#* masked_topic_weight_totals_nodps:
#*   attr:
#*     fillcolor: '3'
#*   desc: Psuedorandomly mask site IDs in topic weight totals to match other psuedorandom
#*     IDs.
#*   ext: py
#*   inputs:
#*   - topic_weight_totals_nodps
#* 

import pyspark.sql.functions as F
from pyspark.sql.types import IntegerType,BooleanType,DateType,StringType

def masked_topic_weight_totals_nodps(topic_weight_totals_nodps):
    enabled = True
    if enabled:
        munged = topic_weight_totals_nodps.withColumn('data_partner_id', F.sha2(F.concat(topic_weight_totals_nodps.data_partner_id.cast(StringType()), F.lit('seed')), 256).substr(0, 3))
        return munged
    else:
        return topic_weight_totals_nodps