Global {#pragmas}

Scope

Unless otherwise specified, a pragma affects all the subsequent expressions up to the end of the module where it's used. If necessary and logically possible, you can change the value of this setting several times within a given query to make it different at different execution steps.

There are also special scoped pragmas with the scope defined by the same rules as the scope of named expressions.

Unlike scoped pragmas, global pragmas can only be used in the global scope (not inside DEFINE ACTION and DEFINE SUBQUERY).

AutoCommit

Automatically perform COMMIT after each expression.

TablePathPrefix

Add the specified prefix to the cluster table paths. Works on the same principle as merging paths in a file system: supports references to the parent catalog .. and doesn't require adding a slash to the right. For example,

PRAGMA TablePathPrefix = "home/yql";
SELECT * FROM test;

The prefix is not added if the table name is an absolute path (starts with /).

UDF

Importing all UDFs from the specified library. For the pragma to work, the library must be attached to the query using the icon. Please note that the library must be shared library (.so) and compiled for Linux x64.

When setting a prefix, it's appended before the names of all loaded modules, e.g. CustomPrefixIp::IsIPv4 instead of Ip::IsIPv4. Setting the prefix lets you use different versions of the same UDF.

RuntimeLogLevel

Lets you change the computation logging level (for example, for UDFs) at query runtime or at the UDF signature declaration stage.

UseTablePrefixForEach

EACH uses TablePathPrefix for each list item.

Warning

Action:

• disable: Disable.
• error: Treat as an error.
• default: Revert to the default behavior.

The warning code is returned with the text itself (it's displayed on the right side of the web interface).

Examples:

PRAGMA Warning("error", "*");
PRAGMA Warning("disable", "1101");
PRAGMA Warning("default", "4503");

In this case, all the warnings are treated as errors, except for warnings 1101 (will be disabled) and 4503 (will be processed by default, that is, remain a warning). Since warnings may be added in new YQL releases, use PRAGMA Warning("error", "*"); with caution (at least cover such queries with autotests).

List of warning and error codes

Greetings

Issue the specified text as the query's Info message.

Example: PRAGMA Greetings("It's a good day!");

WarningMsg

Issue the specified text as the query's Warning message.

Example: PRAGMA WarningMsg("Attention!");

DqEngine

When set to "auto", it enables a new compute engine. Computing is made, whenever possible, without creating map/reduce operations. The "force" value unconditionally routes computations to the new engine.

SimpleColumns

SimpleColumns / DisableSimpleColumns

If using SELECT foo.* FROM ... AS foo, delete the foo. prefix from the names of the resulting columns.

It also works for JOIN, but then it may crash if there's a name conflict (which can be resolved through WITHOUT and renaming columns). For JOIN in SimpleColumns mode, an implicit Coalesce is made for key columns: the query SELECT * FROM T1 AS a JOIN T2 AS b USING(key) in the SimpleColumns mode works same as SELECT a.key ?? b.key AS key, ... FROM T1 AS a JOIN T2 AS b USING(key)

CoalesceJoinKeysOnQualifiedAll

CoalesceJoinKeysOnQualifiedAll / DisableCoalesceJoinKeysOnQualifiedAll

Controls implicit Coalesce for the key JOIN columns in the SimpleColumns mode. If the flag is set, key columns are coalesced if there is at least one expression in the format foo.* or * in SELECT: for example, SELECT a.* FROM T1 AS a JOIN T2 AS b USING(key). If the flag isn't set, JOIN keys are coalesced only if there is an asterisk '*' after SELECT.

StrictJoinKeyTypes

StrictJoinKeyTypes / DisableStrictJoinKeyTypes

If the flag is set, JOIN will require a strict match of key types.
By default, JOIN preconverts keys to a shared type, which might result in performance degradation.

StrictJoinKeyTypes is a scoped setting.

AnsiInForEmptyOrNullableItemsCollections

This pragma brings the behavior of the IN operator in accordance with the standard when there's NULL in the left or right side of IN. The behavior of IN when on the right side there is a Tuple with elements of different types also changed. Examples:

1 IN (2, 3, NULL) = NULL (was Just(False))
NULL IN () = Just(False) (was NULL)
(1, null) IN ((2, 2), (3, 3)) = Just(False) (was NULL)

For more information about the IN behavior when operands include NULLs, see here. You can explicitly select the old behavior by specifying the pragma DisableAnsiInForEmptyOrNullableItemsCollections. If no pragma is set, you'll see a warning, and the original variant will be used.

AnsiRankForNullableKeys

Aligns the RANK/DENSE_RANK behavior with the standard if there are optional types in the window sort keys or in the argument of such window functions. It means that:

• The result type is always Uint64 rather than Uint64?;
• NULLs in keys are treated as equal to each other (the current implementation returns NULL).

You can explicitly select the old behavior by specifying the pragma DisableAnsiRankForNullableKeys. If no pragma is set, you'll see a warning, and the original variant will be used.

AnsiCurrentRow

Aligns the implicit setting of a window frame with the standard if there is ORDER BY.

If AnsiCurrentRow is not set, the (ORDER BY key) window is equivalent to (ORDER BY key ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW). The standard requires that such window behaves as (ORDER BY key RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW).

The difference is in CURRENT ROW interpretation. In ROWS mode, CURRENT ROW is interpreted literally: the current string in the partition.
And in RANGE mode, the end of the CURRENT ROW frame means "the last row in the partition with the sorting key equal to the current row".

OrderedColumns

OrderedColumns / DisableOrderedColumns

Output the sequence of columns to SELECT/JOIN/UNION ALL and save it when recording the results. The order of columns is undefined by default.

PositionalUnionAll

Enable the standard columnar execution for UNION ALL. This automatically enables ordered columns.

RegexUseRe2

Use Re2 UDF instead of Pcre for executing REGEX,MATCH,RLIKE, and SQL operators. Re2 UDF supports correct processing of Unicode symbols, unlike Pcre UDF, which is used by default.

ClassicDivision

In the classical version, the result of integer division remains integer (by default). If disabled, the result is always Double.

ClassicDivision is a scoped setting.

CheckedOps

When enabled: if aggregate functions SUM/SUM_IF, binary operations +, -, *, /, %, or unary operation - on integers result in an overflow beyond the range of the target argument or result type, return NULL. When disabled: overflow isn't checked.

Has no effect on floating point or Decimal numbers.

CheckedOps is a scoped setting.

UnicodeLiterals

UnicodeLiterals/DisableUnicodeLiterals

When the mode is enabled, string literals without suffixes like "foo"/'bar'/@@multiline@@ will have the Utf8 type. If disabled, then String.

UnicodeLiterals is a scoped setting.

WarnUntypedStringLiterals

WarnUntypedStringLiterals/DisableWarnUntypedStringLiterals

When the mode is enabled, string literals without a suffix like "foo"/'bar'/@@multiline@@ will prompt a warning. It can be suppressed if you explicitly select the s suffix for the String type or the u suffix for the Utf8 type.

WarnUntypedStringLiterals is a scoped setting.

AllowDotInAlias

Enable dot in names of result columns. This behavior is disabled by default, since the further use of such columns in JOIN is not fully implemented.

WarnUnnamedColumns

Generate a warning if a column name (in the format column[0-9]+) was automatically generated for an unnamed expression in SELECT.

GroupByLimit

Increasing the limit on the number of groups in GROUP BY.

GroupByCubeLimit

Increasing the limit on the number of dimensions in GROUP BY.

Use this option with care: the computational complexity of the query grows exponentially with the number of dimensions.