HTTP 401 on download_attachments_from_page() but not get_attachments_from_content()

[jacenat]

I am trying to download attachments from confluence pages. Because I have been getting 401 errors when calling download_attachments_from_page() (it worked yesterday?) I also included get_attachments_from_content().

def downloadFileFromConfluence(logger, Confluence, confluencePageID, dlFolder):
    logger.info("[" + str(datetime.datetime.now()) + "] --- Start of download from Confluence.")
    try:        
        # Download all attachments from the given page
        attachments = Confluence.get_attachments_from_content(confluencePageID, start=0, limit=50, expand=None, filename=None, media_type=None)
        print(attachments)
        Confluence.download_attachments_from_page(confluencePageID, path=dlFolder)
    except Exception as e:
        print(str(e))
        logger.error("[" + str(datetime.datetime.now()) + "] --- An error occured while accessing Confluence!")
        logger.error("[" + str(datetime.datetime.now()) + "] " + str(e))

The console output from this piece of code is

{'results': [{'id': 'att67907559', 'type': 'attachment', 'ari': 'ari:cloud:confluence:8babfdf0-f329-48b6-9de1-52a8a936740f:attachment/67907559', 'base64EncodedAri': 'YXJpOmNsb3VkOmNvbmZsdWVuY2U6OGJhYmZkZjAtZjMyOS00OGI2LTlkZTEtNTJhOGE5MzY3NDBmOmF0dGFjaG1lbnQvNjc5MDc1NTk=', 'status': 'current', 'title': 'Amaris Overview 220221.pdf', 'macroRenderedOutput': {}, 'metadata': {'mediaType': 'application/pdf'}, 'extensions': {'mediaType': 'application/pdf', 'fileSize': 1519699, 'comment': '', 'mediaTypeDescription': 'PDF Document', 'fileId': '08f7fe95-ad82-42db-9af9-1413c54d3742', 'collectionName': 'contentId-67895904'}, '_expandable': {'container': '/rest/api/content/67895904', 'restrictions': '/rest/api/content/att67907559/restriction/byOperation', 'history': '/rest/api/content/att67907559/history', 'body': '', 'version': '', 'descendants': '/rest/api/content/att67907559/descendant', 'space': '/rest/api/space/IT', 'childTypes': '', 'schedulePublishInfo': '', 'operations': '', 'schedulePublishDate': '', 'children': '/rest/api/content/att67907559/child', 'ancestors': '', 'draftVersion': ''}, '_links': {'webui': '/pages/viewpageattachments.action?pageId=67895904&preview=%2F67895904%2F67907559%2FAmaris+Overview+220221.pdf', 'self': 'https://<our org>.atlassian.net/wiki/rest/api/content/att67907559', 'download': '/download/attachments/67895904/Amaris%20Overview%20220221.pdf?version=1&modificationDate=1646135898000&cacheVersion=1&api=v2'}}], 'start': 0, 'limit': 50, 'size': 1, '_links': {'base': 'https://<our org>.atlassian.net/wiki', 'context': '/wiki', 'self': 'https://<our org>.atlassian.net/wiki/rest/api/content/67895904/child/attachment'}}
HTTP error occurred while downloading attachments: Unauthorized (401)

Getting the list of the (one) attachment(s) works. But once I want to download them, the call gives me an "Unauthorized (401)". I checked this also via stepping through, just so I am sure that download_attachments_from_page() is throwing that 401.
I MUST be doing something wrong. The user and api token can curl the page without problems on the command line.

Before I re-write the rest calls myself (which I'd rather not), I thought it can't hurt to ask here.

[MarcoFierimonte]

I thought I was going crazy, but I have the same issue...exactly the same. Even my code was working well yesterday, and today I can’t download any attachments. Could this be a global server error?

[jacenat]

For reference here: I did open a ticket with atlassian, but their stance is: "curl with that user works fine -> not our prob" which I understand. But I did not change/update atlassian-python-api since at least monday (probably last week) and it did work yesterday.

I have no idea why this would even the case. I am really confused why something works and other things dont.

[MarcoFierimonte]

Yes, that's weird!
I confirm that using the Official Rest API it works, example:

curl --request GET \
    --url 'https://mycompany.atlassian.net/wiki/rest/api/content/500589740/child/attachment/att938510701/download' \
    --user 'mfierimonte@mycompany.com:<api_token>' \
    -L \
    -o "image_att938510701.png" \
    -s -w "%{http_code}\n"

[jacenat]

I was reading through the Confluence changelog here https://developer.atlassian.com/cloud/confluence/changelog/#
For 19.11.2025 there is

19 November 2025
Removed Removed */download/attachments/* internal API’s

Following our [prior deprecation notice](https://developer.atlassian.com/cloud/confluence/changelog/#CHANGE-2735), requests to the following internal API endpoints will now fail.

/download/attachments/{id}/crash-scraper/files/{id}

/download/attachments/{id}/{id}/customer-card-payment-rest

/download/attachments/{id}/{id}

/download/attachments/{id}/{id}/{id}

/download/download/attachments/{id}/{id}

/download/attachments/{id}/{id}/{id}/{id}

/download/attachments/{id}/attachments/{id}/{id}/{id}

/download/attachments/{id}/{id}/image

/download/all_attachments

prior deprecation notice: https://developer.atlassian.com/cloud/confluence/changelog/#CHANGE-2735

I might be mistaken, but this looks eerily like something that could impact here. I am very certain I was still working on Monday, but maybe the replication of this took a few days?

[jacenat]

I had to step away from the project for a bit. Tried again just now and it works again? No errors?
I did not update atlassian-python-api yet (didn't look if there is a new release) so it must have been on Atlassian side.

[jacenat]

@MarcoFierimonte if you can confirm it's working for you, I will close the issue, so it doesn't pollute the issue feed.

[sebastianreloaded]

for me, "download_attachments_from_page" works only when using a personal apitoken. when i use a serviceaccount apitoken it fails with 401.

Edit: Ok, this has nothing todo with service accounts, more with scoped tokens which use a different api.

[dresdner353]

Was also getting the 401 errors. I ended up getting it to work this way..

            download_url = '%s/rest/api/content/%s/child/attachment/%s/download' % (
                    confluence.url, 
                    page_id, 
                    attachment_id
                    )

[Chekov2k]

You can fix the code like this as well using the url snippet from above

class CustomConfluence(Confluence):
    def download_attachments_from_page(self, page_id, path=None, start=0, limit=50, filename=None, to_memory=False):
        """
        Downloads attachments from a Confluence page. Supports downloading all files or a specific file.
        Files can either be saved to disk or returned as BytesIO objects for in-memory handling.

        :param page_id: str
            The ID of the Confluence page to fetch attachments from.
        :param path: str, optional
            Directory where attachments will be saved. If None, defaults to the current working directory.
            Ignored if `to_memory` is True.
        :param start: int, optional
            The start point for paginated attachment fetching. Default is 0. Ignored if `filename` is specified.
        :param limit: int, optional
            The maximum number of attachments to fetch per request. Default is 50. Ignored if `filename` is specified.
        :param filename: str, optional
            The name of a specific file to download. If provided, only this file will be fetched.
        :param to_memory: bool, optional
            If True, attachments are returned as a dictionary of {filename: BytesIO object}.
            If False, files are written to the specified directory on disk.
        :return:
            - If `to_memory` is True, returns a dictionary {filename: BytesIO object}.
            - If `to_memory` is False, returns a summary dict: {"attachments_downloaded": int, "path": str}.
        :raises:
            - FileNotFoundError: If the specified path does not exist.
            - PermissionError: If there are permission issues with the specified path.
            - requests.HTTPError: If the HTTP request to fetch an attachment fails.
            - Exception: For any unexpected errors.
        """
        # Default path to current working directory if not provided
        if not to_memory and path is None:
            path = os.getcwd()

        try:
            # Fetch attachments based on the specified parameters
            if filename:
                # Fetch specific file by filename
                attachments = self.get_attachments_from_content(page_id=page_id, filename=filename)["results"]
                if not attachments:
                    return f"No attachment with filename '{filename}' found on the page."
            else:
                # Fetch all attachments with pagination
                attachments = self.get_attachments_from_content(page_id=page_id, start=start, limit=limit)["results"]
                if not attachments:
                    return "No attachments found on the page."

            # Prepare to handle downloads
            downloaded_files = {}
            for attachment in attachments:
                file_name = attachment["title"] or attachment["id"]  # Use attachment ID if title is unavailable
                download_link = attachment["_links"]["download"]
                download_link = f"/rest/api/content/{page_id}/child/attachment/{attachment['id']}/download"
                # Fetch the file content
                response = self.get(str(download_link), not_json_response=True)

                if to_memory:
                    # Store in BytesIO object
                    file_obj = io.BytesIO(response)
                    downloaded_files[file_name] = file_obj
                else:
                    # Sanitize filename if needed
                    if re.search(r'[<>:"/\\|?*\x00-\x1F]', file_name):
                        sanitized = re.sub(r'[<>:"/\\|?*\x00-\x1F]', "_", file_name)
                        warnings.warn(
                            f"File name '{file_name}' contained invalid characters and was renamed to '{sanitized}'.",
                            UserWarning,
                        )
                        file_name = sanitized
                    file_path = os.path.join(path, file_name)
                    # Save file to disk
                    file_path = os.path.join(path, file_name)
                    with open(file_path, "wb") as file:
                        file.write(response)

            # Return results based on storage mode
            if to_memory:
                return downloaded_files
            else:
                return {"attachments_downloaded": len(attachments), "path": path}
        except NotADirectoryError:
            raise FileNotFoundError(f"The directory '{path}' does not exist.")
        except PermissionError:
            raise PermissionError(f"Permission denied when trying to save files to '{path}'.")
        except requests.HTTPError as http_err:
            raise requests.HTTPError(
                f"HTTP error occurred while downloading attachments: {http_err}",
                response=http_err.response,
                request=http_err.request,
            )
        except Exception as err:
            raise Exception(f"An unexpected error occurred: {err}")