- Recipes
- Amazon S3 CSV file to OpenAI prompt
Connect Amazon S3 CSV file and OpenAI prompt in our serverless environment
Use this template to Read CSV file entries from Amazon S3 bucket using them to send a prompt to OpenAI and generate completions.
Share
Read CSV file entries from Amazon S3 bucket
Used integrations:
JavaScript
Python
class AwsS3SourceReadRemoteCsv {
async init() {
// TODO: Create your aws-s3 credential
// More info at https://yepcode.io/docs/integrations/aws-s3/#credential-configuration
this.awsS3 = yepcode.integration.awsS3("your-aws-s3-credential-name");
}
async fetch(publish, done) {
// TODO: Customize the command to be executed
// More info at: https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/clients/client-s3/interfaces/getobjectcommandinput.html
const getObjectCommand = new GetObjectCommand({
Bucket: "your-bucket-name",
Key: "your-file-name",
});
const getObjectResult = await this.awsS3.send(getObjectCommand);
// Result.Body is a stream of object content
getObjectResult.Body.pipe(
csv.parse({
delimiter: ",",
columns: true,
})
)
.on("data", publish)
.on("end", done);
}
async close() {}
}
import csv
import io
class AwsS3SourceReadRemoteCsv:
def setup(self):
# TODO: Create your S3 credential:
# More info at https://yepcode.io/docs/integrations/aws-s3/#credential-configuration
self.aws_s3_client = yepcode.integration.awsS3("your-s3-credential-name")
# TODO: If your csv does no have headers, you can define them here as a list:
# self.fieldnames = ["column1", "column2", "column3"]
self.fieldnames = None
def generator(self):
# TODO: Customize your bucket name and object key
# See all supported params in: https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/s3/client/get_object.html
response = self.aws_s3_client.get_object(
Bucket="your-bucket-name",
Key="your-file-name",
)
bytes_stream = response["Body"]
csv_file_stream = io.TextIOWrapper(bytes_stream, encoding="utf-8")
reader = csv.DictReader(
csv_file_stream, delimiter=",", fieldnames=self.fieldnames
)
for row in reader:
yield row
def close(self):
pass
Do you need help solving this integration with YepCode?
Let's talkSend a prompt to OpenAI and generate completions
Used integrations:
- JavaScript
- Python
class OpenaiTargetSendPrompt {
async init() {
// TODO: Create your OpenAI credential
// More info at https://yepcode.io/docs/integrations/openai/#credential-configuration
this.openai = yepcode.integration.openai("your-openai-credential-name");
}
async consume(item) {
// TODO: Configure your prompt
// More info at: https://platform.openai.com/docs/api-reference/completions/create
const completion = await this.openai.completions.create({
model: item.model,
prompt: item.prompt,
max_tokens: item.maxTokens,
temperature: item.temperature,
});
// Use the completion for whatever you want
console.log(completion.data.choices[0].text);
}
async close() {}
}
class OpenaiTargetSendPrompt:
def setup(self):
# TODO: Create your OpenAI credential
# More info at https://yepcode.io/docs/integrations/openai/#credential-configuration
self.openai = yepcode.integration.openai("your-openai-credential-name")
def consume(self, generator, done):
# TODO: Configure your prompt
# More info at: https://platform.openai.com/docs/api-reference/completions/create
for item in generator:
completion = self.openai.completions.create(
model="text-davinci-003",
prompt=item.get("prompt"),
max_tokens=7,
temperature=0,
)
# Use the completion for whatever you want
print(completion.choices)
done()
def close(self):
pass
CSVFAQs
YepCode is a SaaS platform that enables the creation, execution and monitoring of integrations and automations using source code in a serverless environment.
We like to call it the Zapier for developers, since we bring all the agility and benefits of NoCode tools (avoid server provisioning, environment configuration, deployments,...), but with all the power of being able to use a programming language like JavaScript or Python.
These recipes are an excellent starting point for creating your own YepCode processes and solving complex integration and automation problems.
You only have to complete the sign up form and your account will be created with our FREE plan (no credit card required).
YepCode has been created with a clear enterprise focus, offering a multi-tenant environment, team management capabilities, high security and auditing standards, Identity Provider (IdP) integrations, and on-premise options. It serves as the Swiss army knife for engineering teams, especially those requiring the extraction or transmission of information to external systems. It excels in scenarios demanding flexibility and adaptability to change within the process.
Sure! You only need to configure YepCode servers to establish a connection with that service. Check our docs page to get more information.