I'm having a staging table and a datawarehouse table, which keep giving me constraint violation. i can't seem to figure out why since DRIVERID and RACEID a combination of those should be unique? How come i get contraint violation - primary key
table
CREATE TABLE QUALIFYING (
QUALIFYID DECIMAL(18,0) IDENTITY NOT NULL,
RACEID DECIMAL(18,0) DEFAULT '0' NOT NULL,
DRIVERID DECIMAL(18,0) DEFAULT '0' NOT NULL,
CONSTRUCTORID DECIMAL(18,0) DEFAULT '0' NOT NULL,
DRIVERNUMBER DECIMAL(18,0) DEFAULT '0' NOT NULL,
DRIVERPOSITION DECIMAL(18,0) DEFAULT NULL,
Q1 VARCHAR(255) UTF8 DEFAULT NULL,
Q2 VARCHAR(255) UTF8 DEFAULT NULL,
Q3 VARCHAR(255) UTF8 DEFAULT NULL,
PRIMARY KEY(QUALIFYID)
);
Staging
CREATE OR REPLACE TABLE STGQUALIFYING(
raceId int DEFAULT '0' NOT NULL,
driverId int DEFAULT '0' NOT NULL,
constructorId int DEFAULT '0' NOT NULL,
driverNumber int DEFAULT '0' NOT NULL,
driverPosition int DEFAULT NULL,
q1 varchar(255) DEFAULT NULL,
q2 varchar(255) DEFAULT NULL,
q3 varchar(255) DEFAULT NULL,
PRIMARY KEY(RACEID, DRIVERID)
);
SQL
MERGE INTO QUALIFYING c
USING STGQUALIFYING n
ON
(n.RACEID = c.RACEID AND n.DRIVERID = c.DRIVERID)
WHEN MATCHED THEN
UPDATE SET
CONSTRUCTORID = n.CONSTRUCTORID, DRIVERNUMBER = n.DRIVERNUMBER, DRIVERPOSITION = n.DRIVERPOSITION, Q1 = n.Q1, Q2 = n.Q2, Q3 = n.Q3
WHEN NOT MATCHED THEN
INSERT (RACEID, DRIVERID, CONSTRUCTORID, DRIVERNUMBER, DRIVERPOSITION, Q1, Q2, Q3) VALUES
(RACEID, DRIVERID, CONSTRUCTORID, DRIVERNUMBER, DRIVERPOSITION, Q1, Q2, Q3);
The EXASolution user manual says:
The content of an identity column applies to the following rules:
If you specify an explicit value for the identity column while inserting a row, then this value is inserted.
In all other cases monotonically increasing numbers are generated by the system, but gaps can occur between the numbers.
and
You should not mistake an identity column with a constraint, i.e. identity columns do not guarantee unique values. But the values are unique as long as values are inserted only implicitly and are not changed manually.
You've put a primary key constraint on your identity column, so it must be unique. Since you are getting duplicates from your merge, either (a) you have, at some point, provided explicit values as in the first bullet above or updated a value manually, and the monotonically increasing sequence has reached a point where it is clashing with those existing values; or (b) there's a bug in their merge. The former seems more likely.
You can look at recently inserted value if you have one, or do a temporary insert of a new row (with merge) to see if it will create a row successfully, and if so whether you already have ID values higher than the one it allocates for that new row. If there are no higher values already, and insert works and merge continues to fail consistently, then it sounds like something you'd need to raise with EXASolution.
Related
I am trying to write a DB2 query that allows me to either update a record if it already exists but if it does not exist it should be inserted. I wrote the following query that should accomplish this:
MERGE INTO OA1P.TLZ712A1 AS PC
USING (
SELECT * FROM OA1P.TLZ712A1
WHERE CALENDAR_ID=13 AND
"PACKAGE"='M2108'
) PC2
ON (PC.ID_PACKAGE_CALENDAR=PC2.ID_PACKAGE_CALENDAR)
WHEN MATCHED THEN
UPDATE SET ACT_DATE = '31.12.2021'
WHEN NOT MATCHED THEN
INSERT ("PACKAGE", ACT_DATE, CALENDAR_ID, PREPTA, MIXED) VALUES ('M2108', '31.12.2021', 13, 0, 0)
This query should attempt to check if a record already exists for the selection criteria. Updating a record seems to be working fine but I am not able to get the "WHEN NOT MATCHED" part to work and inserting a new record. Anyone able to provide some assistance?
The table is used to save the activation date of a certain software package. PACKAGE is the reference to the package table containing the name of the package (eg. "M2108"). CALENDAR_ID refers to a system where the software package will be activated. The actual date is stored in ACT_DATE.
Did not manage to get the DDL into SQLFiddle so I have to provide it here:
CREATE TABLE OA1P.TLZ712A1 (
ID_PACKAGE_CALENDAR INTEGER DEFAULT IDENTITY GENERATED BY DEFAULT NOT NULL,
CALENDAR_ID INTEGER,
"PACKAGE" VARCHAR(10) NOT NULL,
ACT_DATE DATE NOT NULL,
PREPTA SMALLINT DEFAULT 0 NOT NULL,
MIXED SMALLINT DEFAULT 0 NOT NULL,
"COMMENT" VARCHAR(60) NOT NULL,
LAST_MODIFIED_PID CHAR(7) NOT NULL,
ST_STARTID TIMESTAMP NOT NULL,
ST_FROM TIMESTAMP NOT NULL,
ST_TO TIMESTAMP NOT NULL,
CONSTRAINT TLZ712A1_PK PRIMARY KEY (ID_PACKAGE_CALENDAR),
CONSTRAINT CALENDAR FOREIGN KEY (CALENDAR_ID) REFERENCES OA1P.TLZ711A1(ID_CALENDAR) ON DELETE RESTRICT,
CONSTRAINT "PACKAGE" FOREIGN KEY ("PACKAGE") REFERENCES OA1P.TLZ716A1(NAME) ON DELETE RESTRICT
);
CREATE UNIQUE INDEX ILZ712A0 ON OA1P.TLZ712A1 (ID_PACKAGE_CALENDAR);
If your goal is to set ACT_DATE to 31.12.2021 if a row is found with PACKAGE = M2108 and CALENDAR_ID = 13 and if no row is found with these values then insert it, then this could be the answer
MERGE INTO OA1P.TLZ712A1 AS PC
USING (
VALUES ('M2108', 13, date '31.12.2021')
) PC2 ("PACKAGE", CALENDAR_ID, ACT_DATE)
ON (PC."PACKAGE", PC.CALENDAR_ID) = (PC2."PACKAGE", PC2.CALENDAR_ID)
WHEN MATCHED THEN
UPDATE SET ACT_DATE = PC2.ACT_DATE
WHEN NOT MATCHED THEN
INSERT ("PACKAGE", ACT_DATE, CALENDAR_ID, PREPTA, MIXED) VALUES (PC2."PACKAGE", PC2.ACT_DATE, PC2.CALENDAR_ID, 0, 0)
The code is the category of the video, it is represented by one upper case character, excluding I, O,
Q, V, Y and Z, followed by a numeric character.
So far, I took a guess and got this. Any suggestions on how to fix it?
create table channelTable (
channelID number NOT NULL,
ChannelName varchar(100) NOT NULL,
ChannelDate date NOT NULL,
UserName varchar(100) NOT NULL UNIQUE,
TopicCode varchar(4) NOT NULL);
CONSTRAINT channelID_pk PRIMARY KEY (channelID)
CONSTRAINT c_topicCode LIKE '[A-Za-z][0-9] NOT (I,O,Q,N,Y,Z)
);
Some comments:
NOT NULL is not needed for PRIMARY KEY columns.
In Oracle, use VARCHAR2().
Then, I would suggests regular expressions. If the value is supposed to be exactly two characters, then declare it as such:
create table channelTable (
channelID number,
ChannelName varchar(100) NOT NULL,
ChannelDate date NOT NULL,
UserName varchar2(100) NOT NULL UNIQUE,
TopicCode char(2) NOT NULL;
CONSTRAINT channelID_pk PRIMARY KEY (channelID)
CONSTRAINT check (REGEXP_LIKE(c_topicCode, '^[A-HJ-NPR-UYZ][0-9]$')
);
Or perhaps more simply:
CONSTRAINT REGEXP_LIKE(c_topicCode, '^[A-Z][0-9]$') AND NOT REGEXP_LIKE(c_topicCode, '^[IOQNYZ]'))
All that said, I would rather see a table of TopicCodes that is populated with the correct values. Then you can just use a foreign key relationship to define the appropriate codes.
Use the regular expression ^[A-HJ-MPR-X]\d$ to match an upper-case character excluding I,O,Q,N,Y,Z followed by a digit:
CREATE TABLE channels (
id number CONSTRAINT channel__id__pk PRIMARY KEY,
Name varchar(100) CONSTRAINT channel__name__nn NOT NULL,
DateTime date CONSTRAINT channel__date__nn NOT NULL,
UserName varchar(100) CONSTRAINT channel__username__NN NOT NULL
CONSTRAINT channel__username__U UNIQUE,
TopicCode varchar(4),
CONSTRAINT channel__topiccode__chk CHECK ( REGEXP_LIKE( topiccode, '^[A-HJ-MPR-X]\d$' ) )
);
db<>fiddle
Also, you don't need to call the table channeltable just call it channels and you don't need to prefix the column names with the table name and you can name all the constraints (rather than relying on system generated constraint names which makes it much harder to track down issues when you are debugging).
Consider the following check constrait:
create table channelTable (
...
topicCode varchar(4) not null
check(
substr(c_topicCode, 1, 1) not in ('I', 'O', 'Q', 'V', 'Y', 'Z')
and regexp_like(topicCode, '^[A-Z]\d')
),
...
);
The first condition ensures that the code does not start with one of the forbidden characters, the second valides that it stats with an upper alphabetic character, followed by a number.
To avoid using two conditions, an alternative would be to list all allowed characters in the first position:
check(regexp_like(topicCode, '^[ABCDEFGHJKLMNPRSTUVWX]\d'))
This works in Oracle, and in very recent versions of MySQL.
I am experimenting with PostgreSQL coming from SQL using MySQL and I simply wish to create a table with this piece of code which is valid SQL:
CREATE TABLE flat_10
(
pk_flat_id INT(30) DEFAULT 1,
rooms INT(10) UNSIGNED NOT NULL,
room_label CHAR(1) NOT NULL,
PRIMARY KEY (flat_id)
);
I get the error
ERROR: syntax error at or near "("
LINE 3: pk_flat_id integer(30) DEFAULT 1,
I have conducted searches on the web and found no answer and I cant seem to find an answer in the PostgreSQL manual. What am I doing wrong?
I explicitly want to set a limit to the number of digits that can be inserted into the "pk_flat_id" field
I explicitly want to set a limit to the number of digits that can be inserted into the "pk_flat_id" field
Your current table definition does not impose a "size limit" in any way. In MySQL the parameter for the intdata type is only a hint for applications on the display width of the column when displaying it.
You can store the value 2147483647 in an int(1) without any problems.
If you want to limit the values to be stored in an integer column you can use a check constraint:
CREATE TABLE flat_10
(
pk_flat_id bigint DEFAULT 1,
rooms integer NOT NULL,
room_label CHAR(1) NOT NULL,
PRIMARY KEY (flat_id),
constraint valid_number
check (pk_flat_id <= 999999999)
);
The answer is that you use numeric or decimal types. These are documented here.
Note that these types can take an optional precision argument, but you don't want that. So:
CREATE TABLE flat_10
(
pk_flat_id DECIMAL(30) DEFAULT 1,
rooms DECIMAL(10) NOT NULL,
room_label CHAR(1) NOT NULL,
PRIMARY KEY (pk_flat_id)
);
Here is a SQL Fiddle.
I don't think that Postgres supports unsigned decimals. And, it seems like you really want serial types for your keys and the long number of digits is superfluous.
Changing integer to numeric works.
CREATE TABLE flat_10
(
pk_flat_id bigint DEFAULT 1,
rooms numeric NOT NULL,
room_label CHAR(1) NOT NULL,
);
How can I enforce a table to have only one row? Below is what I tried. The UPDATE trigger might work, however, the CREATE trigger definitely will not. For the CREATE, I would like to use SET, however, SET is not supported by SQLite.
CREATE TABLE IF NOT EXISTS `config` (
`id` TINYINT NOT NULL DEFAULT 0,
`subdomain` VARCHAR(45) NOT NULL,
`timezone` CHAR(3) NOT NULL,
`timeout` TINYINT NOT NULL,
`offline` TINYINT NOT NULL,
`hash_config` CHAR(32) NOT NULL,
`hash_points` CHAR(32) NOT NULL,
PRIMARY KEY (`id`));
INSERT INTO config(id,subdomain,timezone,timeout,offline,hash_config,hash_points) VALUES(0,'subdomain','UTC',5,0,'hash_config','hash_points');
CREATE TRIGGER `config_insert_zero`
BEFORE INSERT ON `config`
FOR EACH ROW
BEGIN
-- SET NEW.id=0;
NEW.id=OLD.id;
END;
CREATE TRIGGER `config_update_zero`
BEFORE UPDATE ON `config`
FOR EACH ROW
BEGIN
-- SET NEW.id=0;
NEW.id=OLD.id;
END;
In the general case, to limit the number of rows in a table, you have to prevent any further insert.
In SQLite, this is done with RAISE():
CREATE TRIGGER config_no_insert
BEFORE INSERT ON config
WHEN (SELECT COUNT(*) FROM config) >= 1 -- limit here
BEGIN
SELECT RAISE(FAIL, 'only one row!');
END;
However, if the limit is one, you could instead simply constrain the primary key to a fixed value:
CREATE TABLE config (
id INTEGER PRIMARY KEY CHECK (id = 0),
[...]
);
One idea you may want to consider is to make it appear like the table has only one row. In reality, you keep all previous rows because it's quite possible you will one day want to maintain a history of all past values.
Since there is only one row, there really is no need for an ID column, the purpose of which is to uniquely differentiate each row from all the others. You do need, however, a timestamp which will be used to identify the "one row" which will be the latest row written to the table.
CREATE TABLE `config_history` (
`created` timestamp default current_timestamp,
`subdomain` VARCHAR(45) NOT NULL,
`timezone` CHAR(3) NOT NULL,
`timeout` TINYINT NOT NULL,
`offline` TINYINT NOT NULL,
`hash_config` CHAR(32) NOT NULL,
`hash_points` CHAR(32) NOT NULL,
PRIMARY KEY (`created`)
);
Since you are normally interested in only the last row written (the latest version), the query selects the row with the latest creation date:
select ch.created effective_date, ch.subdomain, ch.timezone, ch.timeout,
ch.offline, ch.hash_config, ch.hash_points
from config_history ch
where ch.created =(
select max( created )
from config_history );
Put a create view config as in front of this query and you have a view that selects only one row, the latest, from the table. Any query against the view returns the one row:
select *
from config;
An instead of trigger on the view can convert Updates to Inserts -- you don't actually want to change a value, just write a new row with the new values. This then becomes the new "current" row.
Now you have what appears to be a table with only one row but you also maintain a complete history of all the past changes ever made to that row.
I have a table with the following definition:
CREATE TABLE url_tracker (
id int not null identity(1, 1),
active bit not null,
install_date int not null,
partner_url nvarchar(512) not null,
local_url nvarchar(512) not null,
public_url nvarchar(512) not null,
primary key(id)
);
And I have a requirement that these three URLs always be unique - any individual URL can appear many times, but the combination of the three must be unique (for a given day).
Initially I thought I could do this:
CREATE UNIQUE INDEX uniques ON url_tracker
(install_date, partner_url, local_url, public_url);
However this gives me back the warning:
Warning! The maximum key length is 900 bytes. The index 'uniques' has maximum
length of 3076 bytes. For some combination of large values, the insert/update
operation will fail.
Digging around I learned about the INCLUDE argument to CREATE INDEX, but according to this question converting the command to use INCLUDE will not enforce uniqueness on the URLs.
CREATE UNIQUE INDEX uniques ON url_tracker (install_date)
INCLUDE (partner_url, local_url, public_url);
How can I enforce uniqueness on several relatively large nvarchar fields?
Resolution
So from the comments and answers and more research I'm concluding I can do this:
CREATE TABLE url_tracker (
id int not null identity(1, 1),
active bit not null,
install_date int not null,
partner_url nvarchar(512) not null,
local_url nvarchar(512) not null,
public_url nvarchar(512) not null,
uniquehash AS HashBytes('SHA1',partner_url+local_url+public_url) PERSISTED,
primary key(id)
);
CREATE UNIQUE INDEX uniques ON url_tracker (install_date,uniquehash);
Thoughts?
I would make a computed column with the hash of the URLs, then make a unique index/constraint on that. Consider making the hash a persisted computed column. It shouldn't have to be recalculated after insertion.
Following the ideas from the conversation in the comments. Assuming that you can change the datatype of the URL to be VARCHAR(900) (or NVARCHAR(450) if you really think you need Unicode URLs) and be happy with the limitation on the length of the URL, this solution could work. This also assumes SQL Server 2008 or better. Please, always specify what version you're working with; sql-server is not specific enough, since solutions can vary greatly depending on the version.
Setup:
USE tempdb;
GO
CREATE TABLE dbo.urls
(
id INT IDENTITY(1,1) PRIMARY KEY,
url VARCHAR(900) NOT NULL UNIQUE
);
CREATE TABLE dbo.url_tracker
(
id INT IDENTITY(1,1) PRIMARY KEY,
active BIT NOT NULL DEFAULT 1,
install_date DATE NOT NULL DEFAULT CURRENT_TIMESTAMP,
partner_url_id INT NOT NULL REFERENCES dbo.urls(id),
local_url_id INT NOT NULL REFERENCES dbo.urls(id),
public_url_id INT NOT NULL REFERENCES dbo.urls(id),
CONSTRAINT unique_urls UNIQUE
(
install_date,partner_url_id, local_url_id, public_url_id
)
);
Insert some URLs:
INSERT dbo.urls(url) VALUES
('http://msn.com/'),
('http://aol.com/'),
('http://yahoo.com/'),
('http://google.com/'),
('http://gmail.com/'),
('http://stackoverflow.com/');
Now let's insert some data:
-- succeeds:
INSERT dbo.url_tracker(partner_url_id, local_url_id, public_url_id)
VALUES (1,2,3), (2,3,4), (3,4,5), (4,5,6);
-- fails:
INSERT dbo.url_tracker(partner_url_id, local_url_id, public_url_id)
VALUES(1,2,3);
GO
/*
Msg 2627, Level 14, State 1, Line 3
Violation of UNIQUE KEY constraint 'unique_urls'. Cannot insert duplicate key
in object 'dbo.url_tracker'. The duplicate key value is (2011-09-15, 1, 2, 3).
The statement has been terminated.
*/
-- succeeds, since it's for a different day:
INSERT dbo.url_tracker(install_date, partner_url_id, local_url_id, public_url_id)
VALUES('2011-09-01',1,2,3);
Cleanup:
DROP TABLE dbo.url_tracker, dbo.urls;
Now, if 900 bytes is not enough, you could change the URL table slightly:
CREATE TABLE dbo.urls
(
id INT IDENTITY(1,1) PRIMARY KEY,
url VARCHAR(2048) NOT NULL,
url_hash AS CONVERT(VARBINARY(32), HASHBYTES('SHA1', url)) PERSISTED,
CONSTRAINT unique_url UNIQUE(url_hash)
);
The rest doesn't have to change. And if you try to insert the same URL twice, you get a similar violation, e.g.
INSERT dbo.urls(url) SELECT 'http://www.google.com/';
GO
INSERT dbo.urls(url) SELECT 'http://www.google.com/';
GO
/*
Msg 2627, Level 14, State 1, Line 1
Violation of UNIQUE KEY constraint 'unique_url'. Cannot insert duplicate key
in object 'dbo.urls'. The duplicate key value is
(0xd111175e022c19f447895ad6b72ff259552d1b38).
The statement has been terminated.
*/