Table 8-1 lists the built-in single-row functions that Oracle CQL provides.

Syntax

Purpose

concat returns char1 concatenated with char2 as a char[] or byte1 concatenated with byte2 as a byte[]. The char returned is in the same character set as char1. Its data type depends on the data types of the arguments.

Using concat, you can concatenate any combination of character, byte, and numeric data types. The concat performs automatic numeric to string conversion.

This function is equivalent to the concatenation operator (||).

Examples

concat Function

Consider the query chr_concat in concat and data stream S4 in concat. Stream S4 has schema (c1 char(10)). The query returns the relation in concat.

Concatenation Operator (||)

Consider the query q264 in Example 8–4 and the data stream S10 in Example 8–5. Stream S10 has schema (c1 integer, c2 char(10)). The query returns the relation in Example 8–6.

Syntax

Purpose

hextoraw converts char containing hexadecimal digits in the char character set to a raw value.

Examples

Consider the query q6 and the data stream SByt. Stream SByt has schema (c1 integer, c2 char(10)). The query returns the relation.

<query id="q6"><![CDATA[ 
    select * from SByt[range 2] 
    where 
        hextoraw(c2) between and hextoraw("5600")
]]></query>

Timestamp   Tuple
1000        1,"51c1"
2000        2,"52"
3000        3,"53aa"
4000        4,"5"
5000         ,"55ef"
6000        6,
h 8000
h 200000000

Timestamp   Tuple Kind  Tuple
3000        +           3,"53aa"
5000        -           3,"53aa"
5000        +            ,"55ef"
7000        -            ,"55ef"

Syntax

Purpose

The length function returns the length of its char or byte expression as an int. length calculates length using characters as defined by the input character set.

For a char expression, the length includes all trailing blanks. If the expression is null, this function returns null.

Examples

Consider the query chr_len and the data stream S2. Stream S2 has schema (c1 char(10), c2 integer). The query returns the relation.

<query id="chr_len"><![CDATA[ 
    select length(c1) from S2[range 5]
]]></query>

Timestamp   Tuple
 1000
 2000       hi
 8000       hi1
 9000
15000       xyz
h 200000000

Timestamp   Tuple Kind  Tuple
 1000:      +           0
 2000:      +           2
 6000:      -           0
 7000:      -           2
 8000:      +           3
 9000:      +           0
13000:      -           3
14000:      -           0
15000:      +           3
20000:      -           3

Syntax

Purpose

lk boolean true if char1 matches the regular expression char2, otherwise it returns false.

This function is equivalent to the LIKE condition. For more information, see .

Examples

Consider the query q291 and the data stream SLk1. Stream SLk1 has schema (first1 char(20), last1 char(20)). The query returns the relation.

<query id="q291"><![CDATA[ 
    select * from SLk1 
    where 
        lk(first1,"^Ste(v|ph)en$") = true
]]></query>

Timestamp   Tuple
1           Steven,King
2           Sten,Harley
3           Stephen,Stiles
4           Steven,Markles
h 200000000

Timestamp   Tuple Kind  Tuple
1:          +           Steven,King
3:          +           Stephen,Stiles
4:          +           Steven,Markles

Syntax

Purpose

nvl lets you replace null (returned as a blank) with a string in the results of a query. If expr1 is null, then NVL returns expr2. If expr1 is not null, then NVL returns expr1.

The arguments expr1 and expr2 can have any data type. If their data types are different, then GGSA implicitly converts one to the other. If they cannot be converted implicitly, GGSA returns an error. The implicit conversion is implemented as follows:

• If expr1 is character data, then GGSA converts expr2 to character data before comparing them and returns VARCHAR2 in the character set of expr1.

• If expr1 is numeric, then GGSA determines which argument has the highest numeric precedence, implicitly converts the other argument to that data type, and returns that data type.

Examples

Consider the query q281 and the data stream SNVL. Stream SNVL has schema (c1 char(20), c2 integer). The query returns the relation.

<query id="q281"><![CDATA[ 
    select nvl(c1,"abcd") from SNVL
]]></query>

Timestamp   Tuple
1              ,1
2            ab,2
3           abc,3
4              ,4
h 200000000

Timestamp   Tuple Kind  Tuple
1:          +           abcd
2:          +           ab
3:          +           abc
4:          +           abcd

Syntax

Purpose

prev returns the value of the stream attribute (function argument identifier2) of the event that occurred previous to the current event and which belongs to the partition to which the current event belongs. It evaluates to NULL if there is no such previous event.

The type of the specified stream element may be any of:

• integer

• bigint

• float

• double

• byte

• char

• interval

• timestamp.

The return type of this function depends on the type of the specified stream attribute (function argument identifier2).

Where:

• identifier1.identifier2 : identifier1 is the name of a correlation variable used in the PATTERN clause and defined in the DEFINE clause and identifier2 is the name of a stream attribute whose value in the previous event should be returned by prev.

• const_int: if this argument has a value n, then it specifies the nth previous event in the partition to which the current event belongs. The value of the attribute (specified in argument identifier2) in the nth previous event will be returned if such an event exists, NULL otherwise.

• const_bigint: specifies a time range duration in nanoseconds and should be used if you are interested in previous events that occurred only within a certain range of time before the current event.

If the query uses PARTITION BY with the prev function and input data will include many different partition key values (meaning many partitions), then total memory consumed for storing the previous event(s) per partition could be large. In such cases, consider using the time range duration (the third argument, possibly with a large range value) so that this memory can be reclaimed wherever possible.

Examples

prev(identifier1.identifier2)

Consider query q2 and the data stream S1. Stream S1 has schema (c1 integer). This example defines pattern A as A.c1 = prev(A.c1). In other words, pattern A matches when the value of c1 in the current stream element matches the value of c1 in the stream element immediately before the current stream element. The query returns the stream.

<query id="q2"><![CDATA[ 
    select 
        T.Ac1,
        T.Cc1 
    from 
        S1 
    MATCH_RECOGNIZE ( 
        MEASURES 
            A.c1 as Ac1, 
            C.c1 as Cc1 
        PATTERN(A B+ C) 
        DEFINE 
            A as A.c1 = prev(A.c1), 
            B as B.c1 = 10, 
            C as C.c1 = 7
    ) as T
]]></query>

Timestamp   Tuple
1000        35
3000        35
4000        10
5000         7

Timestamp   Tuple Kind  Tuple
5000:       +           35,7

prev(identifier1.identifier2, const_int)

Consider query q35 and the data stream S15. Stream S15 has schema (c1 integer, c2 integer). This example defines pattern A as A.c1 = prev(A.c1,3). In other words, pattern A matches when the value of c1 in the current stream element matches the value of c1 in the third stream element before the current stream element. The query returns the stream.

<query id="q35"><![CDATA[ 
    select T.Ac1 from S15 
    MATCH_RECOGNIZE ( 
        MEASURES 
            A.c1 as Ac1 
        PATTERN(A) 
        DEFINE 
            A as (A.c1 = prev(A.c1,3) )
    ) as T
]]></query>

Timestamp   Tuple
 1000       45,20
 2000       45,30
 3000       45,30
 4000       45,30
 5000       45,30
 6000       45,20
 7000       45,20
 8000       45,20
 9000       43,40
10000       52,10
11000       52,30
12000       43,40
13000       52,50
14000       43,40
15000       43,40

Timestamp   Tuple Kind  Tuple
 4000:       +          45
 5000:       +          45
 6000:       +          45
 7000:       +          45
 8000:       +          45
12000:       +          43
13000:       +          52
15000:       +          43

prev(identifier1.identifier2, const_int, const_bigint)

Consider query q36 and the data stream S15. Stream S15 has schema (c1 integer, c2 integer). This example defines pattern A as A.c1 = prev(A.c1,3,5000000000L). In other words, pattern A matches when:

• the value of c1 in the current event equals the value of c1 in the third previous event of the partition to which the current event belongs, and

• the difference between the timestamp of the current event and that third previous event is less than or equal to 5000000000L nanoseconds.

The query returns the output stream. Notice that in the output stream, there is no output at 8000. The following example shows the contents of the partition (partitioned by the value of the c2 attribute) to which the event at 8000 belongs.

Timestamp   Tuple
1000        45,20
6000        45,20
7000        45,20
8000        45,20

As the following example shows, even though the value of c1 in the third previous event (the event at 1000) is the same as the value c1 in the current event (the event at 8000), the range condition is not satisfied. This is because the difference in the timestamps of these two events is more than 5000000000 nanoseconds. So it is treated as if there is no previous tuple and prev returns NULL so the condition fails to match.

<query id="q36"><![CDATA[ 
    select T.Ac1 from S15 
    MATCH_RECOGNIZE ( 
        PARTITION BY
            c2 
        MEASURES 
            A.c1 as Ac1 
        PATTERN(A) 
        DEFINE 
            A as (A.c1 = prev(A.c1,3,5000000000L) )
    ) as T
]]></query>

Timestamp   Tuple
 1000       45,20
 2000       45,30
 3000       45,30
 4000       45,30
 5000       45,30
 6000       45,20
 7000       45,20
 8000       45,20
 9000       43,40
10000       52,10
11000       52,30
12000       43,40
13000       52,50
14000       43,40
15000       43,40

Timestamp   Tuple Kind  Tuple
5000:       +           45

Syntax

Purpose

rawtohex converts byte containing a raw value to hexadecimal digits in the CHAR character set.

Examples

Consider the query byte_to_hex and the data stream S5. Stream S5 has schema (c1 integer, c2 byte(10)). This query uses the rawtohex function to convert a ten byte raw value to the equivalent ten hexadecimal digits in the character set of your current locale. The query returns the relation.

<query id="byte_to_hex"><![CDATA[ 
    select rawtohex(c2) from S5[range 4]
]]></query>

Timestamp   Tuple
1000        1,"51c1"
2000        2,"52"
2500        7,"axc"
3000        3,"53aa"
4000        4,"5"
5000         ,"55ef"
6000        6,
h 8000
h 200000000

Timestamp   Tuple Kind  Tuple
 1000:      +           51c1
 2000:      +           52
 3000:      +           53aa
 4000:      +           05
 5000:      -           51c1
 5000:      +           55ef
 6000:      -           52
 6000:      +
 7000:      -           53aa
 8000:      -           05
 9000:      -           55ef
10000:      -

Syntax

Purpose

systimestamp returns the system date, including fractional seconds and time zone, of the system on which the GGSA server resides. The return type is TIMESTAMP WITH TIME ZONE.

Examples

Consider the query q106 and the data stream S0. Stream S0 has schema (c1 float, c2 integer). The query returns the relation.

<query id="q106"><![CDATA[ 
    select * from S0 
    where 
        case c2 
            when 10 then null 
            when 20 then null 
            else systimestamp() 
        end > "07/06/2007 14:13:33"
]]></query>

Timestamp   Tuple
     1000   0.1 ,10
     1002   0.14,15
   200000   0.2 ,20
   400000   0.3 ,30
   500000   0.3 ,35
   600000       ,35
h 800000
100000000   4.04,40
h 200000000

Timestamp   Tuple Kind  Tuple
     1002:  +           0.14,15
   400000:  +           0.3 ,30
   500000:  +           0.3 ,35
   600000:  +               ,35
100000000:  +           4.04,40

Syntax

Purpose

Input/Output Types

The input/output types for this function are as follows:

Examples

Consider the query q282 and the data stream S11. Stream S11 has schema (c1 integer, name char(10)). The query returns the relation.

<query id="q282"><![CDATA[ 
    select nvl(to_bigint(c1), 5.2) from S11
]]></query>

Timestamp   Tuple
  10        1,abc
2000         ,ab
3400        3,abc
4700         ,a
h 8000
h 200000000

Timestamp   Tuple Kind  Tuple
  10:       +           1
2000:       +           5.2
3400:       +           3
4700:       +           5.2

Syntax

Purpose

to_boolean returns a value of true or false for its bigint or integer expression argument.

Examples

Consider the query q282 and the data stream S11. Stream S11 has schema (c1 integer, name char(10)). The query returns the relation.

<view id="v2" schema="c1 c2" ><![CDATA[
    select to_boolean(c1), c1 from tkboolean_S3 [now] where c2 = 0.1
]]></view><query id="q1"><![CDATA[ 
    select * from v2
]]></query>

Timestamp   Tuple
1000        -2147483648, 0.1
2000        2147483647, 0.2
3000        12345678901, 0.3
4000        -12345678901, 0.1
5000        9223372036854775799, 0.2
6000        -9223372036854775799, 0.3
7000        , 0.1
8000        10000000000, 0.2
9000        60000000000, 0.3
h 200000000

Timestamp   Tuple Kind  Tuple
1000      +           true,-2147483648
1000      -           true,-2147483648
4000      +           true,-12345678901
4000      -           true,-12345678901
7000      +           ,
7000      -           ,

Syntax

Purpose

to_char returns a char value for its integer, double, bigint, float, timestamp, or interval expression argument. If the bigint argument exceeds the char precision, GGSA returns an error.

Examples

Consider the query q282 and the data stream S11. Stream S11 has schema (c1 integer, name char(10)). The query returns the relation.

<query id="q1"><![CDATA[ 
    select to_char(c1), to_char(c2), to_char(c3), to_char(c4), to_char(c5), to_char(c6) 
    from S1
]]></query>

Timestamp   Tuple
1000        99,99999, 99.9, 99.9999, "4 1:13:48.10", "08/07/2004 11:13:48", cep

Timestamp   Tuple Kind  Tuple
1000:       +           99,99999,99.9,99.9999,4 1:13:48.10,08/07/2004 11:13:48

Syntax

Purpose

to_double returns a double value for its bigint, integer, or float expression argument. If the bigint argument exceeds the double precision, GGSA returns an error.

Examples

Consider the query q282 and the data stream S11. Stream S11 has schema (c1 integer, name char(10)). The query returns the relation.

<query id="q282"><![CDATA[ 
    select nvl(to_double(c1), 5.2) from S11
]]></query>

Timestamp   Tuple
  10        1,abc
2000         ,ab
3400        3,abc
4700         ,a
h 8000
h 200000000

Timestamp   Tuple Kind  Tuple
  10:       +           1
2000:       +           5.2
3400:       +           3
4700:       +           5.2

Syntax

Purpose

to_float returns a float number equivalent of its bigint or integer argument. If the bigint argument exceeds the float precision, GGSA returns an error.

Examples

Consider the query q1 and the data stream S11. Stream S1 has schema (c1 integer, name char(10)). The query returns the relation.

<query id="q1"><![CDATA[ 
    select nvl(to_float(c1), 5.2) from S11
]]></query>

Timestamp   Tuple
  10        1, abc
2000         , ab
3400        3, abc
4700         , a
h 8000
h 200000000

Timestamp   Tuple Kind  Tuple
10:+ 1.02000:+ 5.23400:+ 3.04700:+ 5.2

Syntax

Purpose

to_timestamp converts char literals that conform to java.text.SimpleDateFormat format models to timestamp data types. There are two forms of the to_timestamp function distinguished by the number of arguments:

• char: this form of the to_timestamp function converts a single char argument that contains a char literal that conforms to the default java.text.SimpleDateFormat format model (MM/dd/yyyy HH:mm:ss) into the corresponding timestamp data type.

• char1, char2: this form of the to_timestamp function converts the char1 argument that contains a char literal that conforms to the java.text.SimpleDateFormat format model specified in the second char2 argument into the corresponding timestamp data type.

• long: this form of the to_timestamp function converts a single long argument that represents the number of nanoseconds since the standard base time known as "the epoch", namely January 1, 1970, 00:00:00 GMT, into the corresponding timestamp data type represented as a number in milliseconds since "the epoch" with a date format that conforms to the default java.text.SimpleDateFormat format model (MM/dd/yyyy HH:mm:ss).

Examples

Consider the query q277 and the data stream STs2. Stream STs2 has schema (c1 integer, c2 char(20)). The query returns the relation.

<query id="q277"><![CDATA[ 
    select * from STs2 
    where 
        to_timestamp(c2,"yyMMddHHmmss") = to_timestamp("09/07/2005 10:13:48")
]]></query>

Timestamp   Tuple
1           1,"040807111348"
2           2,"050907101348"
3           3,"041007111348"
4           4,"060806111248"
h 200000000

Timestamp   Tuple Kind  Tuple
2:          +           2,050907101348