From d712d43c589173376e2348d2dfa294a247407747 Mon Sep 17 00:00:00 2001
From: Joel Balcaen <jobalcaen+github@gmail.com>
Date: Fri, 26 Apr 2024 11:02:37 -0300
Subject: [PATCH] new lamda

---
 lambdas/document_filler/lambda.tf            | 64 ++++++++++++++++++++
 lambdas/document_filler/output.tf            |  7 +++
 lambdas/document_filler/src/index.py         | 37 +++++++++++
 lambdas/document_filler/src/requirements.txt |  1 +
 lambdas/document_filler/variables.tf         | 15 +++++
 terraform/modules.tf                         |  8 +++
 6 files changed, 132 insertions(+)
 create mode 100644 lambdas/document_filler/lambda.tf
 create mode 100644 lambdas/document_filler/output.tf
 create mode 100644 lambdas/document_filler/src/index.py
 create mode 100644 lambdas/document_filler/src/requirements.txt
 create mode 100644 lambdas/document_filler/variables.tf

diff --git a/lambdas/document_filler/lambda.tf b/lambdas/document_filler/lambda.tf
new file mode 100644
index 0000000..49ea866
--- /dev/null
+++ b/lambdas/document_filler/lambda.tf
@@ -0,0 +1,64 @@
+locals {
+  lambda_function_name = "levio-esta-document-filler"
+  timeout              = 30
+  runtime              = "python3.11"
+  powertools_layer_arn = "arn:aws:lambda:${var.aws_region}:017000801446:layer:AWSLambdaPowertoolsPythonV2:67"
+}
+
+data "aws_caller_identity" "current" {}
+
+
+module "lambda_function_container_image" {
+  source = "terraform-aws-modules/lambda/aws"
+  function_name = local.lambda_function_name
+  handler       = "index.lambda_handler"
+  publish       = true
+  runtime = local.runtime
+  timeout = local.timeout
+  layers  = [local.powertools_layer_arn]
+  source_path = "${path.module}/src"
+  s3_bucket   = var.lambda_storage_bucket
+  memory_size              = 256
+  role_name                = "${local.lambda_function_name}-role"
+  attach_policy_statements = true
+
+  policy_statements = {
+    log_group = {
+      effect = "Allow"
+      actions = [
+        "logs:CreateLogGroup"
+      ]
+      resources = [
+        "arn:aws:logs:*:*:*"
+      ]
+    }
+
+    s3 = {
+      effect = "Allow"
+      actions = [
+        "s3:Get*",
+        "s3:List*",
+        "s3:Describe*",
+        "s3:PutObject",
+        "s3-object-lambda:Get*",
+        "s3-object-lambda:List*",
+        "s3-object-lambda:WriteGetObjectResponse"
+      ]
+      resources = var.allowed_s3_resources
+    }
+
+    log_write = {
+      effect = "Allow"
+
+      resources = [
+        "arn:aws:logs:*:*:log-group:/aws/${local.lambda_function_name}/*:*"
+      ]
+
+      actions = [
+        "logs:CreateLogStream",
+        "logs:PutLogEvents",
+      ]
+    }
+
+  }
+}
diff --git a/lambdas/document_filler/output.tf b/lambdas/document_filler/output.tf
new file mode 100644
index 0000000..cd2e3af
--- /dev/null
+++ b/lambdas/document_filler/output.tf
@@ -0,0 +1,7 @@
+output "lambda_function_arn" {
+  value = module.lambda_function_container_image.lambda_function_arn
+}
+
+output "lambda_function_name" {
+  value = module.lambda_function_container_image.lambda_function_name
+}
diff --git a/lambdas/document_filler/src/index.py b/lambdas/document_filler/src/index.py
new file mode 100644
index 0000000..7393037
--- /dev/null
+++ b/lambdas/document_filler/src/index.py
@@ -0,0 +1,37 @@
+import boto3
+from docx import Document
+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 document_key with the text_to_fill
+    """
+    s3_arn = event['doc_s3_arn']
+    replacements = event['replacements']
+
+    bucket_name = s3_arn.split(':')[5].split('/')[0]
+    key = '/'.join(s3_arn.split(':')[5].split('/')[1:])
+
+    print(f"Download bucket: {bucket_name}, key: {key}")
+
+    file_obj = s3.get_object(Bucket=bucket_name, Key=key)
+    file_content = file_obj['Body'].read()
+
+    doc = Document(BytesIO(file_content))
+
+    for paragraph in doc.paragraphs:
+        for replacement in replacements:
+            if replacement['document_key'] in paragraph.text:
+                paragraph.text = paragraph.text.replace(replacement['document_key'], replacement['text_to_fill'])
+
+    output_stream = BytesIO()
+    doc.save(output_stream)
+
+    s3.put_object(Bucket=bucket_name, Key=key, Body=output_stream.getvalue())
+
+    return {
+        'statusCode': 200,
+        'body': f'Successfully modified {key} and uploaded to {bucket_name}'
+    }
\ No newline at end of file
diff --git a/lambdas/document_filler/src/requirements.txt b/lambdas/document_filler/src/requirements.txt
new file mode 100644
index 0000000..7a6646f
--- /dev/null
+++ b/lambdas/document_filler/src/requirements.txt
@@ -0,0 +1 @@
+docx
\ No newline at end of file
diff --git a/lambdas/document_filler/variables.tf b/lambdas/document_filler/variables.tf
new file mode 100644
index 0000000..45ce8a5
--- /dev/null
+++ b/lambdas/document_filler/variables.tf
@@ -0,0 +1,15 @@
+variable "lambda_storage_bucket" {
+  type     = string
+  nullable = false
+}
+
+variable "aws_region" {
+  type     = string
+  nullable = false
+}
+
+variable "allowed_s3_resources" {
+  type     = list(string)
+  nullable = false
+  description = "values for the s3 resources that the lambda function can access"
+}
\ No newline at end of file
diff --git a/terraform/modules.tf b/terraform/modules.tf
index ecae810..d966afb 100644
--- a/terraform/modules.tf
+++ b/terraform/modules.tf
@@ -275,4 +275,12 @@ 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"
+  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}/*"]
+}
+
+