| 1 |
# coding: utf-8 |
| 2 |
|
| 3 |
# Copyright 2019, 2025 IBM All Rights Reserved. |
| 4 |
# |
| 5 |
# Licensed under the Apache License, Version 2.0 (the "License"); |
| 6 |
# you may not use this file except in compliance with the License. |
| 7 |
# You may obtain a copy of the License at |
| 8 |
# |
| 9 |
# http://www.apache.org/licenses/LICENSE-2.0 |
| 10 |
# |
| 11 |
# Unless required by applicable law or agreed to in writing, software |
| 12 |
# distributed under the License is distributed on an "AS IS" BASIS, |
| 13 |
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 14 |
# See the License for the specific language governing permissions and |
| 15 |
# limitations under the License. |
| 16 |
# from ibm_cloud_sdk_core.authenticators import Authenticator |
| 17 |
import datetime |
| 18 |
import gzip |
| 19 |
import io |
| 20 |
import json as json_import |
| 21 |
import re |
| 22 |
from os import getenv, environ, getcwd |
| 23 |
from os.path import isfile, join, expanduser |
| 24 |
from typing import List, Union |
| 25 |
from urllib.parse import urlparse, parse_qs |
| 26 |
|
| 27 |
import dateutil.parser as date_parser |
| 28 |
from .logger import get_logger |
| 29 |
|
| 30 |
logger = get_logger() |
| 31 |
|
| 32 |
|
| 33 |
class GzipStream(io.RawIOBase): |
| 34 |
"""Compress files on the fly. |
| 35 |
|
| 36 |
GzipStream is a helper class around the gzip library. It helps to |
| 37 |
compress already opened files (file-like objects) on the fly, so |
| 38 |
there is no need to read everything into the memory and call the |
| 39 |
`compress` function on it. |
| 40 |
The GzipFile is opened on the instance itself so it needs to act |
| 41 |
as a file-like object. |
| 42 |
|
| 43 |
Args: |
| 44 |
input: the source of the data to be compressed. |
| 45 |
It can be a file-like object, bytes or string. |
| 46 |
""" |
| 47 |
|
| 48 |
def __init__(self, source: Union[io.IOBase, bytes, str]): |
| 49 |
self.buffer = b'' |
| 50 |
|
| 51 |
if isinstance(source, io.IOBase): |
| 52 |
# The input is already a file-like object, use it as-is. |
| 53 |
self.uncompressed = source |
| 54 |
elif isinstance(source, str): |
| 55 |
# Strings must be handled with StringIO. |
| 56 |
self.uncompressed = io.StringIO(source) |
| 57 |
else: |
| 58 |
# Handle the rest as raw bytes. |
| 59 |
self.uncompressed = io.BytesIO(source) |
| 60 |
|
| 61 |
self.compressor = gzip.GzipFile(fileobj=self, mode='wb') |
| 62 |
|
| 63 |
def read(self, size: int = -1) -> bytes: |
| 64 |
"""Compresses and returns the requested size of data. |
| 65 |
|
| 66 |
Args: |
| 67 |
size: how many bytes to return. -1 to read and compress the whole file |
| 68 |
""" |
| 69 |
compressed = b'' |
| 70 |
|
| 71 |
if (size < 0) or (len(self.buffer) < size): |
| 72 |
for raw in self.uncompressed: |
| 73 |
# We need to encode text like streams (e.g. TextIOWrapper) to bytes. |
| 74 |
if isinstance(raw, str): |
| 75 |
raw = raw.encode() |
| 76 |
|
| 77 |
self.compressor.write(raw) |
| 78 |
|
| 79 |
# Stop compressing if we reached the max allowed size. |
| 80 |
if 0 < size < len(self.buffer): |
| 81 |
self.compressor.flush() |
| 82 |
break |
| 83 |
else: |
| 84 |
self.compressor.close() |
| 85 |
|
| 86 |
if size < 0: |
| 87 |
# Return all data from the buffer. |
| 88 |
compressed = self.buffer |
| 89 |
self.buffer = b'' |
| 90 |
else: |
| 91 |
# If we already have enough data in our buffer |
| 92 |
# return the desired chunk of bytes |
| 93 |
compressed = self.buffer[:size] |
| 94 |
# then remove them from the buffer. |
| 95 |
self.buffer = self.buffer[size:] |
| 96 |
|
| 97 |
return compressed |
| 98 |
|
| 99 |
def flush(self) -> None: |
| 100 |
"""Not implemented.""" |
| 101 |
# Since this "pipe" sits between 2 other stream (source/read -> target/write) |
| 102 |
# it wouldn't be worth to implemet flushing. |
| 103 |
pass |
| 104 |
|
| 105 |
def write(self, compressed: bytes) -> None: |
| 106 |
"""Append the compressed data to the buffer |
| 107 |
|
| 108 |
This happens when the target stream calls the `read` method and |
| 109 |
that triggers the gzip "compressor". |
| 110 |
""" |
| 111 |
self.buffer += compressed |
| 112 |
|
| 113 |
def close(self) -> None: |
| 114 |
"""Closes the underlying file-like object.""" |
| 115 |
self.uncompressed.close() |
| 116 |
|
| 117 |
|
| 118 |
def has_bad_first_or_last_char(val: str) -> bool: |
| 119 |
"""Returns true if a string starts with any of: {," ; or ends with any of: },". |
| 120 |
|
| 121 |
Args: |
| 122 |
val: The string to be tested. |
| 123 |
|
| 124 |
Returns: |
| 125 |
Whether or not the string starts or ends with bad characters. |
| 126 |
""" |
| 127 |
return val is not None and (val.startswith('{') or val.startswith('"') or val.endswith('}') or val.endswith('"')) |
| 128 |
|
| 129 |
|
| 130 |
def remove_null_values(dictionary: dict) -> dict: |
| 131 |
"""Create a new dictionary without keys mapped to null values. |
| 132 |
|
| 133 |
Args: |
| 134 |
dictionary: The dictionary potentially containing keys mapped to values of None. |
| 135 |
|
| 136 |
Returns: |
| 137 |
A dict with no keys mapped to None. |
| 138 |
""" |
| 139 |
if isinstance(dictionary, dict): |
| 140 |
return {k: v for (k, v) in dictionary.items() if v is not None} |
| 141 |
return dictionary |
| 142 |
|
| 143 |
|
| 144 |
def cleanup_values(dictionary: dict) -> dict: |
| 145 |
"""Create a new dictionary with boolean values converted to strings. |
| 146 |
|
| 147 |
Ex. true -> 'true', false -> 'false'. |
| 148 |
{ 'key': true } -> { 'key': 'true' } |
| 149 |
|
| 150 |
Args: |
| 151 |
dictionary: The dictionary with keys mapped to booleans. |
| 152 |
|
| 153 |
Returns: |
| 154 |
The dictionary with certain keys mapped to s and not booleans. |
| 155 |
""" |
| 156 |
if isinstance(dictionary, dict): |
| 157 |
return {k: cleanup_value(v) for (k, v) in dictionary.items()} |
| 158 |
return dictionary |
| 159 |
|
| 160 |
|
| 161 |
def cleanup_value(value: any) -> any: |
| 162 |
"""Convert a boolean value to string.""" |
| 163 |
if isinstance(value, bool): |
| 164 |
return 'true' if value else 'false' |
| 165 |
return value |
| 166 |
|
| 167 |
|
| 168 |
def strip_extra_slashes(value: str) -> str: |
| 169 |
"""Combine multiple trailing slashes to a single slash""" |
| 170 |
if value.endswith('//'): |
| 171 |
return value.rstrip('/') + '/' |
| 172 |
return value |
| 173 |
|
| 174 |
|
| 175 |
def datetime_to_string(val: datetime.datetime) -> str: |
| 176 |
"""Convert a datetime object to string. |
| 177 |
|
| 178 |
If the supplied datetime does not specify a timezone, |
| 179 |
it is assumed to be UTC. |
| 180 |
|
| 181 |
Args: |
| 182 |
val: The datetime object. |
| 183 |
|
| 184 |
Returns: |
| 185 |
datetime serialized to iso8601 format. |
| 186 |
""" |
| 187 |
if isinstance(val, datetime.datetime): |
| 188 |
if val.tzinfo is None: |
| 189 |
return val.isoformat() + 'Z' |
| 190 |
val = val.astimezone(datetime.timezone.utc) |
| 191 |
return val.isoformat().replace('+00:00', 'Z') |
| 192 |
return val |
| 193 |
|
| 194 |
|
| 195 |
def string_to_datetime(string: str) -> datetime.datetime: |
| 196 |
"""De-serializes string to datetime. |
| 197 |
|
| 198 |
Args: |
| 199 |
string: string containing datetime in iso8601 format. |
| 200 |
|
| 201 |
Returns: |
| 202 |
the de-serialized string as a datetime object. |
| 203 |
""" |
| 204 |
val = date_parser.parse(string) |
| 205 |
if val.tzinfo is not None: |
| 206 |
return val |
| 207 |
return val.replace(tzinfo=datetime.timezone.utc) |
| 208 |
|
| 209 |
|
| 210 |
def string_to_datetime_list(string_list: List[str]) -> List[datetime.datetime]: |
| 211 |
"""De-serializes each string in a list to a datetime. |
| 212 |
|
| 213 |
Args: |
| 214 |
string_list: list of strings containing datetime in iso8601 format. |
| 215 |
|
| 216 |
Returns: |
| 217 |
the de-serialized list of strings as a list of datetime objects. |
| 218 |
""" |
| 219 |
if not isinstance(string_list, list): |
| 220 |
raise ValueError( |
| 221 |
"Invalid argument type: " + str(type(string_list)) + ". Argument string_list must be of type List[str]" |
| 222 |
) |
| 223 |
datetime_list = [] |
| 224 |
for string_val in string_list: |
| 225 |
datetime_list.append(string_to_datetime(string_val)) |
| 226 |
return datetime_list |
| 227 |
|
| 228 |
|
| 229 |
def datetime_to_string_list(datetime_list: List[datetime.datetime]) -> List[str]: |
| 230 |
"""Convert a list of datetime objects to a list of strings. |
| 231 |
|
| 232 |
Args: |
| 233 |
datetime_list: The list of datetime objects. |
| 234 |
|
| 235 |
Returns: |
| 236 |
list of datetimes serialized as strings in iso8601 format. |
| 237 |
""" |
| 238 |
if not isinstance(datetime_list, list): |
| 239 |
raise ValueError( |
| 240 |
"Invalid argument type: " |
| 241 |
+ str(type(datetime_list)) |
| 242 |
+ ". Argument datetime_list must be of type List[datetime.datetime]" |
| 243 |
) |
| 244 |
string_list = [] |
| 245 |
for datetime_val in datetime_list: |
| 246 |
string_list.append(datetime_to_string(datetime_val)) |
| 247 |
return string_list |
| 248 |
|
| 249 |
|
| 250 |
def date_to_string(val: datetime.date) -> str: |
| 251 |
"""Convert a date object to string. |
| 252 |
|
| 253 |
Args: |
| 254 |
val: The date object. |
| 255 |
|
| 256 |
Returns: |
| 257 |
date serialized to `YYYY-MM-DD` format. |
| 258 |
""" |
| 259 |
if isinstance(val, datetime.date): |
| 260 |
return str(val) |
| 261 |
return val |
| 262 |
|
| 263 |
|
| 264 |
def string_to_date(string: str) -> datetime.date: |
| 265 |
"""De-serializes string to date. |
| 266 |
|
| 267 |
Args: |
| 268 |
string: string containing date in 'YYYY-MM-DD' format. |
| 269 |
|
| 270 |
Returns: |
| 271 |
the de-serialized string as a date object. |
| 272 |
""" |
| 273 |
return date_parser.parse(string).date() |
| 274 |
|
| 275 |
|
| 276 |
def string_to_bool(string: str) -> bool: |
| 277 |
"""Converts 'string' to a bool value. |
| 278 |
Args: |
| 279 |
string: the value to convert. This should be some form of "true" or "false" |
| 280 |
(e.g. "True", "TrUE", "False", "FaLsE", etc.) |
| 281 |
|
| 282 |
Returns: |
| 283 |
the boolean value |
| 284 |
""" |
| 285 |
return string.strip().lower() == 'true' |
| 286 |
|
| 287 |
|
| 288 |
def get_query_param(url_str: str, param: str) -> str: |
| 289 |
"""Return a query parameter value from url_str |
| 290 |
|
| 291 |
Args: |
| 292 |
url_str: the URL from which to extract the query |
| 293 |
parameter value |
| 294 |
param: the name of the query parameter whose value |
| 295 |
should be returned |
| 296 |
|
| 297 |
Returns: |
| 298 |
the value of the `param` query parameter as a string |
| 299 |
|
| 300 |
Raises: |
| 301 |
ValueError: if errors are encountered parsing `url_str` |
| 302 |
""" |
| 303 |
if not url_str: |
| 304 |
return None |
| 305 |
url = urlparse(url_str) |
| 306 |
if not url.query: |
| 307 |
return None |
| 308 |
query = parse_qs(url.query, strict_parsing=True) |
| 309 |
values = query.get(param) |
| 310 |
return values[0] if values else None |
| 311 |
|
| 312 |
|
| 313 |
def convert_model(val: any) -> dict: |
| 314 |
"""Convert a model object into an equivalent dict. |
| 315 |
|
| 316 |
Arguments: |
| 317 |
val: A dict or a model object |
| 318 |
|
| 319 |
Returns: |
| 320 |
A dict representation of the input object. |
| 321 |
""" |
| 322 |
if isinstance(val, dict): |
| 323 |
return val |
| 324 |
if hasattr(val, "to_dict"): |
| 325 |
return val.to_dict() |
| 326 |
# Consider raising a ValueError here in the next major release |
| 327 |
return val |
| 328 |
|
| 329 |
|
| 330 |
def convert_list(val: Union[str, List[str]]) -> str: |
| 331 |
"""Convert a list of strings into comma-separated string. |
| 332 |
|
| 333 |
Arguments: |
| 334 |
val: A string or list of strings |
| 335 |
|
| 336 |
Returns: |
| 337 |
A comma-separated string of the items in the input list. |
| 338 |
""" |
| 339 |
if isinstance(val, str): |
| 340 |
return val |
| 341 |
if isinstance(val, list) and all(isinstance(x, str) for x in val): |
| 342 |
return ",".join(val) |
| 343 |
# Consider raising a ValueError here in the next major release |
| 344 |
return val |
| 345 |
|
| 346 |
|
| 347 |
def read_external_sources(service_name: str) -> dict: |
| 348 |
"""Look for external configuration of a service. |
| 349 |
|
| 350 |
Try to get config from external sources, with the following priority: |
| 351 |
1. Credentials file(ibm-credentials.env) |
| 352 |
2. Environment variables |
| 353 |
3. VCAP Services(Cloud Foundry) |
| 354 |
|
| 355 |
Args: |
| 356 |
service_name: The service name |
| 357 |
|
| 358 |
Returns: |
| 359 |
A dictionary containing relevant configuration for the service if found. |
| 360 |
""" |
| 361 |
logger.debug('Retrieving config properties for service \'%s\'', service_name) |
| 362 |
config = {} |
| 363 |
|
| 364 |
config = __read_from_credential_file(service_name) |
| 365 |
|
| 366 |
if not config: |
| 367 |
config = __read_from_env_variables(service_name) |
| 368 |
|
| 369 |
if not config: |
| 370 |
config = __read_from_vcap_services(service_name) |
| 371 |
logger.debug('Retrieved %d properties', len(config)) |
| 372 |
return config |
| 373 |
|
| 374 |
|
| 375 |
def __read_from_env_variables(service_name: str) -> dict: |
| 376 |
"""Return a config object based on environment variables for a service. |
| 377 |
|
| 378 |
Args: |
| 379 |
service_name: The name of the service to look for in env variables. |
| 380 |
|
| 381 |
Returns: |
| 382 |
A set of service configuration key-value pairs. |
| 383 |
""" |
| 384 |
config = {} |
| 385 |
for key, value in environ.items(): |
| 386 |
_parse_key_and_update_config(config, service_name, key, value) |
| 387 |
return config |
| 388 |
|
| 389 |
|
| 390 |
def __read_from_credential_file(service_name: str, *, separator: str = '=') -> dict: |
| 391 |
"""Return a config object based on credentials file for a service. |
| 392 |
|
| 393 |
Args: |
| 394 |
service_name: The name of the service to look for in env variables. |
| 395 |
|
| 396 |
Keyword Args: |
| 397 |
separator: The character to split on to de-serialize a line into a key-value pair. |
| 398 |
|
| 399 |
Returns: |
| 400 |
A set of service configuration key-value pairs. |
| 401 |
""" |
| 402 |
default_credentials_file_name = 'ibm-credentials.env' |
| 403 |
|
| 404 |
# 1. ${IBM_CREDENTIALS_FILE} |
| 405 |
credential_file_path = getenv('IBM_CREDENTIALS_FILE') |
| 406 |
|
| 407 |
# 2. <current-working-directory>/ibm-credentials.env |
| 408 |
if credential_file_path is None: |
| 409 |
file_path = join(getcwd(), default_credentials_file_name) |
| 410 |
if isfile(file_path): |
| 411 |
credential_file_path = file_path |
| 412 |
|
| 413 |
# 3. <user-home-directory>/ibm-credentials.env |
| 414 |
if credential_file_path is None: |
| 415 |
file_path = join(expanduser('~'), default_credentials_file_name) |
| 416 |
if isfile(file_path): |
| 417 |
credential_file_path = file_path |
| 418 |
|
| 419 |
config = {} |
| 420 |
if credential_file_path is not None: |
| 421 |
try: |
| 422 |
with open(credential_file_path, 'r', encoding='utf-8') as fobj: |
| 423 |
for line in fobj: |
| 424 |
key_val = line.strip().split(separator, 1) |
| 425 |
if len(key_val) == 2: |
| 426 |
key = key_val[0] |
| 427 |
value = key_val[1] |
| 428 |
_parse_key_and_update_config(config, service_name, key, value) |
| 429 |
except OSError: |
| 430 |
# just absorb the exception and make sure we return an empty response |
| 431 |
config = {} |
| 432 |
|
| 433 |
return config |
| 434 |
|
| 435 |
|
| 436 |
def _parse_key_and_update_config(config: dict, service_name: str, key: str, value: str) -> None: |
| 437 |
service_name = service_name.replace(' ', '_').replace('-', '_').upper() |
| 438 |
if key.startswith(service_name): |
| 439 |
config[key[len(service_name) + 1 :]] = value |
| 440 |
|
| 441 |
|
| 442 |
def __read_from_vcap_services(service_name: str) -> dict: |
| 443 |
"""Return a config object based on the vcap services environment variable. |
| 444 |
|
| 445 |
Args: |
| 446 |
service_name: The name of the service to look for in the vcap. |
| 447 |
|
| 448 |
Returns: |
| 449 |
A set of service configuration key-value pairs. |
| 450 |
""" |
| 451 |
vcap_services = getenv('VCAP_SERVICES') |
| 452 |
vcap_service_credentials = {} |
| 453 |
if vcap_services: |
| 454 |
services = json_import.loads(vcap_services) |
| 455 |
for key in services.keys(): |
| 456 |
for i in range(len(services[key])): |
| 457 |
if vcap_service_credentials and isinstance(vcap_service_credentials, dict): |
| 458 |
break |
| 459 |
if services[key][i].get('name') == service_name: |
| 460 |
vcap_service_credentials = services[key][i].get('credentials', {}) |
| 461 |
if not vcap_service_credentials: |
| 462 |
if service_name in services.keys(): |
| 463 |
service = services.get(service_name) |
| 464 |
if service: |
| 465 |
vcap_service_credentials = service[0].get('credentials', {}) |
| 466 |
|
| 467 |
if vcap_service_credentials and isinstance(vcap_service_credentials, dict): |
| 468 |
new_vcap_creds = {} |
| 469 |
# cf |
| 470 |
if vcap_service_credentials.get('username') and vcap_service_credentials.get('password'): |
| 471 |
new_vcap_creds['AUTH_TYPE'] = 'basic' |
| 472 |
new_vcap_creds['USERNAME'] = vcap_service_credentials.get('username') |
| 473 |
new_vcap_creds['PASSWORD'] = vcap_service_credentials.get('password') |
| 474 |
vcap_service_credentials = new_vcap_creds |
| 475 |
elif vcap_service_credentials.get('iam_apikey'): |
| 476 |
new_vcap_creds['AUTH_TYPE'] = 'iam' |
| 477 |
new_vcap_creds['APIKEY'] = vcap_service_credentials.get('iam_apikey') |
| 478 |
vcap_service_credentials = new_vcap_creds |
| 479 |
elif vcap_service_credentials.get('apikey'): |
| 480 |
new_vcap_creds['AUTH_TYPE'] = 'iam' |
| 481 |
new_vcap_creds['APIKEY'] = vcap_service_credentials.get('apikey') |
| 482 |
vcap_service_credentials = new_vcap_creds |
| 483 |
return vcap_service_credentials |
| 484 |
|
| 485 |
|
| 486 |
# A regex that matches an "application/json" mimetype. |
| 487 |
json_mimetype_pattern = re.compile('^application/json(\\s*;.*)?$') |
| 488 |
|
| 489 |
|
| 490 |
def is_json_mimetype(mimetype: str) -> bool: |
| 491 |
"""Returns true if 'mimetype' is a JSON-like mimetype, false otherwise. |
| 492 |
|
| 493 |
Args: |
| 494 |
mimetype: The mimetype to check. |
| 495 |
|
| 496 |
Returns: |
| 497 |
true if mimetype is a JSON-line mimetype, false otherwise. |
| 498 |
""" |
| 499 |
return mimetype is not None and json_mimetype_pattern.match(mimetype) is not None |