pgoapi
pgoapi copied to clipboard
tejado
Botting
I cloned a while ago from @tejado and have been working on implementing features... Got to say things are going well and I'm willing to answer questions people have, but not willing to copy/paste source code since it would lead to more questions and at this point I'm divergent from the current code base by several revisions (some of them structural).
I do not claim to have the best or most advanced bot, but it does what I want it to do... it just leveled a brand new test account from lvl 1 -> 20 in under a day (servers were stable).
Here is what I have figured out thus far:
- looping a specified set of waypoints OR a random-walk based on seeded start location
- use all pokestops we encounter
- catching all pokemon that are map visible as we walk ** catching uses curve balls and probably a "too" high chance of excellent throws ** catching uses appropriate level of pokeball based on difficulty of pokemon ** catching uses razz berries as necessary based of difficulty of pokemon
- inventory knowledge - I know how much I have of each item (except evolution candies for now....)
- pokemon inventory control (aka we transfer pokemons to the professor for candy when overly full, config options to keep what you want by type/nickname/cp-level)
- ability to scan our Pokedex to know what pokemons we have seen or not seen yet
- bag item control (delete items we don't care about if we run out of space... also cap our pokeballs so we don't keep too many)
- egg hatching detection
- automatic egg incubation
- detection of bad network conditions or session-key changes by server with auto-reconnect after timeout
- code for using Lucky Eggs (currently not automated as when to do it is debatable .. )
- code for automatic evolution of pokemons (although untested still)
Feel free to ask questions, but like I said, I won't be releasing code or copy/pasting solutions, but can help with proto formats.
detection of bad network conditions or session-key changes by server with auto-reconnect after timeout -> miss Tell me more about this, i don't know too much about sessions
I believe your evolution candies are listed in your inventory as well in the dictionary of the pokemon.
{
"inventory_item_data": {
"pokemon_family": {
"family_id": 124,
"candy": 20
}
},
"modified_timestamp_ms": 1469131340176
}
Could you help me out how to trigger egg incubation ? In the inventory I can find if I have a available incubator and its id, how can to get my eggs id How do I trigger it ?
@dnsBlah they are listed in dictionary of pokemons but they all are 0 apparently when you walk it... thanks for the pokemon_family data... that will work
Egg Incubation - the proto is slightly wrong as checked (at least time last time I checked). Notice I have "repeated" whereas what's checked in is "optional". @tejado doesn't provide the .proto files here (not sure why) so it won't be an easy fix for you guys - but that fix is only necessary if you want to incubate multiple eggs at the same time.
message EggIncubators {
repeated EggIncubator egg_incubator = 1;
}
Below should explain how to use incubators (the code is using my structure, but should give you an idea of how to use it).
def useIncubator(api):
egg_ids = []
inc_ids = []
inv = getInventory(api)
if inv:
for items in inv.inventory_delta.inventory_items:
pk = items.inventory_item_data.pokemon
if pk and pk is not None:
if pk.is_egg and not pk.egg_km_walked_start:
egg_ids.append(pk.id)
ics = items.inventory_item_data.egg_incubators
for ic in ics.egg_incubator: # <-- requires the fixed proto I mentioned, othewise can't loop
if ic and ic is not None:
if ic.item_id:
inc_ids.append(ic.item_id)
log.debug('Available Eggs for Incubation: %s', egg_ids)
log.debug('Available Incubators: %s', inc_ids)
for i in range(0, min(len(egg_ids),len(inc_ids))):
api.use_item_egg_incubator(item_id=inc_ids[i], pokemon_id=egg_ids[i])
inc_use = api.call(False)
log.info('Using Egg Incubator "%s" on "%d"...' % (inc_ids[i], egg_ids[i]))
inc_pay = getPayload(inc_use)
if inc_pay:
inc_resp = rsub.UseItemEggIncubatorResponse()
inc_resp.ParseFromString(inc_pay)
print(inc_resp)
else:
log.error('Bad Payload - Using Egg Incubator')
@ztukaz - regarding sessions, honestly I'm not super familiar with that. @tejado might honestly know better as this would be a API library / networking / connectivity issue. What I know is from here: https://github.com/Mila432/Pokemon_Go_API/blob/master/api.pyy
@Nostrademous Thanks for the hints I don't think the statements are right, as a pokemon['is_egg'] == true does not tell me if its active or not. Instead the incubator tells me with start_ and target_km_walked But can't link a incubator to an egg :-/
Or is that the reason why I need a other proto to request the inventory? or just try all 9 eggs until positive result?
@dnsBlah pokemon['is_egg'] tells you it's an egg pokemon['egg_km_walked_start'] is only present if it is an active egg (and can only be present if it's an egg) <-- this is why I check that it is NOT present to know it's an available egg
With both those checks you can then record pokemon['id'] as a valid egg for incubation. That is the 'id' you pass as the 2nd argument to api.use_item_egg_incubator()
Also, looks liek POGO fixed the "repeated" 20 hrs ago.... heh, didnt' realize. https://github.com/AeonLucid/POGOProtos/blob/master/src/POGOProtos/Inventory/EggIncubators.proto
Hi @Nostrademous Like I mentioned, the egg does not show anything about start...
{
"inventory_item_data": {
"pokemon": {
"creation_time_ms": 1469191037546,
"captured_cell_id": 5171996602586365952,
"is_egg": true,
"id": 11891607393952159418
}
},
"modified_timestamp_ms": 1469191037546
}
(I have 9 eggs *max ofcourse) All 9 look like this, but I'm walking 1.
{
"inventory_item_data": {
"egg_incubators": {
"egg_incubator": {
"target_km_walked": 153.98680114746094,
"start_km_walked": 148.98680114746094,
"pokemon_id": -4500287810622364247,
"item_type": 901,
"incubator_type": 1,
"item_id": "EggIncubatorProto-1928901772253991081"
}
}
},
"modified_timestamp_ms": 1469219038432
},
See what I mean :) I understand the call thanks of your hint above. Seems like almost all calls are just 2 parameters?
But, yet, I don't see a link between the incubator and the egg's I tried partials of the incibator item_id, of the pokemon_id None relates to any of the eggs in my inventory
@dnsBlah I just verified and I guess it changed, you are right, 'km' stuff no longer listed for active eggs. I guess you loop over the eggs and try each one.
If you try to incubate an egg already in use it will return a RESPONSE with: ERROR_POKEMON_ALREADY_INCUBATING = 6;
If you get that, just try a different egg: Results are:
enum Result {
UNSET = 0;
SUCCESS = 1;
ERROR_INCUBATOR_NOT_FOUND = 2;
ERROR_POKEMON_EGG_NOT_FOUND = 3;
ERROR_POKEMON_ID_NOT_EGG = 4;
ERROR_INCUBATOR_ALREADY_IN_USE = 5;
ERROR_POKEMON_ALREADY_INCUBATING = 6;
ERROR_INCUBATOR_NO_USES_REMAINING = 7;
}
It also seems like that the incubators are not shown in the inventory list... Just 1 (the unlimited one), and I have 3
@dnsBlah are you using ITEM_ID=902 for the "extra incubators"? Will have to check if they show up or not, I don't really care actually as when I use incubators I enumerate them differently.
I'm guessing i have it working now.. However I don't get unique id's from my item: 902 But will see tomorrow..
@dnsBlah
def useIncubator(api):
egg_ids = []
inc_ids = []
inv = getInventory(api)
if inv:
for items in inv.inventory_delta.inventory_items:
pk = items.inventory_item_data.pokemon
if pk and pk is not None:
if pk.is_egg:
egg_ids.append(pk.id)
for items in inv.inventory_delta.inventory_items:
ics = items.inventory_item_data.egg_incubators
for ic in ics.egg_incubator:
if ic and ic is not None:
if ic.item_id: # and not ic.target_km_walked:
inc_ids.append(ic.item_id)
log.debug('All Eggs: %s', egg_ids)
log.debug('All Incubators: %s', inc_ids)
for inc in inc_ids:
for egg in egg_ids:
api.use_item_egg_incubator(item_id=inc, pokemon_id=egg)
inc_use = api.call(False)
inc_pay = getPayload(inc_use)
if inc_pay:
inc_resp = rsub.UseItemEggIncubatorResponse()
inc_resp.ParseFromString(inc_pay)
if inc_resp.result == 1:
log.info('Using Egg Incubator "%s" on "%d"...' % (inc, egg))
log.info(inc_resp)
elif inc_resp.result == 6:
continue
elif inc_resp.result == 5:
log.info('Incubator "%s" already in use' % inc)
break
else:
log.error('Bad Payload - Using Egg Incubator')
@Nostrademous
@tejado doesn't provide the .proto files here (not sure why) so it won't be an easy fix for you guys - but that fix is only necessary if you want to incubate multiple eggs at the same time.
Did you imported in this project manually?
@Nostrademous With an other proto class its working fine now :) Also finding other incubators :-)
@dnsBlah glad you got it working :)
I just recently added mass evolving capability to my bot with the use of a Lucky Egg (if I can do more than 125 evolutions at once). This required tracking proper candy and candy requirements (which I do now).
Also added level-up detection and level-up-prize gathering.
Leveled a brand new account to level 22 in under 12 hours - played by bot from level 1 on (without the use of any evolution or lucky-egg).
@Nostrademous How did you approach gym ? It should be something like, StartGymBattle, AttackGym and then Put pokemon right?
With StartGymBattle I get always status = 3 and empty results
gym_id = id of the gym (fort)
attacking_pokemon_ids = [idpok1,idpok2]
defending_pokemon_id=id_first_defender
player_latitude=lat
player_longitude=lng
{'responses': {'START_GYM_BATTLE': {}}, 'status_code': 3, 'auth_ticket': {'expire_timestamp_ms': 1469664544675L, 'start': 'tbXTF4N9/wntFc4f0PV6KlBm7kfwsflM2jW/PTtO2tS8jTG9YYAlBBcFTag9PbRo1TFX9NwYDOAiAmuIEpg6bygUFhoHklL52T4i2OWkOgU=', 'end': 'n5kcn21Ojw+YUJZDUBPwwg=='}, 'request_id': 8145806132888207460L}
@Nostrademous Please share the level-up-prize gathering. :-D However I'm sure that the items are gathered anyways when you reach the next level tho.. Still I think it would be good to simulate as much as possible. Because lots of times I end up with more items than my bag can handle ;-)
@ztukaz tackling gyms next - haven't started yet
@dnsBlah - for leveling use the LEVEL_UP_REWARDS call and you must supply your current level (which you can obtain from Inventory ItemData PlayerStats. The response message will have all the info you need, and unlocked new items which you can now start getting at pokestops.
@ztukaz isn't status_code 3 stating that it is a NEUTRAL GYM? Meaning you can't attack it as no one owns it?
Hi @Nostrademous
That's so cool! I now get AWARED_ALREADY Guess I will have to monitor before I level up, to see what the response might be :)
Any idea when this is called in the real game? In the heartbeat ? Btw you also know how the heartbeat actually looks like? :D
Thanks!
@ztukaz @tejado Working on gyms... but apparently I keep failing my request... no idea why: any thoughts?
def battleGym(fort):
log.info("<>________________________________________________________________________<>")
log.info("%s GYM NEARBY", TeamColor.DESCRIPTOR.values[fort.owned_by_team].name)
# get gym details
gym_proto = getGymDetails(fort)
if not gym_proto: return 0
log.info(gym_proto)
log.info("<>________________________________________________________________________<>")
##############################################
# gather necessary details to start a battle #
##############################################
# check if gym is neutral/unowned - if so deploy to it
if fort.owned_by_team == 0: # Neutral
deployPokemonToGym(gym_proto)
return 0
# save a boolean checking if we are on the same team
sameTeam = gl.my_team == fort.owned_by_team
# find out who is defending
defending_pokemon = []
for defender in gym_proto.gym_state.memberships:
defending_pokemon.append(defender.pokemon_data)
# determine who we will attack with
num = 6
if sameTeam: num = 1
my_attackers = pickAttackers(num) # this will be a list []
#log.error('Function not finished, returning out...')
#return 0
# start the battle
log.info('Gym ID: %s' % gym_proto.gym_state.fort_data.id)
log.info('Attacker list: %s' % str(my_attackers))
log.info('Defender: %d' % defending_pokemon[0].id)
gl.api.start_gym_battle(gym_id=gym_proto.gym_state.fort_data.id, attacking_pokemon_ids=my_attackers, defending_pokemon_id=defending_pokemon[0].id, player_latitude=gl.p_lat, player_longitude=gl.p_lng)
start_reply = gl.api.call(False)
start_payload = getPayload(start_reply)
Printing the request I see:
[23:45:59] [ INFO] Gym ID: 555ba115a93546aca7851fed86d3ef6c.16
[23:45:59] [ INFO] Attacker list: [17xxxxxx list of IDs xxxx 2L]
[23:45:59] [ INFO] Defender: 568051736162843098
[23:46:00] [ERROR] startGymBattle() request failed to get payload
and printing the gym_proto result:
[23:45:59] [ INFO] gym_state {
fort_data {
id: "555ba115a93546aca7851fed86d3ef6c.16"
last_modified_timestamp_ms: 1469762856688
latitude: 38.854506
longitude: -77.349488
owned_by_team: RED
guard_pokemon_id: FLAREON
enabled: true
gym_points: 2000
}
memberships {
pokemon_data {
id: 568051736162843098
pokemon_id: FLAREON
cp: 1002
stamina: 130
stamina_max: 130
move_1: EMBER_FAST
move_2: FLAMETHROWER
owner_name: "Firehawk04xx"
height_m: 0.884167253971
weight_kg: 9.47041511536
individual_attack: 15
individual_defense: 14
individual_stamina: 1
cp_multiplier: 0.481684952974
pokeball: ITEM_GREAT_BALL
battles_attacked: 10
battles_defended: 1
creation_time_ms: 1469753995514
num_upgrades: 2
additional_cp_multiplier: 0.0181734859943
nickname: "Flareon"
}
trainer_public_profile {
name: "Firehawk04xx"
level: 13
avatar {
skin: 2
hair: 2
shirt: 1
hat: 2
shoes: 1
backpack: 2
}
}
}
}
@Nostrademous I was waiting for you for the same problem, but still didn't get out
@tejado @ztukaz so I get it to work sometimes. The fix is modifying the StartGymBattleMessage() protobuf. the defender_id needs to be fixed64.
However it fails at time when the defender_id happens to be larger than int64 but still smaller than uint64. Python by default treats everything as signed.. but I checked your rpc_api and you use "setattr()" in the _build_sub_requests() function which is pass by reference.. so I am not sure what the issue is. However, there is a pattern for me. For IDs larger than int64 but still 8-bytes I get no reponse payload from server. This also happens for me in Incense-based Pokemon encounters.
I'm just getting responses like: Session.py - line 407 - startGymBattle()
status_code: 3
request_id: 951201794807
returns: ""
Which are not good I guess status 3? Or I will need the request_id to perform AttackGymMessage <-> Response for the actual battle I guess Then still I think status 3 is not good
According to the responses it should be, ERROR_GYM_NEUTRAL ? The gym is not neutral! :-/
@Nostradamous where exactly do you change what in the request message ?
@dnsBlah if you go into the /pgoapi/protos/POGOProtos_src/Networking/Requests/Messages/StartGymBattleMessage.proto
you need to change the defending_id from uint64 to fixed64
then you need to recompile the proto using "protoc" which you need to have installed of version 3.0.0b4
Ahhh I need to do that before compiling it!? Thanks will have a look :)
awesome ;-) but its not always working, however using the same defender, so that part has nothing to do with the fixed64, it's fine I guess I am experiencing some server issues throughout the whole day already tho.
Guess its a matter of a loop now and send back the actions battle_log.battle_actions to the BattleAction proto, together with the battle_id and some BattleAction's... Not sure how to create a BattleAction fo... send an fast-attack for special
@Nostrademous can you show some examples of running / reading forts from map_objects (i keep getting a weird error, at work now so can't supply it at this time)?
You need bigger delay's... You can only call map_objects per 5 seconds. Besides that there are hardly any pokemon shown anymore. Only within a range of 50-70 meters. Dispite the fact your scanning a bigger area
@joshk6656 did @dnsBlah comments help you? It's hard to tell what your errors were. If you are not getting good fort data from map_objects it's probably b/c you are soft-banned for teleporting too much. Otherwise it is fairly straight forward.