pyspark.sql.functions.str_to_map#

pyspark.sql.functions.str_to_map(text, pairDelim=None, keyValueDelim=None)[source]#

Map function: Converts a string into a map after splitting the text into key/value pairs using delimiters. Both pairDelim and keyValueDelim are treated as regular expressions.

Added in version 3.5.0.

Parameters:

textColumn or str: Input column or strings. A column that evaluates to a string.
pairDelimColumn or str, optional: Delimiter to use to split pairs. Default is comma (,). A column that evaluates to a string.
keyValueDelimColumn or str, optional: Delimiter to use to split key/value. Default is colon (:). A column that evaluates to a string.

Returns:

Column: A new column of map type where each string in the original column is converted into a map. Returns a column that evaluates to a map.

Examples

Example 1: Using default delimiters

>>> from pyspark.sql import functions as sf
>>> df = spark.createDataFrame([("a:1,b:2,c:3",)], ["e"])
>>> df.select(sf.str_to_map(df.e)).show(truncate=False)
+------------------------+
|str_to_map(e, ,, :)     |
+------------------------+
|{a -> 1, b -> 2, c -> 3}|
+------------------------+

Example 2: Using custom delimiters

>>> from pyspark.sql import functions as sf
>>> df = spark.createDataFrame([("a=1;b=2;c=3",)], ["e"])
>>> df.select(sf.str_to_map(df.e, sf.lit(";"), sf.lit("="))).show(truncate=False)
+------------------------+
|str_to_map(e, ;, =)     |
+------------------------+
|{a -> 1, b -> 2, c -> 3}|
+------------------------+

Example 3: Using different delimiters for different rows

>>> from pyspark.sql import functions as sf
>>> df = spark.createDataFrame([("a:1,b:2,c:3",), ("d=4;e=5;f=6",)], ["e"])
>>> df.select(sf.str_to_map(df.e,
...   sf.when(df.e.contains(";"), sf.lit(";")).otherwise(sf.lit(",")),
...   sf.when(df.e.contains("="), sf.lit("=")).otherwise(sf.lit(":"))).alias("str_to_map")
... ).show(truncate=False)
+------------------------+
|str_to_map              |
+------------------------+
|{a -> 1, b -> 2, c -> 3}|
|{d -> 4, e -> 5, f -> 6}|
+------------------------+

Example 4: Using a column of delimiters

>>> from pyspark.sql import functions as sf
>>> df = spark.createDataFrame([("a:1,b:2,c:3", ","), ("d=4;e=5;f=6", ";")], ["e", "delim"])
>>> df.select(sf.str_to_map(df.e, df.delim, sf.lit(":"))).show(truncate=False)
+---------------------------------------+
|str_to_map(e, delim, :)                |
+---------------------------------------+
|{a -> 1, b -> 2, c -> 3}               |
|{d=4 -> NULL, e=5 -> NULL, f=6 -> NULL}|
+---------------------------------------+

Example 5: Using a column of key/value delimiters

>>> from pyspark.sql import functions as sf
>>> df = spark.createDataFrame([("a:1,b:2,c:3", ":"), ("d=4;e=5;f=6", "=")], ["e", "delim"])
>>> df.select(sf.str_to_map(df.e, sf.lit(","), df.delim)).show(truncate=False)
+------------------------+
|str_to_map(e, ,, delim) |
+------------------------+
|{a -> 1, b -> 2, c -> 3}|
|{d -> 4;e=5;f=6}        |
+------------------------+