Freshrss: [greader-api] feed ids

Created on 13 Jul 2020 · 9Comments · Source: FreshRSS/FreshRSS

based on a change in the past the api has weird inconsistencies right now.

feed ids used to be only feed/1 but got converted to google-api-standard strings: #3049

The API-endpoints below still use the old format:

item/idsgreader.php#L683
stream/contents greader.php#L627

it still only accepts something like this feed/1, but it should be instead accepting feed/http://feeds.feedburner.com/alistapart/main

currently i am also working with the items/contents and i am unable to get any other response then 400: Bad Request!, but this can be a problem on my side.

API

Source

Stunkymonkey

All 9 comments

@Stunkymonkey Could you please post cURL examples, so that I can better understand the problem(s)?

The quickadd API still returns something like "streamId":"feed/238"

curl -s -H "Authorization:GoogleLogin auth=alice/..." -X POST 'https://freshrss.example.net/api/greader.php/reader/api/0/subscription/quickadd?quickadd=http://feeds.arstechnica.com/arstechnica/science'
{"numResults":1,"query":"http://feeds.arstechnica.com/arstechnica/science","streamId":"feed/238","streamName":"Science – Ars Technica"}

But yes, I can try to add the ability to address feeds by their (unstable) URL if you like.

Alkarex on 14 Jul 2020

For me its fine to stick with feed/<id>, but I am not sure if this could have some incompatibilities with other tools.

I just wanted to be sure, that the api-wrapper I write for the news-flash are done correctly.

Stunkymonkey on 14 Jul 2020

Ok @Stunkymonkey
So far, I have not spotted any compatibility issue with the apps we tried:
https://github.com/FreshRSS/FreshRSS/blob/master/README.md#apis--native-apps
So is it good after all?

Alkarex on 14 Jul 2020

then its fine for me.

Maybe you could help me in providing a request for items/contents. Currently all my attempts ended in a "400: Bad Request!"

Stunkymonkey on 14 Jul 2020

Example:

curl -s -H "Authorization:GoogleLogin auth=alice/..." -d 'i=1594735486442505&i=1594553684923713' \
'https://freshrss.example.net/api/greader.php/reader/api/0/stream/items/contents'

{
    "id": "user/-/state/com.google/reading-list",
    "updated": 1594735739,
    "items": [{
            "id": "tag:google.com,2005:reader/item/0005aa674a2e1c09",
            "crawlTimeMsec": "1594735486442",
            "timestampUsec": "1594735486442505",
            "published": 1594734660,
            "title": "AMD Ryzen 9 3900XT Memory Scaling Performance Under 100 Different Tests",
            "summary": {
                "content": "For those thinking of picking up one of the new AMD Ryzen 3000XT series processors and weighing whether it's worthwhile on your budget picking up DDR4-3600 memory or other higher frequency DDR4 modules, here are some fresh benchmark results with the Ryzen 9 3900XT looking at 100 different tests on Linux and showing how the performance changes from DDR4-2133 through DDR4-3800..."
            },
            "alternate": [{
                    "href": "http://www.phoronix.com/scan.php?page=news_item&px=Ryzen-9-3900XT-DDR4-Memory"
                }
            ],
            "categories": ["user/-/state/com.google/reading-list", "user/-/label/Informatique Int."],
            "origin": {
                "streamId": "feed/145",
                "title": "Phoronix"
            }
        }, {
            "id": "tag:google.com,2005:reader/item/0005aa3cf5f71541",
            "crawlTimeMsec": "1594553684923",
            "timestampUsec": "1594553684923713",
            "published": 1594551840,
            "title": "Fedora Approves Of Making Nano The Default Terminal Text Editor, Other Features Accepted",
            "summary": {
                "content": "At this week's Fedora Engineering and Steering Committee (FESCo) meeting, more features were approved for the Fedora 33 release due out this fall..."
            },
            "alternate": [{
                    "href": "http://www.phoronix.com/scan.php?page=news_item&px=Fedora-33-Nano-Is-The-Default"
                }
            ],
            "categories": ["user/-/state/com.google/reading-list", "user/-/label/Informatique Int."],
            "origin": {
                "streamId": "feed/145",
                "title": "Phoronix"
            }
        }
    ]
}

However, I do not think this is an efficient call for typical synchronisation. Please check first my suggestion:
https://github.com/FreshRSS/FreshRSS/issues/2566#issuecomment-541317776

Alkarex on 14 Jul 2020

❤1

yes thanks. First I want to implement all methods, and then use the ones which fit perfectly.

thanks for your help

Stunkymonkey on 14 Jul 2020

👍1

Ok; you probably only need a little subset of the API for your purpose, though :-)
I think it helps to take existing apps, and check their calls.

One can produce full logs in FreshRSS by uncommenting those two lines:
https://github.com/FreshRSS/FreshRSS/blob/master/p/api/greader.php#L929-L930

Here is for instance a trace for News+, which I use as a reference because it has an efficient synchronisation approach:

Alkarex on 14 Jul 2020

hi @Alkarex,
i started working on the freshrss-newsflash implementation.
Due to my exams, I am not sure how much I can work on it.

I think what you want to tell me with your messages about the syncing process:
use the filter as restrictive as possible (specially restrictive timestamps) and only sync the wanted/updated items.
did i understand this right?

Stunkymonkey on 13 Aug 2020

@Stunkymonkey My main advise was to save time by focussing on the bits of the API you really need, and also take inspiration from the precise calls from other apps.
When it comes to syncing the state of articles changed by the client, using IDs is fine. But in order to massively retrieve new articles (or to get changed articles status from the server), using a method returning all articles matching a filter is more efficient.

Alkarex on 17 Aug 2020

❤1

Was this page helpful?

0 / 5 - 0 ratings