path: root/doc/src/sgml/json.sgml

<!-- doc/src/sgml/json.sgml -->

<sect1 id="datatype-json">
 <title><acronym>JSON</> Types</title>

 <indexterm zone="datatype-json">
  <primary>JSON</primary>
 </indexterm>

 <indexterm zone="datatype-json">
  <primary>JSONB</primary>
 </indexterm>

 <para>
  JSON data types are for storing JSON (JavaScript Object Notation)
  data, as specified in <ulink url="http://rfc7159.net/rfc7159">RFC
  7159</ulink>. Such data can also be stored as <type>text</type>, but
  both JSON data types have the advantage of enforcing that each
  stored value is a valid JSON value.  There are also related support
  functions available; see <xref linkend="functions-json">.
 </para>

 <para>
  There are two JSON data types: <type>json</> and <type>jsonb</>.
  Both accept <emphasis>almost</emphasis> identical sets of values as
  input.  The major practical difference is one of efficiency.  The
  <type>json</> data type stores an exact copy of the input text,
  which processing functions must continually reparse, while
  <type>jsonb</> data is stored in a decomposed binary format that
  makes it slightly less efficient to input due to added serialization
  overhead, but significantly faster to process, since it never needs
  reparsing.  <type>jsonb</> also supports advanced
  <acronym>GIN</acronym> indexing, which is a further significant
  advantage.
 </para>

 <para>
  The other difference between the types is that the <type>json</>
  type is guaranteed to contain an exact copy of the input, including
  preservation of semantically insignificant white space, and the
  order of keys within JSON objects (although <type>jsonb</> will
  preserve trailing zeros within a JSON number). Also, because the
  exact text is kept, if a JSON object within the value contains the
  same key more than once, and has been stored using the <type>json</>
  type, all the key/value pairs are kept.  In that case, the
  processing functions consider the last value as the operative one.
  By contrast, <type>jsonb</> does not preserve white space, does not
  preserve the order of object keys, and does not keep duplicate
  object keys.  Only the last value for a key specified in the input
  is kept.
 </para>

 <para>
  In general, most applications will prefer to store JSON data as
  <type>jsonb</>, unless there are quite specialized needs.
 </para>

 <para>
  <productname>PostgreSQL</productname> allows only one server
  encoding per database.  It is therefore not possible for the JSON
  types to conform rigidly to the specification unless the server
  encoding is UTF-8. Attempts to directly include characters which
  cannot be represented in the server encoding will fail; conversely,
  characters which can be represented in the server encoding but not
  in UTF-8 will be allowed.  <literal>\uXXXX</literal> escapes are
  allowed regardless of the server encoding, and are checked only for
  syntactic correctness.
 </para>

 <sect2 id="json-types">
  <title>Mapping of RFC-7159/JSON Primitive Types to <productname>PostgreSQL</productname> Types</title>
  <table id="json-type-mapping-table">
     <title>Mapping of type correspondence, notes</title>
     <tgroup cols="3">
      <thead>
       <row>
        <entry><productname>PostgreSQL</productname> type</entry>
        <entry>RFC-7159/JSON primitive type</entry>
        <entry>Notes</entry>
       </row>
      </thead>
      <tbody>
       <row>
        <entry><type>text</></entry>
        <entry><type>string</></entry>
        <entry>See general introductory notes on encoding and JSON</entry>
       </row>
       <row>
        <entry><type>numeric</></entry>
        <entry><type>number</></entry>
        <entry><literal>NaN</literal> and <literal>infinity</literal> values are disallowed</entry>
       </row>
       <row>
        <entry><type>boolean</></entry>
        <entry><type>boolean</></entry>
        <entry>Only lowercase <literal>true</literal> and <literal>false</literal> values are accepted</entry>
       </row>
       <row>
        <entry><type>unknown</></entry>
        <entry><type>null</></entry>
        <entry>SQL <literal>NULL</literal> is orthogonal.  NULL semantics do not apply.</entry>
       </row>
      </tbody>
     </tgroup>
   </table>
  <para>
    Primitive types described by <acronym>RFC</> 7159 are effectively
    internally mapped onto native
    <productname>PostgreSQL</productname> types.  Therefore, there are
    some very minor additional constraints on what constitutes valid
    <type>jsonb</type> that do not apply to the <type>json</type>
    type, or to JSON in the abstract, that pertain to limits on what
    can be represented by the underlying type system.  These
    implementation-defined restrictions are permitted by
    <acronym>RFC</> 7159.  However, in practice problems are far more
    likely to occur in other implementations which internally
    represent the <type>number</> JSON primitive type as IEEE 754
    double precision floating point values, which <acronym>RFC</> 7159
    explicitly anticipates and allows for.  When using JSON as an
    interchange format with such systems, the danger of losing numeric
    precision in respect of data originally stored by
    <productname>PostgreSQL</productname> should be considered.
  </para>
  <para>
    Conversely, as noted above there are some minor restrictions on
    the input format of JSON primitive types that do not apply to
    corresponding <productname>PostgreSQL</productname> types.
  </para>

 </sect2>

 <sect2 id="json-querying">
  <title>Querying <type>jsonb</type> documents effectively</title>
  <para>
   Representing data as JSON can be considerably more flexible than
   the traditional relational data model, which is compelling in
   environments where requirements are fluid.  It is quite possible
   for both approaches to co-exist and complement each other within
   the same application.  However, even for applications where maximal
   flexibility is desired, it is still recommended that JSON documents
   have a somewhat fixed structure.  This structure is typically
   unenforced (though enforcing some business rules declaratively is
   possible), but makes it easier to write queries that usefully
   summarize a set of <quote>documents</> (datums) in a table.
  </para>
  <para>
   <type>jsonb</> data is subject to the same concurrency control
   considerations as any other datatype when stored in a table.
   Although storing large documents is practicable, in order to ensure
   correct behavior row-level locks are, quite naturally, acquired as
   rows are updated.  Consider keeping <type>jsonb</> documents at a
   manageable size in order to decrease lock contention among updating
   transactions.  Ideally, <type>jsonb</> documents should each
   represent an atomic datum that business rules dictate cannot
   reasonably be further subdivided into smaller atomic datums that
   can be independently modified.
  </para>
 </sect2>
 <sect2 id="json-keys-elements">
  <title><type>jsonb</> Input and Output Syntax</title>
  <para>
   In effect, <type>jsonb</> has an internal type system whose
   implementation is defined in terms of several particular ordinary
   <productname>PostgreSQL</productname> types.  The SQL parser does
   not have direct knowledge of the internal types that constitute a
   <type>jsonb</>.
  </para>
  <para>
   The following are all valid <type>jsonb</> expressions:
  <programlisting>
-- Simple scalar/primitive value (explicitly required by RFC-7159)
SELECT '5'::jsonb;

-- Array of heterogeneous, primitive-typed elements
SELECT '[1, 2, "foo", null]'::jsonb;

-- Object of heterogeneous key/value pairs of primitive types
-- Note that key values are always strings
SELECT '{"bar": "baz", "balance": 7.77, "active":false}'::jsonb;
  </programlisting>
  </para>
  <para>
   Note the distinction between scalar/primitive values as elements,
   keys and values.
  </para>
 </sect2>
 <sect2 id="json-containment">
  <title><type>jsonb</> containment</title>
  <indexterm>
    <primary>jsonb</primary>
    <secondary>containment</secondary>
  </indexterm>
  <para>
    Testing <quote>containment</> is an important capability of
    <type>jsonb</>.  There is no parallel set of facilities for the
    <type>json</> type.  Containment is the ability to determine if
    one <type>jsonb</> document has contained within it another one.
    <type>jsonb</> is nested, and so containment semantics are nested;
    technically, top-down, unordered <emphasis>subtree isomorphism</>
    may be tested.  Containment is conventionally tested using the
    <literal>@&gt;</> operator, which is made indexable by various
    operator classes discussed later in this section.
  </para>
  <programlisting>
-- Simple scalar/primitive values may contain only each other:
SELECT '"foo"'::jsonb @> '"foo"'::jsonb;

-- The array on the right hand side is contained within the one on the
-- left hand side:
SELECT '[1, 2, 3]'::jsonb @> '[1, 3]'::jsonb;

-- The object with a single pair on the right hand side is contained
-- within the object on the left hand side:
SELECT '{"product": "PostgreSQL", "version": 9.4, "jsonb":true}'::jsonb @> '{"version":9.4}'::jsonb;

-- The array on the right hand side is not contained within the array
-- containing a nested array on the left hand side:
SELECT '[1, 2, [1, 3]]'::jsonb @> '[1, 3]'::jsonb;

-- But with a layer of nesting, it is:
SELECT '[1, 2, [1, 3]]'::jsonb @> '[[1, 3]]'::jsonb;
  </programlisting>
  <para>
    It is both a sufficient and a necessary condition for nesting
    levels to <quote>line up</> for one <type>jsonb</> to contain
    within it another.  Under this definition, objects and arrays
    cannot <quote>line up</>, not least because objects contain
    key/value pairs, while arrays contain elements.
  </para>
  <para>
    As a special exception to the general principle that nesting
    levels should <quote>line up</>, an array may contain a raw scalar:
  </para>
  <programlisting>
-- This array contains the raw scalar value:
SELECT '["foo", "bar"]'::jsonb @> '"bar"'::jsonb;
-- The special exception is not reciprocated -- non-containment is indicated here:
SELECT '"bar"'::jsonb @> '["bar"]'::jsonb;
  </programlisting>
  <para>
    Objects are better suited for testing containment when there is a
    great deal of nesting involved, because unlike arrays they are
    internally optimized for searching, and do not need to be searched
    linearly within a single <type>jsonb</> document.
  </para>
  <programlisting>
-- The right-hand side object is contained in this example:
SELECT '{"p":1, "a":{"b":3, "q":11}, "i":77}'::jsonb @> '{"a":{"b":3}}'::jsonb;
  </programlisting>
  <para>
    The various containment operators, along with all other JSON
    operators and support functions are documented fully within <xref
    linkend="functions-json">, <xref
    linkend="functions-jsonb-op-table">.
  </para>
 </sect2>
 <sect2 id="json-indexing">
  <title><type>jsonb</> GIN Indexing</title>
  <indexterm>
    <primary>jsonb</primary>
    <secondary>indexes on</secondary>
  </indexterm>
  <para>
    <type>jsonb</> GIN indexes can be used to efficiently search among
    more than one possible key/value pair within a single
    <type>jsonb</> datum/document, among a large number of such
    documents within a column in a table (i.e. among many rows).
  </para>
  <para>
    <type>jsonb</> has GIN index support for the <literal>@&gt;</>,
    <literal>?</>, <literal>?&amp;</> and <literal>?|</> operators.
    The default GIN operator class makes all these operators
    indexable:
  </para>
  <programlisting>
-- GIN index (default opclass)
CREATE INDEX idxgin ON api USING GIN (jdoc);

-- GIN jsonb_hash_ops index
CREATE INDEX idxginh ON api USING GIN (jdoc jsonb_hash_ops);
  </programlisting>
  <para>
    The non-default GIN operator class <literal>jsonb_hash_ops</>
    supports indexing the <literal>@&gt;</> operator only.
  </para>
  <para>
    Consider the example of a table that stores JSON documents
    retrieved from a third-party web service, with a documented schema
    definition.  An example of a document retrieved from this web
    service is as follows:
    <programlisting>
{
    "guid": "9c36adc1-7fb5-4d5b-83b4-90356a46061a",
    "name": "Angela Barton",
    "is_active": true,
    "company": "Magnafone",
    "address": "178 Howard Place, Gulf, Washington, 702",
    "registered": "2009-11-07T08:53:22 +08:00",
    "latitude": 19.793713,
    "longitude": 86.513373,
    "tags": [
        "enim",
        "aliquip",
        "qui"
    ]
}
    </programlisting>
    If a GIN index is created on the table that stores these
    documents, <literal>api</literal>, on its <literal>jdoc</>
    <type>jsonb</> column, we can expect that queries like the
    following may make use of the index:
    <programlisting>
-- Note that both key and value have been specified
SELECT jdoc->'guid', jdoc->'name' FROM api WHERE jdoc @&gt; '{"company": "Magnafone"}';
    </programlisting>
    However, the index could not be used for queries like the
    following, due to the aforementioned nesting restriction:
    <programlisting>
SELECT jdoc->'guid', jdoc->'name' FROM api WHERE jdoc -> 'tags' ? 'qui';
    </programlisting>
    Still, with judicious use of expressional indexing, the above
    query can use an index scan.  If there is a requirement to find
    those records with a particular tag quickly, and the tags have a
    high cardinality across all documents, defining an index as
    follows is an effective approach to indexing:
  <programlisting>
-- Note that the "jsonb -> text" operator can only be called on an
-- object, so as a consequence of creating this index the root "jdoc"
-- datum must be an object.  This is enforced during insertion.
CREATE INDEX idxgin ON api USING GIN ((jdoc -> 'tags'));
  </programlisting>
  </para>
  <para>
    Expressional indexes are discussed in <xref
    linkend="indexes-expressional">.
  </para>
  <para>
    For the most flexible approach in terms of what may be indexed,
    sophisticated querying on nested structures is possible by
    exploiting containment.  At the cost of having to create an index
    on the entire structure for each row, and not just a nested
    subset, we may exploit containment semantics to get an equivalent
    result with a non-expressional index on the entire <quote>jdoc</>
    column, <emphasis>without</> ever having to create additional
    expressional indexes against the document (provided only
    containment will be tested).  While the index will be considerably
    larger than our expression index, it will also be much more
    flexible, allowing arbitrary structured searching.  Such an index
    can generally be expected to help with a query like the following:
  </para>
  <programlisting>
SELECT jdoc->'guid', jdoc->'name' FROM api WHERE jdoc @&gt; '{"tags": ["qui"]}';
  </programlisting>
  <para>
   For full details of the semantics that these indexable operators
   implement, see <xref linkend="functions-json">, <xref
   linkend="functions-jsonb-op-table">.
  </para>
 </sect2>
 <sect2 id="json-opclass">
  <title><type>jsonb</> non-default GIN operator class</title>
  <indexterm>
    <primary>jsonb</primary>
    <secondary>indexes on</secondary>
  </indexterm>
  <para>
    Although only the <literal>@&gt;</> operator is made indexable, a
    <literal>jsonb_hash_ops</literal> operator class GIN index has
    some notable advantages over an equivalent GIN index of the
    default GIN operator class for <type>jsonb</type>.  Search
    operations typically perform considerably better, and the on-disk
    size of a <literal>jsonb_hash_ops</literal> operator class GIN
    index can be much smaller.
  </para>
 </sect2>
 <sect2 id="json-btree-indexing">
  <title><type>jsonb</> B-Tree and hash indexing</title>
  <para>
    <type>jsonb</type> comparisons and related operations are
    <emphasis>type-wise</>, in that the underlying
    <productname>PostgreSQL</productname> datatype comparators are
    invoked recursively, much like a traditional composite type.
  </para>
  <para>
    <type>jsonb</> also supports <type>btree</> and <type>hash</>
    indexes.  Ordering between <type>jsonb</> datums is:
    <synopsis>
      <replaceable>Object</replaceable> > <replaceable>Array</replaceable> > <replaceable>Boolean</replaceable> > <replaceable>Number</replaceable> > <replaceable>String</replaceable> > <replaceable>Null</replaceable>

      <replaceable>Object with n pairs</replaceable> > <replaceable>object with n - 1 pairs</replaceable>

      <replaceable>Array with n elements</replaceable> > <replaceable>array with n - 1 elements</replaceable>
    </synopsis>
      Subsequently, individual primitive type comparators are invoked.
      All comparisons of JSON primitive types occurs using the same
      comparison rules as the underlying
      <productname>PostgreSQL</productname> types.  Strings are
      compared lexically, using the default database collation.
      Objects with equal numbers of pairs are compared:
    <synopsis>
      <replaceable>key-1</replaceable>, <replaceable>value-1</replaceable>, <replaceable>key-2</replaceable> ...
    </synopsis>
      Note however that object keys are compared in their storage order, and in particular,
      since shorter keys are stored before longer keys, this can lead to results that might be
      unintuitive, such as:
      <programlisting>{ "aa": 1, "c": 1} > {"b": 1, "d": 1}</programlisting>
      Similarly, arrays with equal numbers of elements are compared:
    <synopsis>
      <replaceable>element-1</replaceable>, <replaceable>element-2</replaceable> ...
    </synopsis>
  </para>
 </sect2>
</sect1>