI in a way to add features to the second variable axis defined by the user, so I did the following steps:

  1. Add a new variable axis defined by the user.

  2. Add the variables defined by the same user as the first axis had during the second axis.

  3. a copy of global variables that the first axis necessary to use independent global variables for the second axis.

  4. develop the FPGA VI while-loops with the circuits of same for axis 2 and ensure that the circuits for axis 2 use their own global variables, the right modules, their user-defined variables and change the index of the axis from 0 to 1 (don't know exactly how the axis index is defined).

  4B. I also tried to copy all the loops for each axis had their independent loops. Then adapt the global variables, modules, the variables defined by the user, index of the axis.

  The FPGA VI compiles and runs, but when I get to the Panel of interactive Test for both axes and try to enable them, it throws an error-70228 (Controller unable to confirm that the drive is enabled).

  I'm doing something wrong? Is it perhaps a simpler way to get two 9501 modules working with SoftMotion?

  I can provide my two examples of published projects (same loop loop / independent), if necessary.

  LabVIEW installation:

  LabVIEW 2014f1

  Module FPGA 2014

  In time real Module 2014

  SoftMotion 2014f1

  cRIO-drivers 2014f1

  Xilinx tools 14.7 2014

  I hope someone can help me, thanks in advance.

  I managed to make it work.

  I was missing that a Subvi in the FPGA VI engine control missed their axis index 0. I couldn't even they worked with an index of the axis until I looked into the block diagram of those.

  So what happened was that my routines on the second axis disabled immediately the first axis because of bad axis index.

  Definition of the index of 0 and 1 axis (axes 1 and 2) for each unique Subvi of motor control was what was missing.

• Restrict lines with separate columns

  Hello
  
  I have all of the following documents:
  Name Dept
  
  John management
  Peter HR
  John Marketing
  Peter finance
  Harry operations
  
  Now I want to select only one record of each to John, Peter and Harry... it's name should be separate, Department it can be any one of the folders.
  Like this
  Name Dept
  Management of John (or Marketing)
  Peter HR (or Finance)
  Harry operations
  
  You could take the first record of a name and filter any file with the same name. Another method would be as well. I tried
  to make a request for this for some time without success.
  
  Thank you very much.

  keep things simple:

  SQL> WITH det AS
    2       (
    3          SELECT 'John' na_me, 'Management' dept
    4            FROM DUAL
    5          UNION ALL
    6          SELECT 'Peter', 'HR'
    7            FROM DUAL
    8          UNION ALL
    9          SELECT 'John', 'Marketing'
   10            FROM DUAL
   11          UNION ALL
   12          SELECT 'Peter', 'Finance'
   13            FROM DUAL
   14          UNION ALL
   15          SELECT 'Harry', 'Operations'
   16            FROM DUAL)
   17  SELECT   na_me, MIN(dept)
   18      FROM det
   19  GROUP BY na_me;
  
  NA_ME MIN(DEPT)
  ----- ----------
  Harry Operations
  John  Management
  Peter Finance
  

• Replace repeating lines with white and don't show the distinct values

  Hi all
  
  I have a SQL query that looks like this:
  
  SELECT Item_id, item_details, category, group INC.
  OF my_items
  WHERE THE...
  
  The results are as follows:
  
  Item_id item_details category group
  
  001 keyboard PC - 1 Grp1
  001 keyboard PC - 2 Grp1
  003 monitor PC - 1 Grp 1
  003 monitor PC - 2 Grp 2
  
  
  What I really want, it's like to have repeated as a draft values and keep only the distinct values. I want the output to be like this:
  
  Item_id item_details category group
  
  001 keyboard PC - 1 Grp1
  PC-2
  003 monitor PC - 1 Grp 1
  PC - 2 Grp 2
  
  
  Is there a way we can do that? Help, please

  Here's a way using the Lag as mentioned by Tubby...

  SQL> ed
  Wrote file afiedt.buf
  
    1  with t as (select '001' as item_id, 'keyboard' as item_details, 'PC-1' as category, 'Grp1' as grp from dual union all
    2             select '001', 'keyboard', 'PC-2', 'Grp1' from dual union all
    3             select '003', 'monitor', 'PC-1', 'Grp1' from dual union all
    4             select '003', 'monitor', 'PC-2', 'Grp2' from dual)
    5  -- end of test data
    6  select decode(item_id, lag(item_id) over (partition by item_id order by category, grp), null, item_id) as item_id
    7        ,decode(item_details, lag(item_details) over (partition by item_id order by category, grp), null, item_details) as item_details
    8        ,decode(category, lag(category) over (partition by item_id order by category, grp), null, category) as category
    9        ,decode(grp, lag(grp) over (partition by item_id order by category, grp), null, grp) as grp
   10  from t
   11* order by t.item_id, t.category, t.grp
  SQL> /
  
  ITE ITEM_DET CATE GRP
  --- -------- ---- ----
  001 keyboard PC-1 Grp1
               PC-2
  003 monitor  PC-1 Grp1
               PC-2 Grp2
  
  SQL>
  

• How can I play files CSV lines with a different number of columns?

  Hi all

  I am trying to load CSV with DIAdem 2014 files, and I found there is a great tool, "use".

  However, with the CSV use tool, there are critical problem with my data files. (See photo)

  As you can see the CSV file attached, it includes a lot of lines with different columns.

  And here, use read the number of columns of 1st line (e.g.: 4 columns) and set it as the number of columns to read.

  So I can not load all of the data (e.g.: 53 columns) if first row data were lower than columns.

  Can anyone recommend any approach, please?

  Kind regards

  Young

  Admin 24/03/16 Note: files deleted by request of the user

  I just add a few lines of script and fix the related use. (I just added the code to get a direct look).

  The uri of the file will appear as a new kind of load in the dialog file tiara is intalled by double click

  Would be nice if you could provide some information, how the data was created. Maybe we can it add as official of our web page to use plugin.

  Option Explicit
  
  Void ReadStore (File)
  File.Formatter.Delimiters = «»
  
    File.Formatter.LineFeeds = \n
  
    File.Formatter.DecimalPoint = '. '.
  Dim startLine: startLine = file. GetNextLine()
  
    <>InStr (startLine, ' [Tenergy Bus Log Data] "" ") then
  
      call RaiseError()
  
    end if
  And that not File.Position = File.Size
  
      Dim groupName: groupName = file. GetNextStringValue (eString)
  
      If it isn't root. ChannelGroups.Exists (groupName) then
  
        root. ChannelGroups.Add (groupName)
  
      end if
  
      Dim grp: set grp = root. ChannelGroups (groupName)
  
      Dim i: i = 1
  
      do for real
  
        Dim val: val = file. GetNextStringValue (eR64)
  
        If isempty (val) then
  
          Exit
  
        end if
  
        < i="">
  
          dial the grp. Channels.Add ("Brand" & I, eR64)
  
        end if
  
        protected chObj: set chObj = grp. Channels (i)
  
        chObj.Values (chObj.Size + 1) = val
  
        i = i + 1
  
      loop
  
      Call File.SkipLine)
  
    Wend
  End Sub

• Put 'do shell script' multi-line with comments

  I would like to integrate a Bash in Applescript script, as it is on several lines with comments.

  The only documentation useful, that I can find is tn2065, but as far as I can see, it does not say how.

  What is the correct syntax?

  Maybe I'm missing something, but you can just write your script bash as a text object and passing to the shell script:

  game of longBashCommand to "# some relevant comments.

  CD /somedir

  / bin/something - foobar

  # one other comment here

  / bin/Gisèle

  ' # one other comment.

  the shell script longBashCommand

  The trickiest part is perhaps the need to escape some characters such as quotation marks in the command shell.

• Draw lines with smooth curves

  Hello

  I am doing a map for an underground system in a gaming PC, and I want something similar to a Transport for London has:

  http://www.tfl.gov.UK/assets/downloads/standard-tube-map.gif

  I am a total newbie with Adobe Illustrator, but I try it as I have heard that this is using Transport for London.

  My question is then, what would be the best way to accomplish several lines with different colors and smooth curves, like the lines in my table of reference?

  Thanks in advance,

  Martin

  Martin,

  Complete what by says Kurt, you can draw traces with segments of straight lines and then cut corners later, using the effect > styles > corners (you can object > decaying appearance in order to get real rounding).

• Convert a line into several lines

  Hello
  
  I have a Calendar_1 table with 4 columns and 1 row
  
  CREATE TABLE Calendar_1
  (
  Sunday (1) tank,.
  Monday (1) tank,.
  Tuesday (1) tank,.
  Wednesday char (1)
  );
  
  INSERT INTO Calendar_1 VALUES ('A', ' B ', 'C' ' d ' ");
  COMMIT;
  
  I need to get my result under a single column 'DAYS' as shown below
  
  DAYS
  --------
  A
  B
  C
  D
  
  Could you please provide me with the required SQL code
  
  Concerning
  -Learnsequel

  Like this

  select decode(l, 1, sunday, 2, monday, 3, tuesday, 4, wednesday) days
    from calendar_1
    cross join (select level l from dual connect by level <= 4) 
  

• Help to extract data Clob to a column in a table that has several lines in a table

  Help to extract data Clob to a column in a table that has several lines in a table

  It works for only the first line

  CREATE or REPLACE DIRECTORY XMLDIR AS ' / orabackups';

  Grant read, write on DIRECTORY XMLDIR to the public;

  () dbms_xslprocessor.clob2file

  beef in CLOB,

  XMLDIR IN VARCHAR2,

  "testfile2.txt" IN VARCHAR2);

  DECLARE

  buf CLOB.

  BEGIN

  SELECT H15_DOC

  IN buf

  OF H15TEST. H15_STAGE

  where rownum = 1;

  dbms_xslprocessor.clob2file (buf, 'XMLDIR', 'testfile2.txt');

  END;

  /

  -This error code

  CREATE or REPLACE DIRECTORY XMLDIR AS ' / orabackups';

  Grant read, write on DIRECTORY XMLDIR to the public;

  () dbms_xslprocessor.clob2file

  XData in CLOB,

  XMLDIR IN VARCHAR2,

  "testfile2.txt" IN VARCHAR2,

  CSID in NUMBER: = 0);

  DECLARE

  CURSOR xmlmycur IS SELECT H15_DOC

  OF H15TEST. H15_STAGE

  where rownum = 102140;

  l_clob CLOB.

  XData CLOB.

  BEGIN

  DBMS_LOB.CREATETEMPORARY (l_clob, true);

  DBMS_LOB.CREATETEMPORARY (xdata, true);

  OPEN xmlmycur.

  LOOP

  SEEK xmlmycur INTO xdata;

  dbms_xslprocessor.clob2file (xdata, 'XMLDIR', 'testfile2.txt');

  EXIT WHEN xmlmycur % notfound;

  END LOOP;

  CLOSE Xmlmycur;

  END;

  /

  ORA-21560: 3 argument is null, invalid or out of range

  ORA-06512: at "SYS." DBMS_LOB", line 991

  ORA-06512: at "XDB". DBMS_XSLPROCESSOR', line 324

  ORA-06512: at line 15 level

  FYI, it seems that the file being created will exceed 5 gig

  Yes, it is a known issue with DBMS_XSLPROCESSOR.clob2file with big files.

  Here is a slightly modified version of the code that I posted earlier, fixing WRITE_ERROR except:

  DECLARE

  v_file utl_file.file_type;

  -write a unique to clob

  procedure write_clob (p_content in clob) is

  v_buffer varchar2 (32767).

  V_POSITION pls_integer: = 1;

  v_amount pls_integer: = 32767;

  Start

  loop

  Start

  DBMS_LOB. Read (p_content, v_amount, v_position, v_buffer);

  exception

  When no_data_found then exit;

  end;

  UTL_FILE.put_raw (v_file, utl_raw.cast_to_raw (v_buffer), true);

  V_POSITION: = v_position + v_amount;

  end loop;

  end;

  BEGIN

  -Open the file

  v_file: = utl_file.fopen ('TEST_DIR', 'testfile2.txt', 'wb', 32767);

  -loops through the lines

  for r in)

  Select h15_doc

  of h15test.h15_stage

  )

  loop

  write_clob (r.h15_doc);

  end loop;

  UTL_FILE.fclose (v_file);

  END;

  /