I solved this problem like this:
Add a class to the rasa/core/utils.py file, this class is like this:
class MyEncoder(json.JSONEncoder):
def default(self, obj):
if not isinstance(obj, Dict):
# return obj.__class__.__name__
return hash(obj)
else:
return super(MyEncoder, self).default(obj)
then modify get_dict_hash() to
def get_dict_hash(data: Dict, encoding: Text = "utf-8") -> Text:
"""Calculate the md5 hash of a dictionary.
这里加入了cls=MyEncoder的类作为参数
"""
return md5(json.dumps(data, cls=MyEncoder, sort_keys=True).encode(encoding)).hexdigest()
The whole file rasa/core/utils.py is like this:
# -*- coding: utf-8 -*-
import argparse
import asyncio
import json
import logging
import re
import sys
from pathlib import Path
from typing import Union
from asyncio import Future
from hashlib import md5, sha1
from io import StringIO
from typing import Any, Dict, List, Optional, Set, TYPE_CHECKING, Text, Tuple, Callable
import aiohttp
from aiohttp import InvalidURL
from sanic import Sanic
from sanic.views import CompositionView
import rasa.utils.io as io_utils
from rasa.utils.endpoints import read_endpoint_config
# backwards compatibility 1.0.x
# noinspection PyUnresolvedReferences
from rasa.utils.endpoints import concat_url
logger = logging.getLogger(__name__)
if TYPE_CHECKING:
from random import Random
def configure_file_logging(logger_obj: logging.Logger, log_file: Optional[Text]):
if not log_file:
return
formatter = logging.Formatter("%(asctime)s [%(levelname)-5.5s] %(message)s")
file_handler = logging.FileHandler(log_file)
file_handler.setLevel(logger_obj.level)
file_handler.setFormatter(formatter)
logger_obj.addHandler(file_handler)
def module_path_from_instance(inst: Any) -> Text:
"""Return the module path of an instance's class."""
return inst.__module__ + "." + inst.__class__.__name__
def dump_obj_as_json_to_file(filename: Text, obj: Any) -> None:
"""Dump an object as a json string to a file."""
dump_obj_as_str_to_file(filename, json.dumps(obj, indent=2))
def dump_obj_as_str_to_file(filename: Text, text: Text) -> None:
"""Dump a text to a file."""
with open(filename, "w", encoding="utf-8") as f:
# noinspection PyTypeChecker
f.write(str(text))
def subsample_array(
arr: List[Any],
max_values: int,
can_modify_incoming_array: bool = True,
rand: Optional["Random"] = None,
) -> List[Any]:
"""Shuffles the array and returns `max_values` number of elements."""
import random
if not can_modify_incoming_array:
arr = arr[:]
if rand is not None:
rand.shuffle(arr)
else:
random.shuffle(arr)
return arr[:max_values]
def is_int(value: Any) -> bool:
"""Checks if a value is an integer.
The type of the value is not important, it might be an int or a float."""
# noinspection PyBroadException
try:
return value == int(value)
except Exception:
return False
def lazyproperty(fn):
"""Allows to avoid recomputing a property over and over.
Instead the result gets stored in a local var. Computation of the property
will happen once, on the first call of the property. All succeeding calls
will use the value stored in the private property."""
attr_name = "_lazy_" + fn.__name__
@property
def _lazyprop(self):
if not hasattr(self, attr_name):
setattr(self, attr_name, fn(self))
return getattr(self, attr_name)
return _lazyprop
def one_hot(hot_idx, length, dtype=None):
import numpy
if hot_idx >= length:
raise ValueError(
"Can't create one hot. Index '{}' is out "
"of range (length '{}')".format(hot_idx, length)
)
r = numpy.zeros(length, dtype)
r[hot_idx] = 1
return r
def str_range_list(start, end):
return [str(e) for e in range(start, end)]
def generate_id(prefix="", max_chars=None):
import uuid
gid = uuid.uuid4().hex
if max_chars:
gid = gid[:max_chars]
return "{}{}".format(prefix, gid)
def request_input(valid_values=None, prompt=None, max_suggested=3):
def wrong_input_message():
print (
"Invalid answer, only {}{} allowed\n".format(
", ".join(valid_values[:max_suggested]),
",..." if len(valid_values) > max_suggested else "",
)
)
while True:
try:
input_value = input(prompt) if prompt else input()
if valid_values is not None and input_value not in valid_values:
wrong_input_message()
continue
except ValueError:
wrong_input_message()
continue
return input_value
class MyEncoder(json.JSONEncoder):
"""模型指纹化时，检查字典类型"""
def default(self, obj):
if not isinstance(obj, Dict):
# return obj.__class__.__name__
return hash(obj)
else:
return super(MyEncoder, self).default(obj)
# noinspection PyPep8Naming
class HashableNDArray(object):
"""Hashable wrapper for ndarray objects.
Instances of ndarray are not hashable, meaning they cannot be added to
sets, nor used as keys in dictionaries. This is by design - ndarray
objects are mutable, and therefore cannot reliably implement the
__hash__() method.
The hashable class allows a way around this limitation. It implements
the required methods for hashable objects in terms of an encapsulated
ndarray object. This can be either a copied instance (which is safer)
or the original object (which requires the user to be careful enough
not to modify it)."""
def __init__(self, wrapped, tight=False):
"""Creates a new hashable object encapsulating an ndarray.
wrapped
The wrapped ndarray.
tight
Optional. If True, a copy of the input ndaray is created.
Defaults to False.
"""
from numpy import array
self.__tight = tight
self.__wrapped = array(wrapped) if tight else wrapped
self.__hash = int(sha1(wrapped.view()).hexdigest(), 16)
def __eq__(self, other):
from numpy import all
return all(self.__wrapped == other.__wrapped)
def __hash__(self):
return self.__hash
def unwrap(self):
"""Returns the encapsulated ndarray.
If the wrapper is "tight", a copy of the encapsulated ndarray is
returned. Otherwise, the encapsulated ndarray itself is returned."""
from numpy import array
if self.__tight:
return array(self.__wrapped)
return self.__wrapped
def _dump_yaml(obj, output):
import ruamel.yaml
yaml_writer = ruamel.yaml.YAML(pure=True, typ="safe")
yaml_writer.unicode_supplementary = True
yaml_writer.default_flow_style = False
yaml_writer.version = "1.1"
yaml_writer.dump(obj, output)
def dump_obj_as_yaml_to_file(filename: Union[Text, Path], obj: Dict) -> None:
"""Writes data (python dict) to the filename in yaml repr."""
with open(str(filename), "w", encoding="utf-8") as output:
_dump_yaml(obj, output)
def dump_obj_as_yaml_to_string(obj: Dict) -> Text:
"""Writes data (python dict) to a yaml string."""
str_io = StringIO()
_dump_yaml(obj, str_io)
return str_io.getvalue()
def list_routes(app: Sanic):
"""List all the routes of a sanic application.
Mainly used for debugging."""
from urllib.parse import unquote
output = {}
def find_route(suffix, path):
for name, (uri, _) in app.router.routes_names.items():
if name.split(".")[-1] == suffix and uri == path:
return name
return None
for endpoint, route in app.router.routes_all.items():
if endpoint[:-1] in app.router.routes_all and endpoint[-1] == "/":
continue
options = {}
for arg in route.parameters:
options[arg] = "[{0}]".format(arg)
if not isinstance(route.handler, CompositionView):
handlers = [(list(route.methods)[0], route.name)]
else:
handlers = [
(method, find_route(v.__name__, endpoint) or v.__name__)
for method, v in route.handler.handlers.items()
]
for method, name in handlers:
line = unquote("{:50s} {:30s} {}".format(endpoint, method, name))
output[name] = line
url_table = "\n".join(output[url] for url in sorted(output))
logger.debug("Available web server routes: \n{}".format(url_table))
return output
def cap_length(s, char_limit=20, append_ellipsis=True):
"""Makes sure the string doesn't exceed the passed char limit.
Appends an ellipsis if the string is to long."""
if len(s) > char_limit:
if append_ellipsis:
return s[: char_limit - 3] + "..."
else:
return s[:char_limit]
else:
return s
def extract_args(
kwargs: Dict[Text, Any], keys_to_extract: Set[Text]
) -> Tuple[Dict[Text, Any], Dict[Text, Any]]:
"""Go through the kwargs and filter out the specified keys.
Return both, the filtered kwargs as well as the remaining kwargs."""
remaining = {}
extracted = {}
for k, v in kwargs.items():
if k in keys_to_extract:
extracted[k] = v
else:
remaining[k] = v
return extracted, remaining
def all_subclasses(cls: Any) -> List[Any]:
"""Returns all known (imported) subclasses of a class."""
return cls.__subclasses__() + [
g for s in cls.__subclasses__() for g in all_subclasses(s)
]
def is_limit_reached(num_messages, limit):
return limit is not None and num_messages >= limit
def read_lines(filename, max_line_limit=None, line_pattern=".*"):
"""Read messages from the command line and print bot responses."""
line_filter = re.compile(line_pattern)
with open(filename, "r", encoding="utf-8") as f:
num_messages = 0
for line in f:
m = line_filter.match(line)
if m is not None:
yield m.group(1 if m.lastindex else 0)
num_messages += 1
if is_limit_reached(num_messages, max_line_limit):
break
def file_as_bytes(path: Text) -> bytes:
"""Read in a file as a byte array."""
with open(path, "rb") as f:
return f.read()
def get_file_hash(path: Text) -> Text:
"""Calculate the md5 hash of a file."""
return md5(file_as_bytes(path)).hexdigest()
def get_text_hash(text: Text, encoding: Text = "utf-8") -> Text:
"""Calculate the md5 hash for a text."""
return md5(text.encode(encoding)).hexdigest()
def get_dict_hash(data: Dict, encoding: Text = "utf-8") -> Text:
"""Calculate the md5 hash of a dictionary.
这里加入了cls=MyEncoder的自建类
"""
return md5(json.dumps(data, cls=MyEncoder, sort_keys=True).encode(encoding)).hexdigest()
async def download_file_from_url(url: Text) -> Text:
"""Download a story file from a url and persists it into a temp file.
Returns the file path of the temp file that contains the
downloaded content."""
from rasa.nlu import utils as nlu_utils
if not nlu_utils.is_url(url):
raise InvalidURL(url)
async with aiohttp.ClientSession() as session:
async with session.get(url, raise_for_status=True) as resp:
filename = io_utils.create_temporary_file(await resp.read(), mode="w+b")
return filename
def remove_none_values(obj: Dict[Text, Any]) -> Dict[Text, Any]:
"""Remove all keys that store a `None` value."""
return {k: v for k, v in obj.items() if v is not None}
def pad_list_to_size(_list, size, padding_value=None):
"""Pads _list with padding_value up to size"""
return _list + [padding_value] * (size - len(_list))
class AvailableEndpoints(object):
"""Collection of configured endpoints."""
@classmethod
def read_endpoints(cls, endpoint_file):
nlg = read_endpoint_config(endpoint_file, endpoint_type="nlg")
nlu = read_endpoint_config(endpoint_file, endpoint_type="nlu")
action = read_endpoint_config(endpoint_file, endpoint_type="action_endpoint")
model = read_endpoint_config(endpoint_file, endpoint_type="models")
tracker_store = read_endpoint_config(
endpoint_file, endpoint_type="tracker_store"
)
event_broker = read_endpoint_config(endpoint_file, endpoint_type="event_broker")
return cls(nlg, nlu, action, model, tracker_store, event_broker)
def __init__(
self,
nlg=None,
nlu=None,
action=None,
model=None,
tracker_store=None,
event_broker=None,
):
self.model = model
self.action = action
self.nlu = nlu
self.nlg = nlg
self.tracker_store = tracker_store
self.event_broker = event_broker
# noinspection PyProtectedMember
def set_default_subparser(parser, default_subparser):
"""default subparser selection. Call after setup, just before parse_args()
parser: the name of the parser you're making changes to
default_subparser: the name of the subparser to call by default"""
subparser_found = False
for arg in sys.argv[1:]:
if arg in ["-h", "--help"]: # global help if no subparser
break
else:
for x in parser._subparsers._actions:
if not isinstance(x, argparse._SubParsersAction):
continue
for sp_name in x._name_parser_map.keys():
if sp_name in sys.argv[1:]:
subparser_found = True
if not subparser_found:
# insert default in first position before all other arguments
sys.argv.insert(1, default_subparser)
def create_task_error_logger(error_message: Text = "") -> Callable[[Future], None]:
"""Error logger to be attached to a task.
This will ensure exceptions are properly logged and won't get lost."""
def handler(fut: Future) -> None:
# noinspection PyBroadException
try:
fut.result()
except Exception:
logger.exception(
"An exception was raised while running task. "
"{}".format(error_message)
)
return handler
class LockCounter(asyncio.Lock):
"""Decorated asyncio lock that counts how many coroutines are waiting.
The counter can be used to discard the lock when there is no coroutine
waiting for it. For this to work, there should not be any execution yield
between retrieving the lock and acquiring it, otherwise there might be
race conditions."""
def __init__(self) -> None:
super().__init__()
self.wait_counter = 0
async def acquire(self) -> bool:
"""Acquire the lock, makes sure only one coroutine can retrieve it."""
self.wait_counter += 1
try:
return await super(LockCounter, self).acquire() # type: ignore
finally:
self.wait_counter -= 1
def is_someone_waiting(self) -> bool:
"""Check if a coroutine is waiting for this lock to be freed."""
return self.wait_counter != 0