Can´t read csv files from S3

[jplicona]

Hi, I want to read csv and parquet files from S3 AWS in my local machine using VSCode. However, I am not quite sure how to do it. I have been trying to find something on the documentation, but it hasnt been helpful.

using AWS.AWSServices: s3
s3("GET", "s3://test1/test_file.csv")

#GET THE FOLLOWING ERROR:
ERROR: Can't find AWS credentials!

How can i specify the credentials (aws_access_key_id, aws_secret_access_key)? What is the right way to write the julia script?
In python I was able to connect to S3 by using boto3 and awswrangler:

import awswrangler as wr
import boto3

# Configure AWS credentials
session = boto3.Session(
    aws_access_key_id="123456789",
    aws_secret_access_key="asdfghjklqwertyuiozxcvbnm"
)

import awswrangler as wr

df = wr.s3.read_csv(
    path="s3://test1/test_file.csv",
    boto3_session=session
)

print(df)

Help. Thanks in advance

[ararslan]

In case this is still relevant for you nearly one year to the day from when you posted it...

There are a few ways you could go about this. The workflow I typically use is with AWSS3.jl and its S3Path type, which can be passed directly to CSV.read. The only use of AWS.jl in this case is for credential management.

using AWS, AWSS3, CSV, DataFrames

# If you don't have a profile defined in `~/.aws/config`:
config = AWSConfig(AWSCredentials("123456789", "asdfghjklqwertyuiozxcvbnm"))
# Or if you do have a profile, you can use it by name:
config = AWSConfig(; profile="whatever")

df = CSV.read(S3Path("s3://test1/test_file.csv"; config), DataFrame)