- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
Based on the open-source Flink functions, Stream Compute Service has supplemented and optimized some functions. The following table lists some built-in functions.
Built-in Function | Stream Compute Service Definition | Flink Definition | Remarks |
unbase64 | unbase64(string) | - | Stream Compute Service: Decodes a Base64-encoded string. The input is a string and the output is an Array[Byte]. |
decode | DECODE(binary,charset), DECODE(binary) | DECODE(binary, string) | Stream Compute Service: DECODE(binary) uses charset="UTF-8” by default. Flink: Returns NULL if any parameter is NULL. |
split | SPLIT(string, separator) | - | Stream Compute Service: Splits string by separator, and returns a Row object. |
get_row_field_str | GET_ROW_FIELD_STR(row, index) | - | Stream Compute Service: Gets the value of the indexth column in Row object row. index starts from 0. The return value is of VARCHAR type. |
get_row_arity | GET_ROW_ARITY(row) | - | Stream Compute Service: Gets the number of columns in Row object row. |
can_cast_to | CAN_CAST_TO(str, type) | - | Stream Compute Service: Checks whether the value of str can be converted to the type specified by type, and returns a Boolean value. The return value can be used as a condition in a CASE statement. For example, CAN_CAST_TO('123456', 'INTEGER') returns True, and CAN_CAST_TO('a145', 'DOUBLE') returns False. |
date_format_simple | DATE_FORMAT_SIMPLE(timestamp, simple_format) | - | Stream Compute Service: Converts a field of BIGINT(Long) type (which represents a Unix timestamp in milliseconds) to a string using the format that is compatible with Java's SimpleDateFormat. For example, DATE_FORMAT_SIMPLE(unix_ts, 'yyyy-MM-dd HH:mm:ss') returns a string similar to "2020-01-01 12:13:14". The input type is Long and the output type is String. |
to_timestamp | TO_TIMESTAMP(string, simple_format) | TO_TIMESTAMP(string1[, string2]) | Stream Compute Service: Converts string to a timestamp using the simple_format, which is compatible with Java's SimpleDateFormat. UTC+8 is used by default. Example: `TO_TIMESTAMP(ts, 'yyyy-MM-dd HH:mm:ss')' The input type is String and the output type is Timestamp.Flink: Same. |
timestamp_to_long | TIMESTAMP_TO_LONG(timestamp), TIMESTAMP_TO_LONG(timestamp, mode) | - | Stream Compute Service: Converts a TIMESTAMP type parameter to a value of BIGINT (Long) type. If mode is 'SECOND', converts to a Unix timestamp in seconds, for example, 1548403425. If mode is any other value or not specified, converts to a Unix timestamp in milliseconds, for example, 1548403425512. |
get_tumble_start | get_tumble_start(timestamp,windowSize), get_tumble_start(timestamp,offset,windowSize) | - | Stream Compute Service: Gets the start time of a window. The default offset is 0. |
if_null_str | IF_NULL_STR(str, defaultValue) | - | Stream Compute Service: Returns the value of str itself if it is not NULL; otherwise, returns the value of defaultValue. |
get_json_object | GET_JSON_OBJECT(json_str, path_str) | - | |
split_index | SPLIT_INDEX(string, separator, index) | SPLIT_INDEX(string1, string2, integer1) | Splits string by separator, and returns the indexth string (VARCHAR) of the split strings. index starts from 0.Stream Compute Service: If string or separator is null, returns string.Flink: Returns null if any parameter is left empty. |
first_str | FIRST_VALUE(expression) | FIRST_VALUE(expression) | Stream Compute Service: Returns the first value in a set of values. Flink: Same. |
last_str | LAST_VALUE(expression) | LAST_VALUE(expression) | Stream Compute Service: Returns the last value in a set of values. Flink: Same. |
explode | EXPLODE(inputStr, separator) | - | Stream Compute Service: Splits a string into a temporary table with multiple rows. This function is a table function, and you need to use the keyword LATERAL TABLE ( ) to reference this dynamically generated temporary table as the right table for JOIN. |
Yes
No
Was this page helpful?