Hi there,
I am trying to query for a right-to-left override (RTLO) character (U+202E) within a process name.
My query in Dev Tools in Kibana is this:
GET /[index_name]/_eql/search
{
"query": """
process where process.name like "*Up\u{200f}date.exe"
"""
}
I use the unicode syntax based on what I had found from a section in the documentation.
This is the error that I received:
{
"error" : {
"root_cause" : [
{
"type" : "parsing_exception",
"reason" : "line 2:37: token recognition error at: '\"*Up\\u'"
}
],
"type" : "parsing_exception",
"reason" : "line 2:37: token recognition error at: '\"*Up\\u'",
"caused_by" : {
"type" : "lexer_no_viable_alt_exception",
"reason" : null
}
},
"status" : 400
}
Leaving a white-space where my query is process where process.name like "*Up \u{200f} date.exe"
, leads to the same error.
I have tried changing the query to process where process.name like "*Up\\u{200f}date.exe"
, which adds an extra backslash. While the query executes successfully, no results are returned.
This is how the process name looks like in Kibana:
As can be seen from my cursor highlight, the RTLO character is inserted after Up
.
Thank you for your time!
Note-to-self: I should change like
to ==
.
Note-to-self-2: Found an old post about replacing unicode characters, but it is for Logstash.