連続画像でもコメント作成等の対応

README.md

@@ -2915,6 +2915,7 @@ Example of a comment object
     "id": "YOUR_COMMENT_ID",
     "taskId": "YOUR_TASK_ID",
     "contentId": "YOUR_CONTENT_ID",
+    "contentName": "01.jpg",
     "type": "text",
     "isResolved": False,
     "points": [185.98, 86.55],
@@ -2970,6 +2971,7 @@ Example of a comment thread object
         "id": "YOUR_COMMENT_ID",
         "taskId": "YOUR_TASK_ID",
         "contentId": "YOUR_CONTENT_ID",
+        "contentName": "01.jpg",
         "type": "text",
         "isResolved": False,
         "points": [185.98, 86.55],
@@ -3001,14 +3003,29 @@ comment = client.create_task_comment(
 )
 ```
 
+For a task with multiple contents (e.g. sequential image), specify which frame
+the comment belongs to with `content_name`. The name is the file name returned
+as `contents[].name` by Get Tasks, and `contents` is not guaranteed to be
+ordered, so match on the name rather than on the array index.
+
+```python
+comment = client.create_task_comment(
+    task_id="YOUR_TASK_ID",
+    content_name="01.jpg",
+    points=[185.98, 86.55],
+    text="comment on frame 01.jpg",
+)
+```
+
 #### Parameters
 
 | Name | Type | Required | Description |
 |------|------|----------|-------------|
 | task_id | str | Yes | Task ID the comment belongs to (the task is resolved by task_id alone) |
 | points | list | Yes | Comment position as a numeric array |
 | text | str | Yes | Comment body text |
-| content_id | str | No | Content ID. If omitted and the task has a single content it is auto-selected; required for multi-content tasks |
+| content_name | str | No | Frame file name (`contents[].name` from Get Tasks). Use this to target a frame of a multi-content task such as sequential image. Mutually exclusive with `content_id` |
+| content_id | str | No | Content ID. Mutually exclusive with `content_name`. If both are omitted and the task has a single content it is auto-selected; one of them is required for multi-content tasks |
 | type | str | No | Comment type. Only `text` is supported (default: `text`) |
 | scale | float | No | Canvas scale |
 | frame | int | No | Frame index for sequential/video (1-indexed) |

fastlabel/__init__.py

@@ -5545,6 +5545,7 @@ def create_task_comment(
         task_id: str = None,
         points: list = None,
         text: str = None,
+        content_name: str = None,
         content_id: str = None,
         type: str = "text",
         scale: float = 0,
@@ -5559,7 +5560,16 @@ def create_task_comment(
         """
         Create a comment, or add a thread (message/reply) to an existing comment.
         The author is recorded as "via API".
+
+        For a task with multiple contents (e.g. sequential image), specify the
+        frame with content_name, which is the file name returned as
+        contents[].name by get_tasks. content_id is also accepted, but the two
+        are mutually exclusive.
         """
+        if content_name is not None and content_id is not None:
+            raise FastLabelInvalidException(
+                "Specify either content_name or content_id, not both.", 422
+            )
         if comment_id is not None:
             return self.api.post_request(
                 "comments/" + comment_id + "/threads", payload={"text": text}
@@ -5574,6 +5584,8 @@ def create_task_comment(
             "scale": scale,
             "frame": frame,
         }
+        if content_name is not None:
+            payload["contentName"] = content_name
         if content_id is not None:
             payload["contentId"] = content_id
         if status is not None: