Support parse_url

docs/additional-functionality/advanced_configs.md

@@ -299,6 +299,7 @@ Name | SQL Function(s) | Description | Default Value | Notes
 <a name="sql.expression.NthValue"></a>spark.rapids.sql.expression.NthValue|`nth_value`|nth window operator|true|None|
 <a name="sql.expression.OctetLength"></a>spark.rapids.sql.expression.OctetLength|`octet_length`|The byte length of string data|true|None|
 <a name="sql.expression.Or"></a>spark.rapids.sql.expression.Or|`or`|Logical OR|true|None|
+<a name="sql.expression.ParseUrl"></a>spark.rapids.sql.expression.ParseUrl|`parse_url`|Extracts a part from a URL|true|None|
 <a name="sql.expression.PercentRank"></a>spark.rapids.sql.expression.PercentRank|`percent_rank`|Window function that returns the percent rank value within the aggregation window|true|None|
 <a name="sql.expression.Pmod"></a>spark.rapids.sql.expression.Pmod|`pmod`|Pmod|true|None|
 <a name="sql.expression.PosExplode"></a>spark.rapids.sql.expression.PosExplode|`posexplode_outer`, `posexplode`|Given an input array produces a sequence of rows for each value in the array|true|None|

docs/compatibility.md

@@ -456,6 +456,20 @@ Spark stores timestamps internally relative to the JVM time zone.  Converting an
 between time zones is not currently supported on the GPU. Therefore operations involving timestamps
 will only be GPU-accelerated if the time zone used by the JVM is UTC.
 
+## URL parsing
+
+In Spark, parse_url is based on java's URI library, while the implementation in the RAPIDS Accelerator is based on regex extraction. Therefore, the results may be different in some edge cases.
+
+These are the known cases where running on the GPU will produce different results to the CPU:
+
+- Spark allow an empty authority component only when it's followed by a non-empty path, 
+  query component, or fragment component. But in plugin, parse_url just simply allow empty 
+  authority component without checking if it is followed something or not. So `parse_url('http://', 'HOST')` will
+  return `null` in Spark, but return `""` in plugin.
+- If input url has a invalid Ipv6 address, Spark will return `null` for all components, but plugin will parse other
+  components except `HOST` as normal. So `http://userinfo@[1:2:3:4:5:6:7:8:9:10]/path?query=1#Ref`'s result will be 
+  `[null,/path,query=1,Ref,http,/path?query=1,userinfo@[1:2:3:4:5:6:7:8:9:10],userinfo]`
+
 ## Windowing
 
 ### Window Functions