Skip to content

Commit

Permalink
SQL: Show/desc commands now support table ids (#33363)
Browse files Browse the repository at this point in the history
Extend SHOW TABLES, DESCRIBE and SHOW COLUMNS to support table
identifiers not just SQL LIKE pattern.
This allows both Elasticsearch-style multi-index patterns and SQL LIKE.
To disambiguate between the two (as the " vs ' can be easy to miss),
the grammar now requires LIKE keyword as a prefix for all LIKE-like
patterns.

Also added some docs comparing the two types of patterns.

Fix #33294

(cherry picked from commit 17c7f99)
  • Loading branch information
costin committed Sep 4, 2018
1 parent 9fef77d commit ad5a9b5
Show file tree
Hide file tree
Showing 25 changed files with 1,400 additions and 946 deletions.
8 changes: 4 additions & 4 deletions docs/reference/sql/concepts.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ NOTE: This documentation while trying to be complete, does assume the reader has

As a general rule, {es-sql} as the name indicates provides a SQL interface to {es}. As such, it follows the SQL terminology and conventions first, whenever possible. However the backing engine itself is {es} for which {es-sql} was purposely created hence why features or concepts that are not available, or cannot be mapped correctly, in SQL appear
in {es-sql}.
Last but not least, {es-sql} tries to obey the https://en.wikipedia.org/wiki/Principle_of_least_astonishment[principle of least suprise], though as all things in the world, everything is relative.
Last but not least, {es-sql} tries to obey the https://en.wikipedia.org/wiki/Principle_of_least_astonishment[principle of least surprise], though as all things in the world, everything is relative.

=== Mapping concepts across SQL and {es}

Expand All @@ -26,7 +26,7 @@ So let's start from the bottom; these roughly are:
|`column`
|`field`
|In both cases, at the lowest level, data is stored in _named_ entries, of a variety of <<sql-data-types, data types>>, containing _one_ value. SQL calls such an entry a _column_ while {es} a _field_.
Notice that in {es} a field can contain _multiple_ values of the same type (esentially a list) while in SQL, a _column_ can contain _exactly_ one value of said type.
Notice that in {es} a field can contain _multiple_ values of the same type (essentially a list) while in SQL, a _column_ can contain _exactly_ one value of said type.
{es-sql} will do its best to preserve the SQL semantic and, depending on the query, reject those that return fields with more than one value.

|`row`
Expand All @@ -43,7 +43,7 @@ Notice that in {es} a field can contain _multiple_ values of the same type (esen

|`catalog` or `database`
|`cluster` instance
|In SQL, `catalog` or `database` are used interchangebly and represent a set of schemas that is, a number of tables.
|In SQL, `catalog` or `database` are used interchangeably and represent a set of schemas that is, a number of tables.
In {es} the set of indices available are grouped in a `cluster`. The semantics also differ a bit; a `database` is essentially yet another namespace (which can have some implications on the way data is stored) while an {es} `cluster` is a runtime instance, or rather a set of at least one {es} instance (typically running distributed).
In practice this means that while in SQL one can potentially have multiple catalogs inside an instance, in {es} one is restricted to only _one_.

Expand All @@ -62,4 +62,4 @@ Multiple clusters, each with its own namespace, connected to each other in a fed

|===

As one can see while the mapping between the concepts are not exactly one to one and the semantics somewhat different, there are more things in common than differences. In fact, thanks to SQL declarative nature, many concepts can move across {es} transparently and the terminology of the two likely to be used interchangebly through-out the rest of the material.
As one can see while the mapping between the concepts are not exactly one to one and the semantics somewhat different, there are more things in common than differences. In fact, thanks to SQL declarative nature, many concepts can move across {es} transparently and the terminology of the two likely to be used interchangeably through-out the rest of the material.
70 changes: 70 additions & 0 deletions docs/reference/sql/language/index-patterns.asciidoc
Original file line number Diff line number Diff line change
@@ -0,0 +1,70 @@
[role="xpack"]
[testenv="basic"]
[[sql-index-pattern]]
== Index patterns

{es-sql} supports two types of patterns for matching multiple indices or tables:

* {es} multi-index

The {es} notation for enumerating, including or excluding <<multi-index,multi index syntax>>
is supported _as long_ as it is quoted or escaped as a table identifier.

For example:

["source","sql",subs="attributes,callouts,macros"]
----
include-tagged::{sql-specs}/docs.csv-spec[showTablesEsMultiIndex]
----

Notice the pattern is surrounded by double quotes `"`. It enumerated `*` meaning all indices however
it excludes (due to `-`) all indices that start with `l`.
This notation is very convenient and powerful as it allows both inclusion and exclusion, depending on
the target naming convention.

* SQL `LIKE` notation

The common `LIKE` statement (including escaping if needed) to match a wildcard pattern, based on one `_`
or multiple `%` characters.

Using `SHOW TABLES` command again:

["source","sql",subs="attributes,callouts,macros"]
----
include-tagged::{sql-specs}/docs.csv-spec[showTablesLikeWildcard]
----

The pattern matches all tables that start with `emp`.

This command supports _escaping_ as well, for example:

["source","sql",subs="attributes,callouts,macros"]
----
include-tagged::{sql-specs}/docs.csv-spec[showTablesLikeEscape]
----

Notice how now `emp%` does not match any tables because `%`, which means match zero or more characters,
has been escaped by `!` and thus becomes an regular char. And since there is no table named `emp%`,
an empty table is returned.

In a nutshell, the differences between the two type of patterns are:

[cols="^h,^,^",options="header"]
|===
| Feature | Multi index | SQL `LIKE`

| Type of quoting | `"` | `'`
| Inclusion | Yes | Yes
| Exclusion | Yes | No
| Enumeration | Yes | No
| One char pattern | No | `_`
| Multi char pattern | `*` | `%`
| Escaping | No | `ESCAPE`

|===

Which one to use, is up to you however try to stick to the same one across your queries for consistency.

NOTE: As the query type of quoting between the two patterns is fairly similar (`"` vs `'`), {es-sql} _always_
requires the keyword `LIKE` for SQL `LIKE` pattern.

2 changes: 2 additions & 0 deletions docs/reference/sql/language/index.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,8 @@ This chapter describes the SQL semantics supported in X-Pack namely:

<<sql-data-types>>:: Data types
<<sql-commands>>:: Commands
<<sql-index-patterns>>:: Index patterns

include::data-types.asciidoc[]
include::syntax/index.asciidoc[]
include::index-patterns.asciidoc[]
4 changes: 2 additions & 2 deletions docs/reference/sql/language/syntax/describe-table.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -6,14 +6,14 @@
.Synopsis
[source, sql]
----
DESCRIBE table
DESCRIBE [table identifier<1>|[LIKE pattern<2>]]
----

or

[source, sql]
----
DESC table
DESC [table identifier<1>|[LIKE pattern<2>]]
----


Expand Down
9 changes: 8 additions & 1 deletion docs/reference/sql/language/syntax/show-columns.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -6,9 +6,15 @@
.Synopsis
[source, sql]
----
SHOW COLUMNS [ FROM | IN ] ? table
SHOW COLUMNS [ FROM | IN ]? [ table identifier<1> | [ LIKE pattern<2> ] ]
----

<1> single table identifier or double quoted es multi index
<2> SQL LIKE pattern

See <<sql-index-pattern, index patterns>> for more information about
patterns.

.Description

List the columns in table and their data type (and other attributes).
Expand All @@ -17,3 +23,4 @@ List the columns in table and their data type (and other attributes).
----
include-tagged::{sql-specs}/docs.csv-spec[showColumns]
----

2 changes: 1 addition & 1 deletion docs/reference/sql/language/syntax/show-functions.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@
.Synopsis
[source, sql]
----
SHOW FUNCTIONS [ LIKE? pattern<1>? ]?
SHOW FUNCTIONS [ LIKE pattern<1>? ]?
----

<1> SQL match pattern
Expand Down
19 changes: 16 additions & 3 deletions docs/reference/sql/language/syntax/show-tables.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,15 @@
.Synopsis
[source, sql]
----
SHOW TABLES [ LIKE? pattern<1>? ]?
SHOW TABLES [ table identifier<1> | [ LIKE pattern<2> ] ]?
----

<1> SQL match pattern
<1> single table identifier or double quoted es multi index
<2> SQL LIKE pattern

See <<sql-index-pattern, index patterns>> for more information about
patterns.


.Description

Expand All @@ -20,7 +25,15 @@ List the tables available to the current user and their type.
include-tagged::{sql-specs}/docs.csv-spec[showTables]
----

The `LIKE` clause can be used to restrict the list of names to the given pattern.
Match multiple indices by using {es} <<multi-index,multi index syntax>>
notation:

["source","sql",subs="attributes,callouts,macros"]
----
include-tagged::{sql-specs}/docs.csv-spec[showTablesEsMultiIndex]
----

One can also use the `LIKE` clause to restrict the list of names to the given pattern.

The pattern can be an exact match:
["source","sql",subs="attributes,callouts,macros"]
Expand Down
20 changes: 12 additions & 8 deletions x-pack/plugin/sql/src/main/antlr/SqlBase.g4
Original file line number Diff line number Diff line change
Expand Up @@ -50,18 +50,18 @@ statement
)*
')')?
statement #debug
| SHOW TABLES (LIKE? pattern)? #showTables
| SHOW COLUMNS (FROM | IN) tableIdentifier #showColumns
| (DESCRIBE | DESC) tableIdentifier #showColumns
| SHOW FUNCTIONS (LIKE? pattern)? #showFunctions
| SHOW TABLES (tableLike=likePattern | tableIdent=tableIdentifier)? #showTables
| SHOW COLUMNS (FROM | IN) (tableLike=likePattern | tableIdent=tableIdentifier) #showColumns
| (DESCRIBE | DESC) (tableLike=likePattern | tableIdent=tableIdentifier) #showColumns
| SHOW FUNCTIONS (likePattern)? #showFunctions
| SHOW SCHEMAS #showSchemas
| SYS CATALOGS #sysCatalogs
| SYS TABLES (CATALOG LIKE? clusterPattern=pattern)?
(LIKE? tablePattern=pattern)?
| SYS TABLES (CATALOG clusterLike=likePattern)?
(tableLike=likePattern | tableIdent=tableIdentifier)?
(TYPE string (',' string)* )? #sysTables
| SYS COLUMNS (CATALOG cluster=string)?
(TABLE LIKE? indexPattern=pattern)?
(LIKE? columnPattern=pattern)? #sysColumns
(TABLE tableLike=likePattern | tableIdent=tableIdentifier)?
(columnPattern=likePattern)? #sysColumns
| SYS TYPES #sysTypes
| SYS TABLE TYPES #sysTableTypes
;
Expand Down Expand Up @@ -189,6 +189,10 @@ predicate
| IS NOT? kind=NULL
;

likePattern
: LIKE pattern
;

pattern
: value=string patternEscape?
;
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -140,49 +140,53 @@ public void resolveNames(String indexWildcard, String javaRegex, EnumSet<IndexTy
boolean retrieveAliases = CollectionUtils.isEmpty(types) || types.contains(IndexType.ALIAS);
boolean retrieveIndices = CollectionUtils.isEmpty(types) || types.contains(IndexType.INDEX);

String[] indices = Strings.commaDelimitedListToStringArray(indexWildcard);
if (retrieveAliases) {
GetAliasesRequest aliasRequest = new GetAliasesRequest()
.local(true)
.aliases(indexWildcard)
.indices(indices)
.aliases(indices)
.indicesOptions(IndicesOptions.lenientExpandOpen());

client.admin().indices().getAliases(aliasRequest, ActionListener.wrap(aliases ->
resolveIndices(indexWildcard, javaRegex, aliases, retrieveIndices, listener),
resolveIndices(indices, javaRegex, aliases, retrieveIndices, listener),
ex -> {
// with security, two exception can be thrown:
// INFE - if no alias matches
// security exception is the user cannot access aliases

// in both cases, that is allowed and we continue with the indices request
if (ex instanceof IndexNotFoundException || ex instanceof ElasticsearchSecurityException) {
resolveIndices(indexWildcard, javaRegex, null, retrieveIndices, listener);
resolveIndices(indices, javaRegex, null, retrieveIndices, listener);
} else {
listener.onFailure(ex);
}
}));
} else {
resolveIndices(indexWildcard, javaRegex, null, retrieveIndices, listener);
resolveIndices(indices, javaRegex, null, retrieveIndices, listener);
}
}

private void resolveIndices(String indexWildcard, String javaRegex, GetAliasesResponse aliases,
private void resolveIndices(String[] indices, String javaRegex, GetAliasesResponse aliases,
boolean retrieveIndices, ActionListener<Set<IndexInfo>> listener) {

if (retrieveIndices) {
GetIndexRequest indexRequest = new GetIndexRequest()
.local(true)
.indices(indexWildcard)
.indices(indices)
.features(Feature.SETTINGS)
.includeDefaults(false)
.indicesOptions(IndicesOptions.lenientExpandOpen());

client.admin().indices().getIndex(indexRequest,
ActionListener.wrap(indices -> filterResults(indexWildcard, javaRegex, aliases, indices, listener),
ActionListener.wrap(response -> filterResults(javaRegex, aliases, response, listener),
listener::onFailure));
} else {
filterResults(indexWildcard, javaRegex, aliases, null, listener);
filterResults(javaRegex, aliases, null, listener);
}
}

private void filterResults(String indexWildcard, String javaRegex, GetAliasesResponse aliases, GetIndexResponse indices,
private void filterResults(String javaRegex, GetAliasesResponse aliases, GetIndexResponse indices,
ActionListener<Set<IndexInfo>> listener) {

// since the index name does not support ?, filter the results manually
Expand Down Expand Up @@ -302,7 +306,6 @@ private static GetIndexRequest createGetIndexRequest(String index) {
return new GetIndexRequest()
.local(true)
.indices(Strings.commaDelimitedListToStringArray(index))
.features(Feature.MAPPINGS)
//lenient because we throw our own errors looking at the response e.g. if something was not resolved
//also because this way security doesn't throw authorization exceptions but rather honours ignore_unavailable
.indicesOptions(IndicesOptions.lenientExpandOpen());
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -121,12 +121,14 @@ public Command visitExplain(ExplainContext ctx) {

@Override
public Object visitShowFunctions(ShowFunctionsContext ctx) {
return new ShowFunctions(source(ctx), visitPattern(ctx.pattern()));
return new ShowFunctions(source(ctx), visitLikePattern(ctx.likePattern()));
}

@Override
public Object visitShowTables(ShowTablesContext ctx) {
return new ShowTables(source(ctx), visitPattern(ctx.pattern()));
TableIdentifier ti = visitTableIdentifier(ctx.tableIdent);
String index = ti != null ? ti.qualifiedIndex() : null;
return new ShowTables(source(ctx), index, visitLikePattern(ctx.likePattern()));
}

@Override
Expand All @@ -136,8 +138,9 @@ public Object visitShowSchemas(ShowSchemasContext ctx) {

@Override
public Object visitShowColumns(ShowColumnsContext ctx) {
TableIdentifier identifier = visitTableIdentifier(ctx.tableIdentifier());
return new ShowColumns(source(ctx), identifier.index());
TableIdentifier ti = visitTableIdentifier(ctx.tableIdent);
String index = ti != null ? ti.qualifiedIndex() : null;
return new ShowColumns(source(ctx), index, visitLikePattern(ctx.likePattern()));
}

@Override
Expand Down Expand Up @@ -172,13 +175,17 @@ else if (value.toUpperCase(Locale.ROOT).equals("TABLE")) {

// if the ODBC enumeration is specified, skip validation
EnumSet<IndexType> set = types.isEmpty() ? null : EnumSet.copyOf(types);
return new SysTables(source(ctx), visitPattern(ctx.clusterPattern), visitPattern(ctx.tablePattern), set, legacyTableType);
TableIdentifier ti = visitTableIdentifier(ctx.tableIdent);
String index = ti != null ? ti.qualifiedIndex() : null;
return new SysTables(source(ctx), visitLikePattern(ctx.clusterLike), index, visitLikePattern(ctx.tableLike), set, legacyTableType);
}

@Override
public Object visitSysColumns(SysColumnsContext ctx) {
Location loc = source(ctx);
return new SysColumns(loc, string(ctx.cluster), visitPattern(ctx.indexPattern), visitPattern(ctx.columnPattern));
TableIdentifier ti = visitTableIdentifier(ctx.tableIdent);
String index = ti != null ? ti.qualifiedIndex() : null;
return new SysColumns(source(ctx), string(ctx.cluster), index, visitLikePattern(ctx.tableLike),
visitLikePattern(ctx.columnPattern));
}

@Override
Expand All @@ -190,4 +197,4 @@ public SysTypes visitSysTypes(SysTypesContext ctx) {
public Object visitSysTableTypes(SysTableTypesContext ctx) {
return new SysTableTypes(source(ctx));
}
}
}
Original file line number Diff line number Diff line change
Expand Up @@ -63,6 +63,7 @@
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.FunctionTemplateContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.GuidEscapedLiteralContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.IntegerLiteralContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.LikePatternContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.LogicalBinaryContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.LogicalNotContext;
import org.elasticsearch.xpack.sql.parser.SqlBaseParser.MatchQueryContext;
Expand Down Expand Up @@ -220,6 +221,11 @@ public Expression visitPredicated(PredicatedContext ctx) {
return pCtx.NOT() != null ? new Not(loc, e) : e;
}

@Override
public LikePattern visitLikePattern(LikePatternContext ctx) {
return ctx == null ? null : visitPattern(ctx.pattern());
}

@Override
public LikePattern visitPattern(PatternContext ctx) {
if (ctx == null) {
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,10 @@ abstract class IdentifierBuilder extends AbstractBuilder {

@Override
public TableIdentifier visitTableIdentifier(TableIdentifierContext ctx) {
if (ctx == null) {
return null;
}

Location source = source(ctx);
ParseTree tree = ctx.name != null ? ctx.name : ctx.TABLE_IDENTIFIER();
String index = tree.getText();
Expand Down
Loading

0 comments on commit ad5a9b5

Please sign in to comment.