59 lines
2.0 KiB
Python
59 lines
2.0 KiB
Python
"""strip presigned query params from avatar URLs
|
|
|
|
Revision ID: a1b2c3d4e5f6
|
|
Revises: 6a41fa07bd94
|
|
Create Date: 2026-02-21 12:00:00.000000
|
|
|
|
"""
|
|
from typing import Sequence, Union
|
|
|
|
from alembic import op
|
|
import sqlalchemy as sa
|
|
|
|
# revision identifiers, used by Alembic.
|
|
revision: str = "a1b2c3d4e5f6"
|
|
down_revision: Union[str, None] = "6a41fa07bd94"
|
|
branch_labels: Union[str, Sequence[str], None] = None
|
|
depends_on: Union[str, Sequence[str], None] = None
|
|
|
|
|
|
def upgrade() -> None:
|
|
conn = op.get_bind()
|
|
|
|
# Extract bare S3 key from presigned URLs.
|
|
# Example: http://localhost:9000/coffee-bucket/avatars/abc.jpg?X-Amz-...
|
|
# -> avatars/abc.jpg
|
|
#
|
|
# Strategy: strip the query string, then remove the scheme+host+bucket prefix.
|
|
# Only touch rows where avatar looks like a full URL (contains '://').
|
|
result = conn.execute(
|
|
sa.text("SELECT id, avatar FROM users WHERE avatar IS NOT NULL AND avatar LIKE :pattern"),
|
|
{"pattern": "%://%"},
|
|
)
|
|
for row in result:
|
|
avatar_url: str = row[1]
|
|
# Remove query string
|
|
path = avatar_url.split("?")[0]
|
|
# Remove scheme + host + bucket prefix: everything up to and including the bucket segment
|
|
# e.g. "http://localhost:9000/coffee-bucket/avatars/abc.jpg" -> "avatars/abc.jpg"
|
|
parts = path.split("/")
|
|
# Find the bucket segment (after host) and take everything after it
|
|
# URL structure: scheme: / / host:port / bucket / key...
|
|
# parts: ['http:', '', 'localhost:9000', 'coffee-bucket', 'avatars', 'abc.jpg']
|
|
try:
|
|
# Skip scheme ('http:'), empty (''), host, bucket -> index 4 onward is the key
|
|
key = "/".join(parts[4:])
|
|
except IndexError:
|
|
continue
|
|
|
|
if key:
|
|
conn.execute(
|
|
sa.text("UPDATE users SET avatar = :key WHERE id = :id"),
|
|
{"key": key, "id": row[0]},
|
|
)
|
|
|
|
|
|
def downgrade() -> None:
|
|
# Data-only migration; cannot restore original presigned URLs.
|
|
pass
|