How Does Instagram's GET/tags/<tag>/media/recent Pagination Actually Work?

Question

I'm trying to use the Real-time Photo Updates API to get all pictures with a specific tag as they come in. Since updates from this API really only tell you that new content exists (but not what it is), I am querying for recent media with my tag whenever I get the notification that something has changed.

The problem I am having is that I'm constantly getting duplicate media returned by that query. The documentation says this API does pagination, but I can't get it to work at all.

The documentation here says to use min_id and max_id in your query string to control pagination, however it says:

• MIN_ID - Return media before this min_id.

• MAX_ID - Return media after this max_id.

This seems backwards (normally min should return items above the minimum and max should return items below the maximum so that when you specify both you get a bounded range).

The actual JSON I get back from my queries includes a pagination object like this:

"pagination":  {
    "next_max_tag_id": "1387272337517",
    "deprecation_warning": "next_max_id and min_id are deprecated for this endpoint; use min_tag_id and max_tag_id instead",
    "next_max_id": "1387272337517",
    "next_min_id": "1387272345517",
    "min_tag_id": "1387272345517",
    "next_url": "https://api.instagram.com/v1/tags/cats/media/recent?access_token=xxx&max_tag_id=1387272337517"
}

The parameter specified in the next_url property is max_tag_id, not max_id like the documentation says.

There is also a deprecation warning that states that next_max_id and min_id are deprecated, but since there are properties in the pagination object with names like that, I don't know whether the query parameters or the object properties the ones that are deprecated..

I would think it means the properties, because the query string never used a parameter of next_max_id, but then the deprecation message says to use min_tag_id and max_tag_id and there is no max_tag_id property on the pagination object (just a next_max_tag_id).

Regardless of all of this conflicting documentation, it doesn't seem to matter what I pass in my query string - I continue to get repeat media in subsequent queries. Can someone help me make sense of this API? All I really want is to get tagged media that is new since my last query.

Answer 1

To get the newest set of grams for a particular tag, use this:

https://api.instagram.com/v1/tags/latergram/media/recent?access_token=TOKEN

From that response, you can get newer grams from the same tag by taking the min_tag_id from the response (under pagination) and build a url like so:

https://api.instagram.com/v1/tags/latergram/media/recent?access_token=TOKEN&min_tag_id=1387332980547

Or you can get the next (older) set of grams by using the next_url parameter from the original response (also under pagination), which looks like:

https://api.instagram.com/v1/tags/latergram/media/recent?access_token=TOKEN&max_tag_id=1387332905573

Make sure your subsequent queries (for new grams of a particular tag) are using the min_tag_id returned by the latest response. I did a few tests and didn't see duplicates, however I was using #latergram and that one has a high volume of posts

Answer 2

@zachallia has answered spot on, but I figure it can't hurt with a sketch:

As the Instagram API says: MIN_TAG_ID Return media before this min_tag_id. MAX_TAG_ID Return media after this max_tag_id.

This is counterintuitive, with a slightly nutty flavor. But still, it is possible to make sense of it.

The /tags/MYTAG/media/recent endpoint will give you grams, ordered by how newly they where tagged with MYTAG. You'll not get all grams, of course, just up to the limit set by Instagram:

|yesteryear ------------------ <---- LIMIT ----> now|

If you use min_tag_id like so /tags/MYTAG/media/recent?min_tag_id=X you'll get grams from X and before (aka older):

|yesteryear ------- <---- LIMIT ---> min ------- now|

If you use max_tag_id like so /tags/MYTAG/media/recent?max_tag_id=Y you'll get grams from Y and after (aka newer):

|yesteryear ------- max <---- LIMIT ---> ------- now|

That's how "max" gets to signify "newer" and "min" gets to signify "older".