LIKE operator working on AWS lambda function but not =

Question

I have a small csv file that looks like that :

is_employee,candidate_id,gender,hesa_type,university
FALSE,b9bb80,Male,Mathematical sciences,Birmingham
FALSE,8e552d,Female,Computer science,Swansea
TRUE,2bc475,Male,Engineering & technology,Aston
TRUE,c3ac8d,Female,Mathematical sciences,Heriot-Watt
FALSE,ceb2fa,Female,Mathematical sciences,Imperial College London

The following lambda function is used to query from an s3bucket.

import boto3
import os
import json

def lambda_handler(event, context):
    BUCKET_NAME = 'foo'
    KEY = 'bar/data.csv'  
    s3 = boto3.client('s3','eu-west-1')
    response = s3.select_object_content(
        Bucket = BUCKET_NAME,
        Key = KEY,
        ExpressionType = 'SQL',
        Expression = 'Select count(*) from s3object s where s.gender like \'%Female%\'',
        InputSerialization = {'CSV': {"FileHeaderInfo": "Use"}},
        OutputSerialization = {'JSON': {}},
    )

    for i in response['Payload']:
        if 'Records' in i:
            query_result = i['Records']['Payload'].decode('utf-8')

    print(list(json.loads(query_result).values())[0])

Now, this works great as I get back a result of 3. But for some reason the same code does not work when changing the like operator to =, results drop down to 0, so no match found. What's happening here ?

Answer 1

So I found the problem. The problem was that the items of the last column were followed by a newline character, which was not understood by the AWS S3 interpreter. So really, a university name was not Swansea, but more Swansea\n.

So s.university = \'Swansea\'' does not work; however, s.university LIKE \'Swansea%\'' does work, and is still a sargable expression.