Organizing Instagram tagged posts

[Fukitsu]

I'm trying to make gallery-dl download a tagged post inside the folder of the tagged user when scraping a profile. For example, when scraping this profile https://www.instagram.com/akunohako/ there's this post https://www.instagram.com/p/Cjs36PIjFWP/ which was made by a different user and I want to download it inside the akunohako folder instead of the xenon_ne folder, same for the filename as if the post were created by akunohako. I've tried with

"tagged": 
            {
                "directory": ["{tagged_username}"]
            },

"tagged": 
            {
                "directory": ["{tagged_users[][username]}"]
            },

and

 "postprocessors":
            [
                {
                    "name": "metadata",
                    "tagged":
                    {
                        "directory": ["{tagged_username}"]
                    }
                }
            ]

but can't get it to work. What's the proper way?

[afterdelight]

this is my tagged config:

"tagged": {
        "directory": ["{tagged_username} [{tagged_owner_id}]", "Tagged"],
        "filename": {
        "": "{date:%Y%m%d}_{username}_{post_shortcode}.{extension}",
        "count > 1": "{date:%Y%m%d}_{username}_{post_shortcode}_{num}.{extension}"
    }
}

[mikf]

{tagged_username}, {tagged_owner_id}, {tagged_full_name} and the "tagged" gdl subcategory are only applicable when downloading from instagram.com/USER/tagged URLs. Just plain instagram.com/USER uses the "posts" subcategory by default.

In your case you'd have to use {tagged_users[0][username]}, i.e. the first item from the list of @-ed users in a post, but that does not always work since there can be multiple tagged users and the one you want not being the first.

What you'd really need is a new metadata field for "coauthors" which gallery-dl does not yet extract, although that would also be a list, or something similar to the user and author distinction that Twitter has.

      "coauthor_producers": [
        {
          "pk": "212047764",
          "username": "akunohako",
          "full_name": "Aku",
          "is_private": false,
          "profile_pic_url": "https://instagram.ftxl3-2.fna.fbcdn.net/v/t51.2885-19/274583097_482362060206481_1212815530004034262_n.jpg?stp=dst-jpg_s150x150&_nc_ht=instagram.ftxl3-2.fna.fbcdn.net&_nc_cat=101&_nc_ohc=U4iWwyTFtiAAX94e6fb&edm=ALQROFkBAAAA&ccb=7-5&oh=00_AT-PpOkf42f6R3iP_gbX_8eWDN5pDE5qrtVz169XqWOMFg&oe=635E1CCD&_nc_sid=30a2ef",
          "profile_pic_id": "2780925354380274334_212047764",
          "is_verified": false
        }
      ],

(unfiltered API data from /p/Cjs36PIjFWP/)

[Fukitsu]

Some workaround I've been using is using a wrapper script inside my Instagram folder with -o base-directory=. and the filenames like this:

"filename":
            {
                "tagged_users and count > 1": "USER {date:%Y-%m-%d - %H_%M_%S} {post_shortcode}_{num}.{extension}",
                "tagged_users": "USER {date:%Y-%m-%d - %H_%M_%S} {post_shortcode}.{extension}",
                "count > 1": "{username} {date:%Y-%m-%d - %H_%M_%S} {post_shortcode}_{num}.{extension}",
                "": "{username} {date:%Y-%m-%d - %H_%M_%S} {post_shortcode}.{extension}"
            },

directory like this:

"directory":
            {
                "": ["."]
            },

and then rename the files replacing USER with the username and deleting them if there are duplicates. Does {coauthors[0][username]} always return the username I'm scraping or is it like {tagged_users[0][username]}?

[mikf]

{coauthors[0][username]} has probably the same problem as {tagged_users[0][username]}, given that it is also a list and therefore can have multiple values.

As I said in https://github.com/mikf/gallery-dl/issues/3107#issuecomment-1291723260, I'll eventually redo the user(name) fields for Instagram and handle them like it is currently done for Twitter, i.e. user[…] is the user the input URL points to and author[…](or some other name) is the actual post author/creator.

[duoside]

Yes, this feature would make scraping much easier. Currently, I'm manually moving the tagged files to the correct folder

[Fovty]

In case anyone else stumbles across the same problem. I solved it as follows (using akunohako as an example):

"directory": {
                "": ["{subcategory}"]
}

gallery-dl https://www.instagram.com/akunohako/ -d "./downloads/akunohako"

Since I run gallery-dl as a subprocess anyway, I can set the destination parameter dynamically (matching the profile name)