update

lambdas/bedrock_invoker/src/index.py

@@ -50,7 +50,6 @@ def lambda_handler(event, context):
         ]
     }
 
-    print(f"Invoke bedrock with this body: ", claude_body)
     bedrock_model = 'anthropic.claude-3-sonnet-20240229-v1:0'
     print(f"Invoke bedock with this model: ", bedrock_model)
     print(f"Invoke claude with system prompt: ", system_prompt)

lambdas/email_sender/src/index.py

@@ -1,3 +1,4 @@
+import json
 import boto3
 from botocore.exceptions import NoCredentialsError
 from email.mime.multipart import MIMEMultipart
@@ -7,6 +8,7 @@
 def lambda_handler(event, context):
     """
     Downloads the attachment_s3_arns into memory and add them as attachments before sending the email with the given params.
+    body is an array of json strings with the following format: {type: "plain/html/xml/csv", message: "Hello World"}
     attachment_s3_arns can be in URI or ARN format
     """
     sender_email = event['sender_email']
@@ -22,7 +24,11 @@ def lambda_handler(event, context):
     msg['Subject'] = subject
     msg['From'] = sender_email
     msg['To'] = destination_email
-    msg.attach(MIMEText(body, 'plain'))
+
+    for mime_text_json in body:
+        mime_text_dict = json.loads(mime_text_json)
+        mime_text = MIMEText(mime_text_dict['message'], mime_text_dict['type'])
+        msg.attach(mime_text)
 
     for attachment in attachment_s3_arns:
         try:

lambdas/document_filler/lambda.tf → lambdas/text_replacer/lambda.tf

@@ -1,5 +1,5 @@
 locals {
-  lambda_function_name = "levio-esta-document-filler"
+  lambda_function_name = "levio-esta-text-replacer"
   timeout              = 30
   runtime              = "python3.11"
   powertools_layer_arn = "arn:aws:lambda:${var.aws_region}:017000801446:layer:AWSLambdaPowertoolsPythonV2:67"
@@ -21,17 +21,17 @@ module "lambda_function_container_image" {
   role_name                = "${local.lambda_function_name}-role"
   attach_policy_statements = true
   # see https://github.com/terraform-aws-modules/terraform-aws-lambda/issues/346 on why this custom path is necessary
-  source_path = [
-    {
-      path = "${path.module}/src"
-      commands = [
-        ":zip",
-        "cd `mktemp -d`",
-        "python3 -m pip install --no-compile --only-binary=:all: --platform=manylinux2014_x86_64 --target=. -r ${abspath(path.module)}/src/requirements.txt",
-        ":zip .",
-      ]
-    }
-  ]
+  # source_path = [
+  #   {
+  #     path = "${path.module}/src"
+  #     commands = [
+  #       ":zip",
+  #       "cd `mktemp -d`",
+  #       "python3 -m pip install --no-compile --only-binary=:all: --platform=manylinux2014_x86_64 --target=. -r ${abspath(path.module)}/src/requirements.txt",
+  #       ":zip .",
+  #     ]
+  #   }
+  # ]
 
   policy_statements = {
     log_group = {

lambdas/document_filler/src/index.py → lambdas/text_replacer/src/index.py

@@ -1,37 +1,31 @@
 import boto3
-from docx import Document
+import re
 from io import BytesIO
 
 s3 = boto3.client('s3')
 
-
 def lambda_handler(event, context):
     """
-    Downloads the given docx and for each replacement item it replaces the matching replacement_key with the replacement_text 
+    Downloads the given text file (txt, html, json, csv, xml, js, py, md etc) and for each replacement item it replaces the matching replacement_key with the replacement_text 
     """
     try:
         s3_arn = event['doc_s3_arn']
         replacements = event['replacements']
 
         s3_bucket, s3_key = s3_arn.replace("s3://", "").replace("arn:aws:s3:::", "").split("/", 1)
 
-
         print(f"Download bucket: {s3_bucket}, key: {s3_key}")
 
         file_obj = s3.get_object(Bucket=s3_bucket, Key=s3_key)
 
-        file_content = file_obj['Body'].read()
+        file_content = file_obj['Body'].read().decode('utf-8')
 
-        doc = Document(BytesIO(file_content))
-
-        for paragraph in doc.paragraphs:
-            for replacement in replacements:
-                if replacement['replacement_key'] in paragraph.text:
-                    print(f"replacing {replacement['replacement_key']} with {replacement['replacement_text'][:10]}")
-                    paragraph.text = paragraph.text.replace(replacement['replacement_key'], replacement['replacement_text'])
+        for replacement in replacements:
+            print(f"replacing {replacement['replacement_key']} with {replacement['replacement_text'][:10]}")
+            file_content = re.sub(re.escape(replacement['replacement_key']), replacement['replacement_text'], file_content)
 
         output_stream = BytesIO()
-        doc.save(output_stream)
+        output_stream.write(file_content.encode('utf-8'))
 
         s3.put_object(Bucket=s3_bucket, Key=s3_key, Body=output_stream.getvalue())
 
@@ -44,4 +38,4 @@ def lambda_handler(event, context):
         return {
             'statusCode': 400,
             'body': str(e)
-        }
+        }

terraform/modules.tf

@@ -275,8 +275,8 @@ module "bedrock_invoker" {
   allowed_s3_resources  = [module.s3_bucket.s3_bucket_arn, "${module.s3_bucket.s3_bucket_arn}/*"]
 }
 
-module "levio_esta_document_filler" {
-  source                = "../lambdas/document_filler"
+module "levio_esta_text_replacer" {
+  source                = "../lambdas/text_replacer"
   lambda_storage_bucket = aws_s3_bucket.lambda_storage.id
   aws_region            = var.aws_region
   allowed_s3_resources  = [module.s3_bucket.s3_bucket_arn, "${module.s3_bucket.s3_bucket_arn}/*"]