Limit length of filename

S3 has a limit of 2kb for metadata:

> the user-defined metadata is limited to 2 KB in size. The size of
> user-defined metadata is measured by taking the sum of the number of
> bytes in the UTF-8 encoding of each key and value.

– https://docs.aws.amazon.com/AmazonS3/latest/dev/UsingMetadata.html#object-metadata

This means we have a limit of 1870 bytes for the filename:
```python
encoded = 'notification_count50000template_id665d26e7-ceac-4cc5-82ed-63d773d21561validTrueoriginal_file_name'.encode('utf-8')
sys.getsizeof(b)
>>> 130
2000-130
>>> 1870
```

Or, in other words, ~918 characters:
```python
sys.getsizeof(('ü'*918).encode('utf-8'))
>>> 1869
```

app/main/views/send.py

@@ -46,6 +46,7 @@ from app.utils import (
     get_errors_for_csv,
     get_help_argument,
     get_template,
+    unicode_truncate,
     user_has_permissions,
 )
 
@@ -555,7 +556,10 @@ def _check_messages(service_id, template_id, upload_id, preview_row, letters_as_
             notification_count=len(recipients),
             template_id=str(template_id),
             valid=True,
-            original_file_name=request.args.get('original_file_name'),
+            original_file_name=unicode_truncate(
+                request.args.get('original_file_name', ''),
+                1872
+            ),
         )
     else:
         session['file_uploads'].pop(upload_id)

app/utils.py

@@ -568,3 +568,8 @@ class GovernmentEmailDomain(AgreementInfo):
             ))
         except StopIteration:
             raise NotGovernmentEmailDomain()
+
+
+def unicode_truncate(s, length):
+    encoded = s.encode('utf-8')[:length]
+    return encoded.decode('utf-8', 'ignore')

tests/app/main/views/test_send.py

@@ -1,4 +1,5 @@
 # -*- coding: utf-8 -*-
+import sys
 import uuid
 from functools import partial
 from glob import glob
@@ -662,6 +663,46 @@ def test_upload_valid_csv_shows_preview_and_table(
             assert normalize_spaces(str(row.select('td')[index + 1])) == cell
 
 
+def test_file_name_truncated_to_fit_in_s3_metadata(
+    client_request,
+    mocker,
+    mock_get_live_service,
+    mock_get_service_template_with_placeholders,
+    mock_get_users_by_service,
+    mock_get_detailed_service_for_today,
+    mock_s3_set_metadata,
+    fake_uuid,
+):
+
+    with client_request.session_transaction() as session:
+        session['file_uploads'] = {
+            fake_uuid: {'template_id': fake_uuid}
+        }
+
+    mocker.patch('app.main.views.send.s3download', return_value="""
+        phone number,name,thing,thing,thing
+        07700900001, A,   foo,  foo,  foo
+    """)
+
+    file_name = 'ü😁' * 2000
+
+    client_request.get(
+        'main.check_messages',
+        service_id=SERVICE_ONE_ID,
+        template_id=fake_uuid,
+        upload_id=fake_uuid,
+        original_file_name=file_name,
+    )
+    assert sys.getsizeof(
+        file_name.encode('utf-8')
+    ) > 2000
+
+    assert sys.getsizeof(''.join((
+        '{}{}'.format(key, value) for key, value in
+        mock_s3_set_metadata.call_args_list[0][1].items()
+    )).encode('utf-8')) == 1998
+
+
 def test_show_all_columns_if_there_are_duplicate_recipient_columns(
     client_request,
     mocker,