Kusto: extract unique words in text

Question

is it possible extract unique word from column with Kusto?

Example text: an example text, an orange, text bold Get only words: an, example, text, orange, bold

I'm trying with this regex:

mytable | extend ff = extract_all(@'(\w+\b)(?!.*\1\b)', info));

Answer 1

you could try this, using set_union() on top of the output of extract_all():

print input = "an example text, an orange, text bold Get only words: an, example, text, orange, bold"
| extend unique_words = set_union(dynamic(null), extract_all(@"(\w+)", input))

input	unique_words
an example text, an orange, text bold Get only words: an, example, text, orange, bold	[ "an", "example", "text", "orange", "bold", "Get", "only", "words" ]