I have a query which works nice when I run it inside sqlplus:
SQL> SELECT T_0.ID AS ATTR_1_, T_0_0.ID AS ATTR_2_,
CASE WHEN ( T_0.ID=1 AND ( T_0_0.ID=3 OR T_0_1.ID='val_1') )
THEN 'val_1' ELSE 'val_2' END AS TXT, T_0_1.ID,
CASE WHEN T_0.ID='boo' THEN 'boo' END AS EXTRA_FIELD
FROM TEST_TABLE T_0
INNER JOIN TEST_TABLE_2 T_0_0 ON ( T_0_0.ATTR=T_0.ID )
INNER JOIN TEST_TABLE_3 T_0_1 ON ( T_0_1.ID = T_0_0.ID )
WHERE ( ( T_0.ID=1 AND T_0_0.ID=3 )
OR T_0_1.ID=2 OR T_0_0.TXT='val_2');
no rows selected
Although, it returns nothing, it still works and does not result in error. However, when I do the same thing in Python, using bindings, I get this error message:
cx_Oracle.DatabaseError: ORA-01722: invalid number
This is how my query looks in Python, before I do cursor.execute:
SELECT T_0.ID AS ATTR_1_, T_0_0.ID AS ATTR_2_,
CASE WHEN ( T_0.ID=:TXT_ AND ( T_0_0.ID=:TXT__ OR T_0_1.ID=:TXT___ ) )
THEN :TXT___ ELSE :TXT____ END AS TXT, T_0_1.ID,
CASE WHEN T_0.ID=:EXTRA_FIELD THEN :EXTRA_FIELD END AS EXTRA_FIELD
FROM TEST_TABLE T_0
INNER JOIN TEST_TABLE_2 T_0_0 ON ( T_0_0.ATTR=T_0.ID )
INNER JOIN TEST_TABLE_3 T_0_1 ON ( T_0_1.ID = T_0_0.ID )
WHERE ( ( T_0.ID=:ID AND T_0_0.ID=:ID_ )
OR T_0_1.ID=:ID__ OR T_0_0.TXT=:TXT )
The query is just a string double-quoted "SELECT ..." . And this is how the dictionary with binding variables looks like:
OrderedDict([('TXT_', 1), ('TXT__', 3), ('TXT___', 'val_1'),
('TXT____', 'val_2'), ('EXTRA_FIELD', 'boo'), ('ID', 1),
('ID_', 3), ('ID__', 2), ('TXT', 'val_2')])
So, as you can see I have a perfect dictionary - number values are just numbers without quotes, string values are just strings with single quotes. I know, you will ask about the schema of the tables. So, here its is:
SQL> SELECT COLUMN_NAME, DATA_TYPE FROM USER_TAB_COLUMNS WHERE
TABLE_NAME = 'TEST_TABLE';
COLUMN_NAME
------------------------------
DATA_TYPE
------------------------------
ID
NUMBER
SQL> SELECT COLUMN_NAME, DATA_TYPE FROM USER_TAB_COLUMNS WHERE
TABLE_NAME = 'TEST_TABLE_2';
COLUMN_NAME
------------------------------
DATA_TYPE
------------------------------
ATTR
NUMBER
ID
NUMBER
TXT
VARCHAR2
SQL> SELECT COLUMN_NAME, DATA_TYPE FROM USER_TAB_COLUMNS
WHERE TABLE_NAME = 'TEST_TABLE_3';
COLUMN_NAME
------------------------------
DATA_TYPE
------------------------------
ID
NUMBER
So, it seems like one and the same query works good in the console, but does not work when using Python. Why is that?
EDIT
And here is a proof - a screen of two console windows. In the first console I run the query in sqlplus, in the second console I print sql query and the dictionary, which is used for binding variables:
EDIT
Oh, it's even more interesting. I was able to reproduce this error in Oracle shell and it looks like Oracle 11c bug. So, look at this:
Please, pay attention to the fact that ID field has a NUMBER type. And then pay attention to these two screens:
In the screen above you can see that everything is ok. However, if we slightly change it by adding OR T_0_1.ID=2 to the WHERE part, then it breaks:
So, this problem is reproducible even in Oracle shell. You can do it, using the schema I provided above.
EDIT
I updated the topic of my question, because it has nothing to do with Python. The whole problem with Oracle itself.
EDIT
BTW. My last comment does not contradict to the beginning part of my investigation. The thing is, if I have some data in TEST_TABLE_3, then the query breaks. And if I delete data, then is starts working. Here is a big proof:
How can data affect correctness of the query??
On your last screen just below the last line of the statement you have
CASE WHEN ( T_O.ID=1 AND ( T_0_0.ID=3 OR T_0_1.ID='VAL_1') )
there's an asterisk (now it helps, but sometimes it could lead in the wrong direction) showing the place of the encountered issue
T_0_1.ID='VAL_1'
in your table ID column is of Number type. 'VAL_1' - is Varchar.
As the comparison rules state:
When comparing a character value with a numeric value, Oracle converts the character data to a numeric value.
see (https://docs.oracle.com/database/121/SQLRF/sql_elements002.htm#SQLRF00214)
when oracle encounters this it tries to cast your string to number - and you get the error
How can data affect correctness of the query??
When there's no data in the table - there's no record returned from the table, hence there's no need the check the value of the column for equality - this comparison is not executed and no error shown
Related
I have a program that has been working for years. Today, we upgraded from SAS 9.4M3 to 9.4M7.
proc setinit
Current version: 9.04.01M7P080520
Since then, I am not able to get the same results as before the upgrade.
Please note that I am querying on Oracle databases directly.
Trying to replicate the issue with a minimal, reproducible SAS table example, I found that the issue disappear when querying on a SAS table instead of on Oracle databases.
Let's say I have the following dataset:
data have;
infile datalines delimiter="|";
input name :$8. id $1. value :$8. t1 :$10.;
datalines;
Joe|A|TLO
Joe|B|IKSK
Joe|C|Yes
;
Using the temporary table:
proc sql;
create table want as
select name,
min(case when id = "A" then value else "" end) as A length 8
from have
group by name;
quit;
Results:
name A
Joe TLO
However, when running the very same query on the oracle database directly I get a missing value instead:
proc sql;
create table want as
select name,
min(case when id = "A" then value else "" end) as A length 8
from have_oracle
group by name;
quit;
name A
Joe
As per documentation, the min() function is behaving properly when used on the SAS table
The MIN function returns a missing value (.) only if all arguments are missing.
I believe this happens when Oracle don't understand the function that SAS is passing it - the min functions in SAS and Oracle are very different and the equivalent in SAS would be LEAST().
So my guess is that the upgrade messed up how is translates the SAS min function to Oracle, but it remains a guess. Does anyone ran into this type of behavior?
EDIT: #Richard's comment
options sastrace=',,,d' sastraceloc=saslog nostsuffix;
proc sql;
create table want as
select t1.name,
min(case when id = 'A' then value else "" end) as A length 8
from oracle_db.names t1 inner join oracle_db.ids t2 on (t1.tid = t2.tid)
group by t1.name;
ORACLE_26: Prepared: on connection 0
SELECT * FROM NAMES
ORACLE_27: Prepared: on connection 1
SELECT UI.INDEX_NAME, UIC.COLUMN_NAME FROM USER_INDEXES UI,USER_IND_COLUMNS UIC WHERE UI.TABLE_NAME='NAMES' AND
UIC.TABLE_NAME='NAMES' AND UI.INDEX_NAME=UIC.INDEX_NAME
ORACLE_28: Executed: on connection 1
SELECT statement ORACLE_27
ORACLE_29: Prepared: on connection 0
SELECT * FROM IDS
ORACLE_30: Prepared: on connection 1
SELECT UI.INDEX_NAME, UIC.COLUMN_NAME FROM USER_INDEXES UI,USER_IND_COLUMNS UIC WHERE UI.TABLE_NAME='IDS' AND
UIC.TABLE_NAME='IDS' AND UI.INDEX_NAME=UIC.INDEX_NAME
ORACLE_31: Executed: on connection 1
SELECT statement ORACLE_30
ORACLE_32: Prepared: on connection 0
select t1."NAME", MIN(case when t2."ID" = 'A' then t1."VALUE" else ' ' end) as A from
NAMES t1 inner join IDS t2 on t1."TID" = t2."TID" group by t1."NAME"
ORACLE_33: Executed: on connection 0
SELECT statement ORACLE_32
ACCESS ENGINE: SQL statement was passed to the DBMS for fetching data.
NOTE: Table WORK.SELECTED_ATTR created, with 1 row and 2 columns.
! quit;
NOTE: PROCEDURE SQL used (Total process time):
real time 0.34 seconds
cpu time 0.09 seconds
Use the SASTRACE= system option to log SQL statements sent to the DBMS.
options SASTRACE=',,,d';
will provide the most detailed logging.
From the prepared statement you can see why you are getting a blank from the Oracle query.
select
t1."NAME"
, MIN ( case
when t2."ID" = 'A' then t1."VALUE"
else ' '
end
) as A
from
NAMES t1 inner join IDS t2 on t1."TID" = t2."TID"
group by
t1."NAME"
The SQL MIN () aggregate function will exclude null values from consideration.
In SAS SQL, a blank value is also interpreted as null.
In SAS your SQL query returns the min non-null value TLO
In Oracle transformed query, the SAS blank '' is transformed to ' ' a single blank character, which is not-null, and thus ' ' < 'TLO' and you get the blank result.
The actual MIN you want to force in Oracle is min(case when id = "A" then value else null end) which #Tom has shown is possible by omitting the else clause.
The only way to see the actual difference is to run the query with trace in the prior SAS version, or if lucky, see the explanation in the (ignored by many) "What's New" documents.
Why are you using ' ' or '' as the ELSE value? Perhaps Oracle is treating a string with blanks in it differently than a null string.
Why not use null in the ELSE clause?
or just leave off the ELSE clause and let it default to null?
libname mylib oracle .... ;
proc sql;
create table want as
select name
, min(case when id = "A" then value else null end) as A length 8
from mylib.have_oracle
group by name
;
quit;
Also try running the Oracle code yourself, instead of using implicit pass thru.
proc sql;
connect to oracle ..... ;
create table want as
select * from connection to oracle
(
select name,
min(case when id = "A" then value else null end) as A length 8
from have_oracle
group by name
)
;
quit;
When I try to reproduce this in Oracle I get the result you are looking for so I suspect it has something to do with SAS (which I'm not familiar with).
with t as (
select 'Joe' name, 'A' id, 'TLO' value from dual union all
select 'Joe' name, 'B' id, 'IKSK' value from dual union all
select 'Joe' name, 'C' id, 'Yes' value from dual
)
select name
, min(case when id = 'A' then value else '' end) as a
from t
group by name;
NAME A
---- ----
Joe TLO
Unrelated, if you are only interested in id = 'A' then a better query would be:
select name
, min(value) as a
from t
where id = 'A'
group by name;
SELECT b.*
FROM buses b,
bus_stations bs,
starts st,
stops_at sa
WHERE st.station_no = ( SELECT station_id
FROM bus_stations
WHERE station_name = "golden mile_Regina"
)
AND sa.station_no = ( SELECT station_id
FROM bus_stations
WHERE station_name = 'westmount_edmonton'
)
ORDER BY DATE;
You can't use double quotes with strings - use single ones, i.e.
WHERE station_name = 'golden mile_Regina'
By the way, are you sure of spelling & letter size? Is it really mixed case, with underscores? Just asking.
Furthermore, you're ordering by DATE - that won't work either, you can't use DATE as a column name (unless you enclose it into double quotes, but I certainly wouldn't recommend that). Have a look at the following example (stupid, yes - setting date to be a number, but I used it just to emphasize that DATE can't be used as a column name):
SQL> create table test (date number);
create table test (date number)
*
ERROR at line 1:
ORA-00904: : invalid identifier
Once you fix that, you'll get unexpected result as there are 4 tables in the FROM clause, but they aren't joined with one another, so that will be a nice Cartesian product.
I am trying to query some jobs from a repo, however I only need the job with the latest start time. I have tried using ROW_NUMBER for this and select only row number 1 for each job, however it doesn't seem to fall through:
SELECT a.jobname||','||a.projectname||','||a.startdate||','||a.enddate||','||
ROW_NUMBER() OVER ( PARTITION BY a.jobname ORDER BY a.startdate DESC ) AS "rowID"
FROM taskhistory a
WHERE a.jobname IS NOT NULL AND a.startdate >= (SYSDATE-1))LIMIT 1 AND rowID = 1;
ERROR at line 7:
ORA-00932: inconsistent datatypes: expected ROWID got NUMBER
Can I please ask for some assistance?
You have aliased your concatenated string "rowID" which is a mistake because it clashes with the Oracle keyword rowid. This is a special datatype, which allows us to identify table rows by their physical location. Find out more.
When you reference the column alias you omitted the fouble quotes. Oracle therefore interprets it as the keyword, rowid, and expects an expression which can be converted to the ROWID datatype.
Double-quoted identifiers are always a bad idea. Avoid them unless truly necessary.
Fixing the column alias will reveal the logic bug in your code. You are concatenating a whole slew of columns together, including the ROW_NUMBER() function, and calling that string "rowID". Clearly that string is never going to equal one, so this will filter out all rows:
and "rowID" = 1
Also LIMIT is not valid in Oracle.
What you need to do is use a sub-query, like this
SELECT a.jobname||','
||a.projectname||','
||a.startdate||','
||a.enddate||','
||to_char(a.rn) as "rowID"
FROM (
SELECT jobname
, projectname
, startdatem
, enddate,
, ROW_NUMBER() OVER ( PARTITION BY jobname
ORDER BY startdate DESC ) AS RN
FROM taskhistory
WHERE jobname IS NOT NULL
AND a.startdate >= (SYSDATE-1)
) a
where a.RN = 1;
Concatenating the projection like that seems an odd thing to do but I don't understand your business requirements.
A weird request maybe but. My boss wants me to create an admin version of a page we have that displays data from an oracle query in a table.
The admin page, instead of displaying the data (query returns 1 row), needs to return the table name and column name
Ex: Instead of:
Name Initial
==================
Bob A
I want:
Name Initial
============================
Users.FirstName Users.MiddleInitial
I realize I can do this in code but would rather just modify the query to return the data I want so I can leave the report generation code mostly alone.
I don't want to do it in a stored procedure.
So when I spit out the data in the report using something like:
blah blah = MyDataRow("FirstName")
I can leave that as is but instead of it displaying "BOB" it would display "Users.FirstName"
And I want to do the query using select * if possible instead of listing all the columns
So for each of the columns I am querying in the * , I want to get (instead of the column value) the tablename.ColumnName or tablename|columnName
hope you are following- I am confusing myself...
pseudo:
select tablename + '.' + Columnname as WhateverTheColumnNameIs
from Table1
left join Table2 on whatever...
Join Table_Names on blah blah
Whew- after writing all this I think I will just do it on the code side.
But if you are up for it maybe a fun challenge
Oracle does not provide an authentic way(there is no pseudocolumn) to get the column name of a table as a result of a query against that table. But you might consider these two approaches:
Extract column name from an xmltype, formed by passing cursor expression(your query) in the xmltable() function:
-- your table
with t1(first_name, middle_name) as(
select 1,2 from dual
), -- your query
t2 as(
select * -- col1 as "t1.col1"
--, col2 as "t1.col2"
--, col3 as "t1.col3"
from hr.t1
)
select *
from ( select q.object_value.getrootelement() as col_name
, rownum as rn
from xmltable('//*'
passing xmltype(cursor(select * from t2 where rownum = 1))
) q
where q.object_value.getrootelement() not in ('ROWSET', 'ROW')
)
pivot(
max(col_name) for rn in (1 as "name", 2 as "initial")
)
Result:
name initial
--------------- ---------------
FIRST_NAME MIDDLE_NAME
Note: In order for column names to be prefixed with table name, you need to list them
explicitly in the select list of a query and supply an alias, manually.
PL/SQL approach. Starting from Oracle 11g you could use dbms_sql() package and describe_columns() procedure specifically to get the name of columns in the cursor(your select).
This might be what you are looking for, try selecting from system views USER_TAB_COLS or ALL_TAB_COLS.
DB: Oracle 11g
Query:
SELECT CASE
WHEN rs.OPTION = '3'
THEN
(SELECT COUNT(DISTINCT ex.EXTS) AS TMPCOL0
FROM CRSIDM.SUB_OPTS ex
INNER JOIN CRSIDM.SUB_OPTS_GRP cg
ON cg.GROUP_ID = ex.GRP_ID
)
ELSE
(SELECT COUNT(DISTINCT ex.EXTS) AS TMPCOL0
FROM CRSIDM.SUB_OPTS ex
INNER JOIN CRSIDM.SUB_OPTS_POL cg
ON cg.GROUP_ID = ex.GRP_ID
)
END AS PROPTS
FROM PR_OPTS
I am getting error 'expected CHAR got NUMBER', here EXTS,GROUP_ID & GRP_ID are numeric. Then how there is a chance of expecting CHAR?
Generally when Oracle compares different datatypes such as a NUMBER with a CHARACTER, implicit conversion kicks in and all is well (provided the data can be converted.) For example, if you have a function that expects a CHARACTER value but you pass it a NUMBER, all is well - Oracle simply converts the NUMBER to character.
E.g. a function like this:
create or replace function get_something(p_id VARCHAR2) return number ...
works if you call it with this:
get_dno(10);
or this:
get_dno('10');
and in SQL:
select * from some_table where numeric_column = '10' -- no problem.
A popular place where you see this kind of error is with the return values in CASE statements. For instance, you'll get that error if you have something like this:
SQL> SELECT CASE WHEN 1 = 1 THEN '1' ELSE 2 END
2 FROM dual
3 ;
SELECT CASE WHEN 1 = 1 THEN '1' ELSE 2 END
*
ERROR at line 1:
ORA-00932: inconsistent datatypes: expected CHAR got NUMBER
(The datatype from the first WHEN clause is what it expects in the other WHEN/ELSE clauses that follow.)
But in your case the WHEN and THEN both return counts - the datatypes are consistent. So, I think you have a red-herring in there.
As Alex mentioned above, OPTION is a keyword and if you try and create a table with that as a column name, Oracle disagrees:
SQL> create table dummy
2 (option varchar2(10)
3 );
(option varchar2(10)
*
ERROR at line 2:
ORA-00904: : invalid identifier
This works:
SQL> create table dummy
2 (option_col varchar2(10)
3 );
Table created.
or you could do it with quotes:
SQL> create table dummy
2 ("option" varchar2(10));
Table created.
But now you're in a world of hurt - you need quotes from now on:
SQL> select option from dummy;
select option from dummy
*
ERROR at line 1:
ORA-00936: missing expression
SQL> select d.option from dummy d;
select d.option from dummy d
*
ERROR at line 1:
ORA-01747: invalid user.table.column, table.column, or column specification
With quotes:
SQL> select d."option" from dummy d;
no rows selected
So, if your query is really giving you "expected CHAR, got NUMBER", it looks to me like something is off.
Essentially, it means some of the fields you are using aren't compatible with each other. It's basically a "type mismatch". Just check to see if any types of CHAR are being used with types of NUMBER. Then you can either switch the type of one, or simply use a conversion as part of the query.
The issue is OPTION = '3', the quotation marks indicate that you're looking for a string containing the solitary character 3.
Try this instead:
SELECT CASE
WHEN rs.OPTION = 3
THEN
(SELECT COUNT(DISTINCT ex.EXTS) AS TMPCOL0
FROM CRSIDM.SUB_OPTS ex
INNER JOIN CRSIDM.SUB_OPTS_GRP cg
ON cg.GROUP_ID = ex.GRP_ID)
ELSE
(SELECT COUNT(DISTINCT ex.EXTS) AS TMPCOL0
FROM CRSIDM.SUB_OPTS ex
INNER JOIN CRSIDM.SUB_OPTS_POL cg
ON cg.GROUP_ID = ex.GRP_ID)
END AS PROPTS
FROM PR_OPTS