如何通过ID而非名称从Google Cloud Storage获取对象？

Got it, let's tackle this problem! First, let's clarify a key detail about Google Cloud Storage (GCS) object IDs: an object's unique ID is actually a structured string that encodes both the bucket name and object name you need for the standard API methods. Typically, it looks like projects/_/buckets/{BUCKET_NAME}/objects/{OBJECT_NAME} (or sometimes the shorter {BUCKET_NAME}/{OBJECT_NAME} format).

Here's how you can use this to fetch your object:

Step 1: Parse the Object ID #

Extract the bucket name and object name from the ID string:

• For an ID like projects/_/buckets/my-bucket/objects/reports/2024/Q1/sales.csv:
  • Bucket name: my-bucket
  • Object name: reports/2024/Q1/sales.csv

Step 2: Use the Standard objects.get Method #

Once you have those two values, you can use the regular objects.get API exactly as documented. Here are examples for common use cases:

REST API Request

Send a GET request to the endpoint, replacing placeholders with your parsed values:

https://storage.googleapis.com/storage/v1/b/{BUCKET_NAME}/o/{OBJECT_NAME}

(Remember to URL-encode the object name if it contains special characters like slashes or spaces.)

Python Client Library Example

from google.cloud import storage

def fetch_object_by_id(object_id):
    # Parse the ID to extract bucket and object name
    id_parts = object_id.split("/")
    bucket_name = id_parts[3]
    object_name = "/".join(id_parts[5:])

    # Initialize GCS client and fetch the object
    storage_client = storage.Client()
    bucket = storage_client.bucket(bucket_name)
    blob = bucket.blob(object_name)
    
    # Get metadata or download content as needed
    object_metadata = blob.metadata
    object_content = blob.download_as_bytes()
    return object_metadata, object_content

Alternative: Use the Full ID Directly #

Some GCS API endpoints accept the complete object ID as the object parameter. You can try passing the full ID directly to objects.get without parsing it first. For example:

GET https://storage.googleapis.com/storage/v1/b/{BUCKET_NAME}/o/{FULL_OBJECT_ID}

Note: You'll need to URL-encode the full ID if it contains special characters to avoid request errors.

Why This Works #

GCS doesn't have a dedicated "get by ID" endpoint, but since the object ID is a unique, structured identifier that directly maps to the bucket and object name, parsing it gives you all the parameters required for the standard API methods. This approach is reliable and aligns with GCS's core object addressing model.

内容的提问来源于stack exchange，提问作者acidboy