[RULE:values_format_logs]
alter parsed_fields = regexcapture(_raw_log,"<(?P<pri>\d{1,3})>(?P<version>\d{1,3})\s(?P<iso_timestamp>\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}[+-]\d{2}:\d{2})\s(?P<hostname>\S+)[\s-]+\\?\"?(?P<remote_addr>\S+)\\?\"?\s(?P<timestamp>\S+)\s(?P<status>\S+)\s(?P<bytes_sent>\S+)\s\"(?P<request>.+)\s\"?(?P<blocked>\d)\s\"?(?P<is_human>\d)\"?\s\"?(?P<block_reason>[^\"]*)\"\s\"(?P<country_name>[^\"]*)\"\s\"(?P<country_code>[^\"]*)\"\s\"(?P<request_id>[^\"]*)\"\s\"(?P<captured_vector>[^\"]*)\"\s(?P<request_time>\S+)\s\"(?P<upstream_addr>[^\"]*)\"?\s(?P<upstream_response_time>[^\"]*)\"?\s\"(?P<upstream_status>\d?\d?\d?)(?P<domain_name>[^\"]*)\"\s\"(?P<host>[^\"]*)\"\s\"(?P<referer>[^\"]*)\"\s\"?(?P<user_agent>[^\"]*)\"\s\"(?P<asn>[^\"]*)\"\s\"(?P<ssl_protocol>[^\"]*)");


[RULE:key_value_format_logs]
alter 
    tmp_get_keys = rtrim(arraystring(arraymap(regextract(_raw_log ,"\s(\S+):\"[^\"]*"), concat("(?P", "<"+"@element"+">", "[^|]+)\|")),""),"\|"),
    tmp_get_values = arraystring(regextract(_raw_log ,"\s\S+:\"([^\"]*)"),"|")

| alter parsed_fields = regexcapture(tmp_get_values, tmp_get_keys)
| fields -tmp_get_keys ,tmp_get_values ;


[RULE:get_hostname_rule]
filter _raw_log ~= "\*\*NF\*\*"
|alter parsed_fields_get_headers = regexcapture(_raw_log,"<(?P<pri>\d{1,3})>(?P<version>\d{1,3})\s(?P<iso_timestamp>\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}(?:\.\d{3})?)(?P<get_offset>[+-]\d{2}:\d{2})\s(?P<hostname>\S+)")
| alter get_time = parse_timestamp("%FT%X%Z", parsed_fields_get_headers -> iso_timestamp+parsed_fields_get_headers -> get_offset)
| alter _time = get_time 
| fields -get_time;


[INGEST:vendor="reblaze", product="waf", target_dataset="reblaze_waf_raw", no_hit = keep]

call get_hostname_rule 
|filter parsed_fields_get_headers -> hostname = "reblazer"
|call key_value_format_logs;

call get_hostname_rule 
|filter parsed_fields_get_headers -> hostname != "reblazer"
|call values_format_logs ;