arista/anta
1
0
Fork 0
Code Issues Activity

Adding upstream version 0.15.0.

Browse source 
Signed-off-by: Daniel Baumann <daniel@debian.org>
This commit is contained in:
Daniel Baumann 2025-02-05 11:39:42 +01:00
parent 6721599912
commit a1777afd4b
Signed by: daniel
GPG key ID: FBB4F0E80A80222F
103 changed files with 79620 additions and 742 deletions
Download patch file Download diff file Expand all files Collapse all files

291
asynceapi/device.py Normal file
View file

@ -0,0 +1,291 @@
# Copyright (c) 2024 Arista Networks, Inc.
# Use of this source code is governed by the Apache License 2.0
# that can be found in the LICENSE file.
# Initially written by Jeremy Schulman at https://github.com/jeremyschulman/aio-eapi
"""asynceapi.Device definition."""
# -----------------------------------------------------------------------------
# System Imports
# -----------------------------------------------------------------------------
from __future__ import annotations
from socket import getservbyname
from typing import TYPE_CHECKING, Any
# -----------------------------------------------------------------------------
# Public Imports
# -----------------------------------------------------------------------------
import httpx
# -----------------------------------------------------------------------------
# Private Imports
# -----------------------------------------------------------------------------
from .aio_portcheck import port_check_url
from .config_session import SessionConfig
from .errors import EapiCommandError
if TYPE_CHECKING:
from collections.abc import Sequence
# -----------------------------------------------------------------------------
# Exports
# -----------------------------------------------------------------------------
__all__ = ["Device"]
# -----------------------------------------------------------------------------
#
# CODE BEGINS
#
# -----------------------------------------------------------------------------
class Device(httpx.AsyncClient):
"""
Represent the async JSON-RPC client that communicates with an Arista EOS device.
This class inherits directly from the
httpx.AsyncClient, so any initialization options can be passed directly.
"""
auth = None
EAPI_OFMT_OPTIONS = ("json", "text")
EAPI_DEFAULT_OFMT = "json"
def __init__( # noqa: PLR0913 # pylint: disable=too-many-arguments
self,
host: str | None = None,
username: str | None = None,
password: str | None = None,
proto: str = "https",
port: str | int | None = None,
**kwargs: Any, # noqa: ANN401
) -> None:
"""
Initialize the Device class.
As a subclass to httpx.AsyncClient, the caller can provide any of those initializers.
Specific parameters for Device class are all optional and described below.
Parameters
----------
host: The EOS target device, either hostname (DNS) or ipaddress.
username: The login user-name; requires the password parameter.
password: The login password; requires the username parameter.
proto: The protocol, http or https, to communicate eAPI with the device.
port: If not provided, the proto value is used to look up the associated
port (http=80, https=443). If provided, overrides the port used to
communite with the device.
Other Parameters
----------------
base_url: str
If provided, the complete URL to the device eAPI endpoint.
auth:
If provided, used as the httpx authorization initializer value. If
not provided, then username+password is assumed by the Caller and
used to create a BasicAuth instance.
"""
self.port = port or getservbyname(proto)
self.host = host
kwargs.setdefault("base_url", httpx.URL(f"{proto}://{self.host}:{self.port}"))
kwargs.setdefault("verify", False)
if username and password:
self.auth = httpx.BasicAuth(username, password)
kwargs.setdefault("auth", self.auth)
super().__init__(**kwargs)
self.headers["Content-Type"] = "application/json-rpc"
async def check_connection(self) -> bool:
"""
Check the target device to ensure that the eAPI port is open and accepting connections.
It is recommended that a Caller checks the connection before involving cli commands,
but this step is not required.
Returns
-------
True when the device eAPI is accessible, False otherwise.
"""
return await port_check_url(self.base_url)
async def cli( # noqa: PLR0913 # pylint: disable=too-many-arguments
self,
command: str | dict[str, Any] | None = None,
commands: Sequence[str | dict[str, Any]] | None = None,
ofmt: str | None = None,
version: int | str | None = "latest",
*,
suppress_error: bool = False,
auto_complete: bool = False,
expand_aliases: bool = False,
req_id: int | str | None = None,
) -> list[dict[str, Any] | str] | dict[str, Any] | str | None:
"""
Execute one or more CLI commands.
Parameters
----------
command:
A single command to execute; results in a single output response
commands:
A list of commands to execute; results in a list of output responses
ofmt:
Either 'json' or 'text'; indicates the output format for the CLI commands.
version:
By default the eAPI will use "version 1" for all API object models.
This driver will, by default, always set version to "latest" so
that the behavior matches the CLI of the device. The caller can
override the "latest" behavior by explicitly setting the version.
suppress_error:
When not False, then if the execution of the command would-have
raised an EapiCommandError, rather than raising this exception this
routine will return the value None.
For example, if the following command had raised
EapiCommandError, now response would be set to None instead.
response = dev.cli(..., suppress_error=True)
auto_complete:
Enabled/disables the command auto-compelete feature of the EAPI. Per the
documentation:
Allows users to use shorthand commands in eAPI calls. With this
parameter included a user can send 'sh ver' via eAPI to get the
output of 'show version'.
expand_aliases:
Enables/disables the command use of User defined alias. Per the
documentation:
Allowed users to provide the expandAliases parameter to eAPI
calls. This allows users to use aliased commands via the API.
For example if an alias is configured as 'sv' for 'show version'
then an API call with sv and the expandAliases parameter will
return the output of show version.
req_id:
A unique identifier that will be echoed back by the switch. May be a string or number.
Returns
-------
One or List of output responses, per the description above.
"""
if not any((command, commands)):
msg = "Required 'command' or 'commands'"
raise RuntimeError(msg)
jsonrpc = self._jsonrpc_command(
commands=[command] if command else commands, ofmt=ofmt, version=version, auto_complete=auto_complete, expand_aliases=expand_aliases, req_id=req_id
)
try:
res = await self.jsonrpc_exec(jsonrpc)
return res[0] if command else res
except EapiCommandError:
if suppress_error:
return None
raise
def _jsonrpc_command( # noqa: PLR0913 # pylint: disable=too-many-arguments
self,
commands: Sequence[str | dict[str, Any]] | None = None,
ofmt: str | None = None,
version: int | str | None = "latest",
*,
auto_complete: bool = False,
expand_aliases: bool = False,
req_id: int | str | None = None,
) -> dict[str, Any]:
"""Create the JSON-RPC command dictionary object."""
cmd: dict[str, Any] = {
"jsonrpc": "2.0",
"method": "runCmds",
"params": {
"version": version,
"cmds": commands,
"format": ofmt or self.EAPI_DEFAULT_OFMT,
},
"id": req_id or id(self),
}
if auto_complete is not None:
cmd["params"].update({"autoComplete": auto_complete})
if expand_aliases is not None:
cmd["params"].update({"expandAliases": expand_aliases})
return cmd
async def jsonrpc_exec(self, jsonrpc: dict[str, Any]) -> list[dict[str, Any] | str]:
"""
Execute the JSON-RPC dictionary object.
Parameters
----------
jsonrpc:
The JSON-RPC as created by the `meth`:_jsonrpc_command().
Raises
------
EapiCommandError
In the event that a command resulted in an error response.
Returns
-------
The list of command results; either dict or text depending on the
JSON-RPC format parameter.
"""
res = await self.post("/command-api", json=jsonrpc)
res.raise_for_status()
body = res.json()
commands = jsonrpc["params"]["cmds"]
ofmt = jsonrpc["params"]["format"]
get_output = (lambda _r: _r["output"]) if ofmt == "text" else (lambda _r: _r)
# if there are no errors then return the list of command results.
if (err_data := body.get("error")) is None:
return [get_output(cmd_res) for cmd_res in body["result"]]
# ---------------------------------------------------------------------
# if we are here, then there were some command errors. Raise a
# EapiCommandError exception with args (commands that failed, passed,
# not-executed).
# ---------------------------------------------------------------------
# -------------------------- eAPI specification ----------------------
# On an error, no result object is present, only an error object, which
# is guaranteed to have the following attributes: code, messages, and
# data. Similar to the result object in the successful response, the
# data object is a list of objects corresponding to the results of all
# commands up to, and including, the failed command. If there was a an
# error before any commands were executed (e.g. bad credentials), data
# will be empty. The last object in the data array will always
# correspond to the failed command. The command failure details are
# always stored in the errors array.
cmd_data = err_data["data"]
len_data = len(cmd_data)
err_at = len_data - 1
err_msg = err_data["message"]
raise EapiCommandError(
passed=[get_output(cmd_data[cmd_i]) for cmd_i, cmd in enumerate(commands[:err_at])],
failed=commands[err_at]["cmd"],
errors=cmd_data[err_at]["errors"],
errmsg=err_msg,
not_exec=commands[err_at + 1 :],
)
def config_session(self, name: str) -> SessionConfig:
"""
return a SessionConfig instance bound to this device with the given session name.
Parameters
----------
name: The config-session name
"""
return SessionConfig(self, name)