Skip to content

Latest commit

 

History

History
466 lines (380 loc) · 15.3 KB

README.md

File metadata and controls

466 lines (380 loc) · 15.3 KB
Raw

SWUbanner

AIO Wialon

license pypi version

AIO Wialon is an async implementation of Python wrapper for Wialon Remote API,

Table of

Installation

pip install py-aiowialon

Start Polling

Open session and start poll AVL events immediately. Look the Wialon Events section to see how we can handle AVL Events on polling

import asyncio
from aiowialon import Wialon

TOKEN = '<Your Wialon API access token>'
HOST = '<API host IP or url>'
wialon = Wialon(host=HOST, token=TOKEN)

if __name__ == "__main__":
    asyncio.run(wialon.start_polling())

Tip

Wialon.start_polling() is not require a manual Wialon.login() call

Wialon API Call

API Call is function that returns Wialon.call() instance Almost all Wialon Remote API services/actions available through dot syntax: wialon.<service>_<action_name>(**params)

To make API call use method of Wialon instance with same name as API endpoint replace / with underscore.

API Call Example

import asyncio
from aiowialon import Wialon, flags

TOKEN = '<Your Wialon API access token>'
wialon = Wialon(token=TOKEN)

async def main():
    await wialon.login()
    # The example of core/search_item API call:
    result = await wialon.core_search_item(id=12345, flags=flags.UnitsDataFlag.ALL)
    print(result)
    await wialon.logout()

asyncio.run(main())

Warning

Some Wialon Remote API methods requires a lock of asynchronous context (execution of reports, loading messages, etc). If you need these methods, it's highly recommended to get acquainted with Critical requests execution section

Batch requests

Use Wialon.batch instead of asyncio.gather to make multiple API calls in a same time. It allows to make just one request to server with few calls. This avoids reaching the server's request limits. And transfers the overhead of processing asynchronous context to the server side. Few Wialon.call() coroutines would be batched to single 'core/batch' request.

# put few calls to a batch method
from aiowialon import Wialon, flags

wialon = Wialon(token=TOKEN)

async def some_func(params1, params2):
    api_calls = [
        wialon.core_search_item(**params1),
        wialon.unit_get_fuel_settings(**params2),
        ...
    ]
    return await wialon.batch(*api_calls, flags_=flags.BatchFlag.EXECUTE_ALL)

Tip

Warning

  • Some requests don't support batch!
  • Don't try to put batch into other batch, it can raise unexpected behaviour
  • Go to the Wialon Remote Api documentation to get details

Multipart requests

Use Wialon.multipart method and MultipartField with API call to but multipart data to request, Put call coroutine and required MultipartField instances to the Wialon.multipart()

from aiowialon import Wialon, MultipartField

wialon = Wialon(token=TOKEN)

async def upload_driver_image():
    event_hash = 'aiowialon_drv_upd'  # custom event hash
    params = {"itemId": 717351, "driverId": 38, "eventHash": event_hash}
    file_path = "driver_img.jpg"
    with open(file_path, 'rb') as f:
        file_data = f.read()

    await wialon.multipart(
        wialon.resource_upload_driver_image(**params),
        *[
            MultipartField(
                name='drivers_dlg_props_upload_image',
                value=file_data,
                filename="image.jpg",
                content_type='image/jpeg'
            )
        ]
    )

Warning

  • Don't try to put multipart requests to batch!
  • Some requests don't support multipart
  • Don't try to put multipart request into batch, it can raise unexpected behaviour
  • Go to the Wialon Remote Api documentation to get details

Shortcuts

Shortcuts are available as efficient solutions for some common actions, like .wlp export

from aiowialon import Wialon
from aiowialon.shortcuts import WLP

wialon = Wialon(token=TOKEN)


async def dump_unit(item_id):
  await wialon.login()
  wlp = await WLP.export_item(wialon, item_id)
  with open(f"{id}.wlp", 'wb') as fp:
    fp.write(wlp)

Wialon Events

The library propose using the polling to handle AVL Events. AVL events is the events that happens on the server and returns to us if we registered it in current session This section references to Wialon AVL Events Docs

Register AVL Events

Firstly we have to register items for AVL events handling in current session. (api reference here)

Bellow is example how to add all AVL Units (vehicles) to handle AVL events of this units in current session, We use there just simple Wialon API Call

from aiowialon import Wialon, flags

wialon = Wialon(token=TOKEN)

async def register_avl_events():
    spec = [
        {
            "type_": "type",
            "data": "avl_unit",
            "flags": flags.UnitsDataFlag.BASE | flags.UnitsDataFlag.POS,
            "mode": 0
        }
    ]
    return await wialon.core_update_data_flags(spec=spec)

On login/logout

We can automate this logic for each session opening by registering on_session_open callback, Use @wialon.on_session_open decorator for this So wialon will login and register avl items to polling before polling start

@wialon.on_session_open
async def register_avl_events(session_login):
    print("Session eid:", session_login['eid'])
    spec = [
        {
            "type_": "type",
            "data": "avl_unit",
            "flags": flags.UnitsDataFlag.BASE | flags.UnitsDataFlag.POS,
            "mode": 0
        }
    ]
    return await wialon.core_update_data_flags(spec=spec)

if __name__ == "__main__":
    asyncio.run(wialon.start_polling())

Also we can add callback on session logout. Use @wialon.on_session_close decorator for this

@wialon.on_session_close
async def on_session_close(session_logout):
    print("Logout event:", session_logout)

Note

  • You can register just one on_session_open callback for Wialon instance
  • You can register just one on_session_close callback for Wialon instance

AVL Events Handling

After polling start and AVL Items registered for polling we can handle the AVL Events. Use @wialon.avl_event_handler() decorator

Register AVL Events handlers

from aiowialon import AvlEvent


@wialon.avl_event_handler()
async def unit_event(event: AvlEvent):
  print("Handler got event:", event)

Put the filter function to the decorator to apply filtering of AVL events

from aiowialon import AvlEvent


@wialon.avl_event_handler(lambda event: event.data.i == 734455)
async def unit_734455_event(event: AvlEvent):
  print("Handler got event from item 734455:", event)

Note

Register handlers in an order in which filters have to be applied. If some handler catched the event, next handler in order will never do.

Remove AVL Events handlers

# use them as you need
wialon.remove_avl_event_handler('handler_name')
wialon.remove_avl_event_handler(handler_func)

Disposable handlers

Use @wialon.avl_event_once to be certain that handler will be removed immediately after single execution

@wialon.avl_event_handler()
@wialon.avl_event_once
async def unit_event(event: AvlEvent):
    print("Handler got event:", event)

Exceptions Handling

The avl_event_handler suppress the callback's WialonError exceptions to protect app to be closed on unexpected behaviour So if u want to handle some specific WialonError, do it in handler's callback scope

Note

You still can get access to response data even if WialonError exception was raised, see next section

from aiowialon import WialonError, WialonAccessDenied


@wialon.avl_event_handler()
async def unit_event(event: AvlEvent):
  try:
    raise WialonAccessDenied  # example of wialon exception raised in callback scope
  except WialonError as err:
    # do something
    pass

Exceptions Handling (Batch)

You still can get access to response data even if WialonError exception was raised It can be usefull for debug or for the batch requests WialonError.reason returns string for single call or list[WialonError] for batch call

async def some_func():
    result = None 
    try:
        result = await wialon.batch(*calls, flags_=flags.BatchFlag.STOP_ON_ERROR)
    except WialonError as err:
        print("Errors", err.reason) # returns a list of WialonErrors for each call in batch
        result = err.result
    finally:
        print("Result", result)

Quick API Help

Use Wialon.help(service_name, action_name) to open Wialon Remote API docs in your system browser

from aiowialon import Wialon

Wialon.help('core', 'search_item')

Advanced usage

Limitations

Adjusting to the Wialon API limitations the Wialon API client limited to 10 connections maximum per session wia asyncio.semaphore Also it limited to 10 requests per second for the session with aiolimiter You can set custom limit of requests per second for your requirements

from aiowialon import Wialon
wialon = Wialon(rps=15)  # set custom requests per second limit

Prevent polling logout

By default start_polling autologout on Exception or on manual stop_polling. You can adjust it to your requirements

from aiowialon import Wialon
wialon = Wialon()  # set custom requests per second limit
wialon.start_polling(token=TOKEN, logout_finally=False)

Critical requests execution

Async session lock

Some requests to services like Render, Reports, Messages requires blocking other requests to be executed together per single session.

  • Use the @Wialon.lock_session decorator to block async loop till your operation done
  • You can apply @Wialon.session_lock also for handlers
  • You can use @Wialon.session_lock inside the methods when inheriting Wialon
import asyncio
from functools import wraps

from aiowialon import Wialon

wialon = Wialon(token=TOKEN)

@wialon.session_lock
async def critical_method(self, params1, params2):
  # For example: execute and export report
  previous_request_timeout = self.timeout  # Store current timeout
  try:
    self.timeout = 600  # Setup request timeout up to 10 minutes
    await self.report_exec_report(**params1)
    self.timeout = previous_request_timeout  # Return previous timeout
    report_result = await self.export_result(**params2)
    return report_result
  finally:
    self.timeout = previous_request_timeout  # Return previous timeout
    await self.report_cleanup_result()

With handlers:

@wialon.avl_event_handler(lambda event: event.data.i == 734455)
@wialon.session_lock
async def unit_event(event: AvlEvent):
  print("Handler got event:", event)
  # simulating long operation
  for i in range(5):
    print("Waiting lock release", i)
    await asyncio.sleep(1)

Timeout for API call

Some API calls requires special timeouts, cause them are processing long. Default timeout for aiohttp request is 5 seconds. You can set custom timeout on some call executing. It mostly usefull with @Wialon.session_lock

@wialon.avl_event_handler()
@wialon.session_lock
async def unit_event(event: AvlEvent):
    try:
        await wialon.wait(wialon.messages_load_last(
            itemId=event.data.i,
            lastTime=event.tm,
            lastCount=10000,
            flags=0x0000,
            flagsMask=0xFF00,
            loadCount=10000
        ), 10)
    except (TimeoutError, WialonError) as err:
        print(err)
    for i in range(5):
        print("Waiting exclusive operation", i, "item:", event.data.i)
        await asyncio.sleep(1)

Extending AIO Wialon

Inherit from Wialon class to add your custom logic and behaviour

  • You can directly use Wialon.request to make requests to special endpoints
  • You can use @wialon.session_lock inside the methods when inheriting Wialon
import json
import asyncio
from aiowialon import Wialon

class CustomWialon(Wialon):
  def __init__(self, **kwargs):
    super().__init__(**kwargs)
    self.__geocode_url = f"{kwargs.get('scheme', 'https')}://geocode-maps.wialon.com/{self.__base_url}/gis_geocode"

  async def geocode_address(self, coords, city_radius, dist_from_unit, txt_dist, flags):
    payload = {
      'coords': coords,
      ...  # other fields
    }
    return await self.request('geocode_fetch', self.__geocode_url, payload=json.dumps(payload))

  async def critical_method(self):
      @self.session_lock
      async def locked_task():
          # simulating long operation
          for i in range(5):
              print("Waiting lock release", i)
              await asyncio.sleep(1)
      return await locked_task()

Debugging

Enable debug messages for aiowialon and aiohttp

import logging
from aiowialon import Wialon, WialonError, flags, AvlEvent
logging.basicConfig(level=logging.DEBUG)

Warning

RISK NOTICE

THE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.

Copyright 2023 Yaroshenko Dmytro (https://github.com/o-murphy)