Merge pull request #19 from nationalarchives/set-mime-type

Set mime type

.gitignore

@@ -3,3 +3,4 @@
 *#
 .DS_Store
 .env
+__pycache__

.pre-commit-config.yaml

@@ -3,24 +3,24 @@ default_stages: [commit]
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.1.0
+    rev: v4.4.0
     hooks:
       - id: trailing-whitespace
       - id: end-of-file-fixer
       - id: check-yaml
 
   - repo: https://github.com/psf/black
-    rev: 22.3.0
+    rev: 23.1.0
     hooks:
       - id: black
 
   - repo: https://github.com/PyCQA/isort
-    rev: 5.10.1
+    rev: 5.12.0
     hooks:
       - id: isort
 
   - repo: https://github.com/PyCQA/flake8
-    rev: 4.0.1
+    rev: 6.0.0
     hooks:
       - id: flake8
         args: ["--config=setup.cfg"]

README.md

@@ -49,8 +49,10 @@ The script scripts/create_json_for_bulk_pdf_regeneration will make that JSON
 file for you, if you want to remake every PDF that's backed by a docx file.
 
 (The eTag is arbitrary but should be a sensible filename fragment, no / )
-Start up:
+
+## Local setup
 
 1. From ds-caselaw-ingester, run `docker-compose up` to launch the Localstack container
-2. From ds-caselaw-pdfconversion, run `script setup-localstack` to set up the queues etc.
-3. From ds-caselaw-pdfconversion, run `docker-compose up` to launch the LibreOffice container
+2. From ds-caselaw-pdfconversion, run `scripts/setup-localstack` to set up the queues etc.
+3. From ds-caselaw-pdfconversion, run `docker-compose up --build` to launch the LibreOffice container
+   (`--build` will ensure the converter script is in the docker container)

docker-compose-localstack.yml

@@ -10,5 +10,5 @@ services:
     ports:
       - 4566:4566
     volumes:
-      - "${TMPDIR:-/tmp}/localstack:/tmp/localstack"
+      - "${LOCALSTACK_VOLUME_DIR:-./volume}:/var/lib/localstack"
       - "/var/run/docker.sock:/var/run/docker.sock"

queue_listener/queue_listener.py

@@ -60,7 +60,10 @@
             # NOTE: there's a risk that some.pdf doesn't exist, we need to handle that case.
             try:
                 s3_client.upload_file(
-                    Bucket=bucket_name, Key=upload_key, Filename=pdf_filename
+                    Bucket=bucket_name,
+                    Key=upload_key,
+                    Filename=pdf_filename,
+                    ExtraArgs={"ContentType": "application/pdf"},
                 )
                 print(f"Uploaded {upload_key}")
             except FileNotFoundError as exception: