SnowConvert AI - Oracle - Sélectionner¶

In this section you could find information about the select query syntax and its conversions.

Note

Certaines parties des codes de sortie sont omises pour des raisons de clarté.

Traduction de sélection globale¶

Sélection simple¶

Oracle :¶

select * from table1;
select col1 from schema1.table1;

Snowflake :¶

select * from
table1;

select col1 from
schema1.table1;

Clause Where¶

Oracle :¶

select col1 from schema1.table1 WHERE col1 = 1 and id > 0 or id < 1;

Snowflake :¶

select col1 from
schema1.table1
WHERE col1 = 1 and id > 0 or id < 1;

Clause Order by¶

Oracle :¶

select col1 from schema1.table1 order by id ASC;

Snowflake :¶

select col1 from
schema1.table1
order by id ASC;

Group by¶

Oracle :¶

select col1 from schema1.table1 GROUP BY id;

Snowflake :¶

select col1 from
schema1.table1
GROUP BY id;

Clause de modèle¶

La clause de modèle n’est pas encore prise en charge.

Clause de limitation de ligne¶

Oracle :¶

-- Using ONLY
select * from TableFetch1 FETCH FIRST 2 ROWS ONLY;
select * from TableFetch1 FETCH FIRST 20 percent ROWS ONLY;
select * from TableFetch1 order by col1 FETCH FIRST 2 ROWS with ties;
select * from TableFetch1 order by col1 FETCH FIRST 20 percent ROWS with ties;

-- Using OFFSET clause
select * from TableFetch1 offset 2 rows FETCH FIRST 2 ROWS ONLY;
select * from TableFetch1 offset 2 rows FETCH FIRST 60 percent rows ONLY;
select * from TableFetch1 
order by col1 offset 2 rows FETCH NEXT 2 ROWs with ties;
select * from TableFetch1 
order by col1 offset 2 rows FETCH FIRST 60 percent ROWs with ties;

-- Using WITH TIES clause
select * from TableFetch1 FETCH FIRST 2 ROWS with ties;
select * from TableFetch1 FETCH FIRST 20 percent ROWS with ties;
select * from TableFetch1 offset 2 rows FETCH NEXT 2 ROWs with ties;
select * from TableFetch1 offset 2 rows FETCH FIRST 60 percent ROWs with ties;

-- Using ORDER BY clause
select * from TableFetch1 order by col1 FETCH FIRST 2 ROWS ONLY;
select * from TableFetch1 order by col1 FETCH FIRST 20 percent ROWS ONLY;
select * from TableFetch1 order by col1 offset 2 rows FETCH FIRST 2 ROWS ONLY;
select * from TableFetch1 
order by col1 offset 2 rows FETCH FIRST 60 percent ROWS ONLY;

select * from TableFetch1 FETCH FIRST ROWS ONLY;

select * from TableFetch1 offset 2 rows;

Snowflake :¶

-- Using ONLY
select * from
TableFetch1
FETCH FIRST 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
ORDER BY
NULL) - 1) / COUNT(*) OVER () < 20 / 100;

select * from
TableFetch1
QUALIFY
RANK() OVER (
order by col1) <= 2;

select * from
TableFetch1
QUALIFY
(RANK() OVER (
order by col1) - 1) / COUNT(*) OVER () < 20 / 100;

-- Using OFFSET clause
select * from
TableFetch1
offset 2 rows FETCH FIRST 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
ORDER BY
NULL) - 1 - 2) / COUNT(*) OVER () < 60 / 100
LIMIT NULL OFFSET 2;

select * from
TableFetch1
QUALIFY
RANK() OVER (
order by col1) - 2 <= 2
LIMIT NULL OFFSET 2;

select * from
TableFetch1
QUALIFY
(RANK() OVER (
order by col1) - 1 - 2) / COUNT(*) OVER () < 60 / 100
LIMIT NULL OFFSET 2;

-- Using WITH TIES clause
select * from
TableFetch1
FETCH FIRST 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
ORDER BY
NULL) - 1) / COUNT(*) OVER () < 20 / 100;

select * from
TableFetch1
offset 2 rows FETCH NEXT 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
ORDER BY
NULL) - 1 - 2) / COUNT(*) OVER () < 60 / 100
LIMIT NULL OFFSET 2;

-- Using ORDER BY clause
select * from
TableFetch1
order by col1
FETCH FIRST 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
order by col1) - 1) / COUNT(*) OVER () < 20 / 100;

select * from
TableFetch1
order by col1 offset 2 rows FETCH FIRST 2 ROWS ONLY;

select * from
TableFetch1
QUALIFY
(ROW_NUMBER() OVER (
order by col1) - 1 - 2) / COUNT(*) OVER () < 60 / 100
LIMIT NULL OFFSET 2;

select * from
TableFetch1
FETCH FIRST 1 ROWS ONLY;

select * from
TableFetch1
LIMIT NULL OFFSET 2;

Note

Dans Oracle, le FETCH / OFFSET WITH TIES est ignoré si aucun ORDER BY n’est spécifié dans le SELECT. Ce cas sera transformé en FETCH / OFFSET avec le mot-clé ONLY dans Snowflake ; notez que dans Snowflake, le mot-clé ONLY n’a aucun effet dans les résultats et n’est utilisé que pour des raisons de lisibilité.

Pivot¶

Snowflake ne prend pas en charge les instructions suivantes :\ - Rename columns\ - Multiple Columns

Oracle :¶

select * from schema1.table1
PIVOT(count(*) as count1 FOR (column1, column2) IN (row1 as rowName));

Snowflake :¶

select * from
schema1.table1
!!!RESOLVE EWI!!! /*** SSC-EWI-0015 - PIVOT/UNPIVOT RENAME COLUMN NOT SUPPORTED ***/!!!
PIVOT (count(*)
                !!!RESOLVE EWI!!! /*** SSC-EWI-0015 - PIVOT/UNPIVOT MULTIPLE COLUMN NOT SUPPORTED ***/!!!
                FOR (column1, column2)
!!!RESOLVE EWI!!! /*** SSC-EWI-0015 - PIVOT/UNPIVOT RENAME COLUMN NOT SUPPORTED ***/!!!
IN (row1 as rowName));

Unpivot¶

Snowflake ne pend pas en charge les instructions suivantes :\ \ - INCLUDE / EXCLUDE NULLS

Oracle :¶

select * from schema1.table1 
UNPIVOT INCLUDE NULLS (column1 FOR column2 IN (ANY, ANY)); 

Snowflake :¶

select * from
schema1.table1
!!!RESOLVE EWI!!! /*** SSC-EWI-0015 - PIVOT/UNPIVOT INCLUDE NULLS NOT SUPPORTED ***/!!!
UNPIVOT ( column1 FOR column2 IN (
ANY,
ANY));

Transformation de JOIN (+) en syntaxe ANSI¶

Danger

Cette traduction est actuellement désactivée et n’est destinée qu’à servir de référence pour les traductions effectuées avec les versions précédentes de SnowConvert AI. Pour la traduction actuelle, consultez la section ci-dessus.

SnowConvert AI traduit la syntaxe de jointure extérieure spéciale (+) NON-ANSI en syntaxe de jointure extérieure ANSI. Cette sous-section présente quelques exemples :

Vers LEFT OUTER JOIN¶

Exemple 1 :

Oracle :¶

-- Additional Params: --OuterJoinsToOnlyAnsiSyntax
SELECT d.department_name,
       e.employee_name      
FROM   departments d, employees e
WHERE  d.department_id = e.department_id (+) 
AND    d.department_id >= 30;

Snowflake :¶

SELECT d.department_name,
       e.employee_name
FROM
       departments d
       LEFT OUTER JOIN
              employees e
              ON d.department_id = e.department_id
WHERE
       d.department_id >= 30;

Exemple 2 :

Oracle :¶

-- Additional Params: --OuterJoinsToOnlyAnsiSyntax
SELECT d.department_name,
       e.employee_name      
FROM   departments d, employees e
WHERE  d.department_id(+)  = e.department_id 
AND    d.department_id >= 30;

Snowflake :¶

SELECT d.department_name,
       e.employee_name
FROM
       employees e
       LEFT OUTER JOIN
              departments d
              ON d.department_id = e.department_id
WHERE
       d.department_id >= 30;

Exemple 3 : jointure multiple

Oracle :¶

-- Additional Params: --OuterJoinsToOnlyAnsiSyntax
SELECT d.department_name,
       e.employee_name      
FROM   departments d, employees e, projects p
WHERE  e.department_id(+) = d.department_id
AND    p.department_id(+) = d.department_id
AND    d.department_id >= 30;

Snowflake :¶

SELECT d.department_name,
       e.employee_name
FROM
       departments d
       LEFT OUTER JOIN
              employees e
              ON e.department_id = d.department_id
       LEFT OUTER JOIN
              projects p
              ON p.department_id = d.department_id
WHERE
       d.department_id >= 30;

Exemple 4 : Jointure avec d’autres types de conditionnels

Oracle :¶

-- Additional Params: --OuterJoinsToOnlyAnsiSyntax
SELECT d.department_name,
       e.employee_name      
FROM   departments d, employees e
WHERE  d.department_id(+)  = e.department_id 
AND    d.location(+) IN ('CHICAGO', 'BOSTON', 'NEW YORK')
AND    d.department_id >= 30;

Snowflake :¶

SELECT d.department_name,
       e.employee_name
FROM
       employees e
       LEFT OUTER JOIN
              departments d
              ON d.department_id = e.department_id
              AND d.location IN ('CHICAGO', 'BOSTON', 'NEW YORK')
WHERE
       d.department_id >= 30;

Exemple 5 : Jointure avec (+) dans une fonction

Oracle :¶

-- Additional Params: --OuterJoinsToOnlyAnsiSyntax
SELECT d.department_name,
       e.employee_name
FROM   departments d, employees e
WHERE SUBSTR(d.department_name, 1, NVL(e.department_id, 1) ) = e.employee_name(+);

Snowflake :¶

SELECT d.department_name,
       e.employee_name
FROM
       departments d
       LEFT OUTER JOIN
              employees e
              ON SUBSTR(d.department_name, 1, NVL(e.department_id, 1) ) = e.employee_name;

Avertissement

Notez que certains des modèles traduits en LEFT OUTER JOIN peuvent récupérer les lignes dans un ordre différent.

Vers CROSSJOIN¶

Exemple 6 : Cas complexe nécessitant l’utilisation de CROSS JOIN

Oracle :¶

SELECT d.department_name,
       e.employee_name,
       p.project_name,
       c.course_name 
FROM   departments d, employees e, projects p, courses c
WHERE
e.salary (+) >= 2000 AND
d.department_id = e.department_id (+)
AND p.department_id = e.department_id(+)
AND c.course_id  = e.department_id(+)
AND d.department_id >= 30;

Snowflake :¶

SELECT d.department_name,
       e.employee_name,
       p.project_name,
       c.course_name
FROM
       departments d
       CROSS JOIN projects p
       CROSS JOIN courses c
       LEFT OUTER JOIN
              employees e
              ON
              e.salary >= 2000
              AND
              d.department_id = e.department_id
              AND p.department_id = e.department_id
              AND c.course_id  = e.department_id
WHERE
       d.department_id >= 30;

Requêtes hiérarchiques¶

Hierarchical queries in Snowflake allow you to organize and retrieve data in a tree-like structure, typically using the CONNECT BY clause. This clause joins a table to itself to process hierarchical data in the table.

Modèles d’échantillons de sources¶

Oracle :¶

SELECT employee_ID, manager_ID, title
FROM employees
START WITH manager_ID = 1
CONNECT BY manager_ID = PRIOR employee_id;

Snowflake :¶

SELECT employee_ID, manager_ID, title
FROM
employees
START WITH manager_ID = 1
CONNECT BY
manager_ID = PRIOR employee_id;

Requête Select Flashback¶

Description¶

Oracle

La requête flashback utilisée dans Oracle permet de récupérer les données passées d’une table, d’une vue ou d’une vue matérialisée. Dans Oracle, les utilisations peuvent être les suivantes :

La restauration de données supprimées ou l’annulation d’un commit incorrect, la comparaison des données actuelles avec les données correspondantes à un moment antérieur, la vérification de l’état des données transactionnelles à un moment donné et les outils de génération de rapports sur les données passées, entre autres. (Documentation sur les requêtes Oracle Flashback).

Snowflake

Le mécanisme équivalent dans Snowflake pour interroger les données du passé est la requête AT | BEGIN. Notez que la seule équivalence concerne les instructions AS OF.

En outre, Snowflake dispose d’une documentation complète sur Time Travel qui permet de cloner des objets tels que des tables, des vues et des schémas dans le cadre de requêtes portant sur des données. Les jours d’accès aux données passées ou supprimées sont limités (90 jours avant le passage au statut Fail-safe). Pour plus d’informations, consultez la documentation sur Snowflake Time Travel.

Syntaxe Oracle

{ VERSIONS BETWEEN
  { SCN | TIMESTAMP }
  { expr | MINVALUE } AND { expr | MAXVALUE }
| AS OF { SCN | TIMESTAMP } expr
}

Syntaxe Snowflake

SELECT ...
FROM ...
  {
   AT( { TIMESTAMP => <timestamp> | OFFSET => <time_difference> | STATEMENT => <id> | STREAM => '<name>' } ) |
   BEFORE( STATEMENT => <id> )
  }
[ ... ]

Note

Notez que l’ID de requête doit faire référence à une requête exécutée dans les 14 jours. Si l’ID de requête fait référence à une requête datant de plus de 14 jours, l’erreur suivante est renvoyée : Erreur : instruction <query_id> introuvable. Pour contourner cette limite, utilisez l’horodatage de la requête référencée. (Snowflake AT | Documentation Before)

Modèles d’échantillons de sources¶

Les données suivantes sont utilisées dans les exemples suivants pour générer les sorties de requête.

Oracle¶

CREATE TABLE Employee (
    EmployeeID NUMBER PRIMARY KEY,
    FirstName VARCHAR2(50),
    LastName VARCHAR2(50),
    EmailAddress VARCHAR2(100),
    HireDate DATE,
    SalaryAmount NUMBER(10, 2)
);

INSERT INTO Employee VALUES (1, 'Bob', 'SampleNameA', 'sample@example.com', TO_DATE('2023-01-15', 'YYYY-MM-DD'), 11111.00);
INSERT INTO Employee VALUES (2, 'Bob', 'SampleNameB', 'sample@example.com', TO_DATE('2023-01-15', 'YYYY-MM-DD'), 11111.00);
INSERT INTO Employee VALUES (3, 'Bob', 'SampleNameC', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);
INSERT INTO Employee VALUES (4, 'Bob', 'SampleNameD', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);
INSERT INTO Employee VALUES (5, 'Bob', 'SampleNameE', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);

Snowflake¶

CREATE OR REPLACE TABLE Employee (
       EmployeeID NUMBER(38, 18) /*** SSC-FDM-0006 - NUMBER TYPE COLUMN MAY NOT BEHAVE SIMILARLY IN SNOWFLAKE. ***/ PRIMARY KEY,
       FirstName VARCHAR(50),
       LastName VARCHAR(50),
       EmailAddress VARCHAR(100),
       HireDate TIMESTAMP /*** SSC-FDM-OR0042 - DATE TYPE COLUMN HAS A DIFFERENT BEHAVIOR IN SNOWFLAKE. ***/,
       SalaryAmount NUMBER(10, 2) /*** SSC-FDM-0006 - NUMBER TYPE COLUMN MAY NOT BEHAVE SIMILARLY IN SNOWFLAKE. ***/
   )
   COMMENT = '{"origin":"sf_sc","name":"snowconvert","version":{"major":1, "minor":0},"attributes":{"component":"oracle"}}'
   ;

   INSERT INTO Employee
   VALUES (1, 'Bob', 'SampleNameA', 'sample@example.com', TO_DATE('2023-01-15', 'YYYY-MM-DD'), 11111.00);

   INSERT INTO Employee
   VALUES (2, 'Bob', 'SampleNameB', 'sample@example.com', TO_DATE('2023-01-15', 'YYYY-MM-DD'), 11111.00);

   INSERT INTO Employee
   VALUES (3, 'Bob', 'SampleNameC', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);

   INSERT INTO Employee
   VALUES (4, 'Bob', 'SampleNameD', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);

   INSERT INTO Employee
   VALUES (5, 'Bob', 'SampleNameE', 'sample@example.com', TO_DATE('2022-03-10', 'YYYY-MM-DD'), 11111.00);

1. AS OF with TIMESTAMP case¶

Oracle¶

SELECT * FROM employees
AS OF TIMESTAMP
TO_TIMESTAMP('2023-09-27 07:00:00', 'YYYY-MM-DD HH:MI:SS')
WHERE last_name = 'SampleName';

Snowflake¶

SELECT * FROM
employees
!!!RESOLVE EWI!!! /*** SSC-EWI-OR0135 - DATA RETENTION PERIOD MAY PRODUCE NO RESULTS ***/!!!
AT (TIMESTAMP =>
TO_TIMESTAMP('2023-09-27 07:00:00', 'YYYY-MM-DD HH:MI:SS'))
WHERE last_name = 'SampleName';

2. AS OF with SCN case¶

Oracle¶

SELECT * FROM employees
AS OF SCN
TO_TIMESTAMP('2023-09-27 07:00:00', 'YYYY-MM-DD HH:MI:SS')
WHERE last_name = 'SampleName';

Snowflake¶

SELECT * FROM
employees
!!!RESOLVE EWI!!! /*** SSC-EWI-0040 - THE 'FLASHBACK QUERY' CLAUSE IS NOT SUPPORTED IN SNOWFLAKE ***/!!!
AS OF SCN
TO_TIMESTAMP('2023-09-27 07:00:00', 'YYYY-MM-DD HH:MI:SS')
WHERE last_name = 'SampleName';

Problèmes connus¶

L’option n’est pas pris en charge lorsqu’elle utilise SCN.
L’instruction VERSION n’est pas prise en charge dans Snowflake.

EWIs connexes¶

SSC-EWI-0040: Instruction non prise en charge.
SSC-EWI-OR0135: La clause actuelle n’est pas prise en charge dans Snowflake.
SSC-FDM-0006: La colonne de type nombre peut ne pas se comporter de la même manière dans Snowflake.
SSC-FDM-OR0042: Le type de date transformé en horodatage a un comportement différent.