/universe/post_universe_ids returning unrelated Corporations when submitting Character names to resolve

[DaneelTrevize]

Bug

Request

curl -X POST "https://esi.tech.ccp.is/latest/universe/ids/" -H "accept: application/json" -H "Content-Type: application/json" -d "[ \"afk x\"]"

Response

{ "corporations": [ { "id": 98364532, "name": "HaHa Hab dich" } ] }

Expected

{ "characters": [ { "id": 95934406, "name": "afk x" } ] }

They exist, as per https://esi.tech.ccp.is/latest/search/?categories=character&search=afk%20x&strict=true, https://zkillboard.com/character/95934406/ and Jita local in-game very recently.

[a-tal]

that's not an unrelated corporation, that corporation's ticker is 'afk x'. it should be returning both categories though...

we should maybe remove this endpoint anyway though. @ccp-zoetrope thoughts?

@DaneelTrevize for an immediate workaround, use the /search/ endpoint instead:

curl -X GET "https://esi.tech.ccp.is/latest/search/?categories=character,corporation&search=afk%20x&strict=true" -H "accept: application/json"

[DaneelTrevize]

Gah, this endpoint was added to finally resolve issue #27 which is most needed for the bulk resolving of Local Scan parsers.
Thanks for the quick response though.

[ccp-zoetrope]

Seems this is cache based again because it works correctly the first time you do it but if it pulls from the cache it will only return "HaHa Hab dich".

@a-tal why should we remove this endpoint?

[DaneelTrevize]

Related caching issues: #811

[Shegox]

got the same problem, I found out that this happens with 5 char long character (["Latch", "Hungo", "Dalic"]) names in my case.
The following is what I saw:
Request:

curl -i --compressed -X POST --header 'Content-Type: application/json' --header 'Accept: application/json' -d '["Latch", "Hungo", "Dalic"]' 'https://esi.tech.ccp.is/latest/universe/ids/?datasource=tranquility&language=en-us'

Response:

{
  "corporations": [
    {
      "id": 98022707,
      "name": "Latchkey Kids"
    },
    {
      "id": 98311751,
      "name": "Im Hungover"
    },
    {
      "id": 98546273,
      "name": "Dallocort Corporation"
    }
  ]
}

Expected 3 Character Ids, is there a patch planned or some workaround?

Edit: I'm doing a single search request for every name now, this gives the expected results, but takes quite longer.

[iay]

Observation: those five-character strings are the corporation tickers for the corporations in question.

[Shegox]

that makes sense @iay , would be cool in that case to allow a flag to filter for character, or return both values under the characters & corperations names.

[DaneelTrevize]

Read up, the first reply was CCP confirming it's tickers-related and that "it should be returning both categories though".

[ccp-zoetrope]

@Shegox the real problem here is that both those corporations and the characters should be showing up, but due to a caching problem they are not. A fix is in the works but fanfest craziness has pushed this back a bit.

[Shegox]

Any update on this issues? I just tested and got on the first try both (chars and corp) returned, afterwards only the corps.
As I'm looking to switch from the /search endpoint for mass characterId lookup and reduce the request amount.
@ccp-zoetrope @a-tal

[jowrjowr]

running into an instance of this on my side

character: bgtop (92482138)
corp: bgtop (98054246)

what's interesting is the corp isn't actually named bgtop but that is rather their ticker

this is what ESI returns:

$ curl -d '["BGTOP"]' -X POST "https://esi.evetech.net/latest/universe/ids/"
{"corporations":[{"id":98054246,"name":"The Circus Corp"}]}

returning everything in the case of degeneracies would be the ideal solution here

the mitigation right now is the search endpoint but that doesn't accept bulk requests

[WhiteRusssian]

@ccp-zoetrope , any idea when we should expect this issue to be fixed? Is this a matter of weeks or months?

[Aidansavage]

Soon(tm)

[ccp-zoetrope]

@WhiteRusssian a fix for this is slated for release so this will be coming out soon

[a-tal]

this fix was pushed to prod