chunkr-ai 0.1.0a6__py3-none-any.whl → 0.1.0a8__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (69) hide show
  1. chunkr_ai/__init__.py +2 -0
  2. chunkr_ai/_base_client.py +3 -3
  3. chunkr_ai/_client.py +31 -3
  4. chunkr_ai/_compat.py +48 -48
  5. chunkr_ai/_constants.py +5 -5
  6. chunkr_ai/_exceptions.py +4 -0
  7. chunkr_ai/_models.py +41 -41
  8. chunkr_ai/_types.py +35 -1
  9. chunkr_ai/_utils/__init__.py +9 -2
  10. chunkr_ai/_utils/_compat.py +45 -0
  11. chunkr_ai/_utils/_datetime_parse.py +136 -0
  12. chunkr_ai/_utils/_transform.py +11 -1
  13. chunkr_ai/_utils/_typing.py +6 -1
  14. chunkr_ai/_utils/_utils.py +0 -1
  15. chunkr_ai/_version.py +1 -1
  16. chunkr_ai/resources/__init__.py +14 -0
  17. chunkr_ai/resources/files.py +3 -3
  18. chunkr_ai/resources/tasks/__init__.py +14 -0
  19. chunkr_ai/resources/tasks/extract.py +393 -0
  20. chunkr_ai/resources/tasks/parse.py +110 -286
  21. chunkr_ai/resources/tasks/tasks.py +64 -32
  22. chunkr_ai/resources/webhooks.py +193 -0
  23. chunkr_ai/types/__init__.py +27 -1
  24. chunkr_ai/types/bounding_box.py +19 -0
  25. chunkr_ai/types/cell.py +39 -0
  26. chunkr_ai/types/cell_style.py +28 -0
  27. chunkr_ai/types/chunk.py +40 -0
  28. chunkr_ai/types/chunk_processing.py +40 -0
  29. chunkr_ai/types/chunk_processing_param.py +42 -0
  30. chunkr_ai/types/extract_configuration.py +24 -0
  31. chunkr_ai/types/extract_output_response.py +62 -0
  32. chunkr_ai/types/file_create_params.py +2 -1
  33. chunkr_ai/types/file_info.py +21 -0
  34. chunkr_ai/types/generation_config.py +29 -0
  35. chunkr_ai/types/generation_config_param.py +29 -0
  36. chunkr_ai/types/llm_processing.py +36 -0
  37. chunkr_ai/types/llm_processing_param.py +36 -0
  38. chunkr_ai/types/ocr_result.py +28 -0
  39. chunkr_ai/types/page.py +27 -0
  40. chunkr_ai/types/parse_configuration.py +64 -0
  41. chunkr_ai/types/parse_configuration_param.py +65 -0
  42. chunkr_ai/types/parse_output_response.py +29 -0
  43. chunkr_ai/types/segment.py +109 -0
  44. chunkr_ai/types/segment_processing.py +228 -0
  45. chunkr_ai/types/segment_processing_param.py +229 -0
  46. chunkr_ai/types/task_extract_updated_webhook_event.py +22 -0
  47. chunkr_ai/types/task_get_params.py +0 -3
  48. chunkr_ai/types/task_list_params.py +7 -1
  49. chunkr_ai/types/task_parse_updated_webhook_event.py +22 -0
  50. chunkr_ai/types/task_response.py +68 -0
  51. chunkr_ai/types/tasks/__init__.py +7 -1
  52. chunkr_ai/types/tasks/extract_create_params.py +47 -0
  53. chunkr_ai/types/tasks/extract_create_response.py +67 -0
  54. chunkr_ai/types/tasks/extract_get_params.py +18 -0
  55. chunkr_ai/types/tasks/extract_get_response.py +67 -0
  56. chunkr_ai/types/tasks/parse_create_params.py +25 -793
  57. chunkr_ai/types/tasks/parse_create_response.py +55 -0
  58. chunkr_ai/types/tasks/parse_get_params.py +18 -0
  59. chunkr_ai/types/tasks/parse_get_response.py +55 -0
  60. chunkr_ai/types/unwrap_webhook_event.py +11 -0
  61. chunkr_ai/types/version_info.py +31 -0
  62. chunkr_ai/types/webhook_url_response.py +9 -0
  63. {chunkr_ai-0.1.0a6.dist-info → chunkr_ai-0.1.0a8.dist-info}/METADATA +14 -13
  64. chunkr_ai-0.1.0a8.dist-info/RECORD +88 -0
  65. chunkr_ai/types/task.py +0 -1225
  66. chunkr_ai/types/tasks/parse_update_params.py +0 -845
  67. chunkr_ai-0.1.0a6.dist-info/RECORD +0 -52
  68. {chunkr_ai-0.1.0a6.dist-info → chunkr_ai-0.1.0a8.dist-info}/WHEEL +0 -0
  69. {chunkr_ai-0.1.0a6.dist-info → chunkr_ai-0.1.0a8.dist-info}/licenses/LICENSE +0 -0
chunkr_ai/__init__.py CHANGED
@@ -24,6 +24,7 @@ from ._exceptions import (
24
24
  InternalServerError,
25
25
  PermissionDeniedError,
26
26
  UnprocessableEntityError,
27
+ APIWebhookValidationError,
27
28
  APIResponseValidationError,
28
29
  )
29
30
  from ._base_client import DefaultHttpxClient, DefaultAioHttpClient, DefaultAsyncHttpxClient
@@ -45,6 +46,7 @@ __all__ = [
45
46
  "APITimeoutError",
46
47
  "APIConnectionError",
47
48
  "APIResponseValidationError",
49
+ "APIWebhookValidationError",
48
50
  "BadRequestError",
49
51
  "AuthenticationError",
50
52
  "PermissionDeniedError",
chunkr_ai/_base_client.py CHANGED
@@ -59,7 +59,7 @@ from ._types import (
59
59
  ModelBuilderProtocol,
60
60
  )
61
61
  from ._utils import is_dict, is_list, asyncify, is_given, lru_cache, is_mapping
62
- from ._compat import PYDANTIC_V2, model_copy, model_dump
62
+ from ._compat import PYDANTIC_V1, model_copy, model_dump
63
63
  from ._models import GenericModel, FinalRequestOptions, validate_type, construct_type
64
64
  from ._response import (
65
65
  APIResponse,
@@ -232,7 +232,7 @@ class BaseSyncPage(BasePage[_T], Generic[_T]):
232
232
  model: Type[_T],
233
233
  options: FinalRequestOptions,
234
234
  ) -> None:
235
- if PYDANTIC_V2 and getattr(self, "__pydantic_private__", None) is None:
235
+ if (not PYDANTIC_V1) and getattr(self, "__pydantic_private__", None) is None:
236
236
  self.__pydantic_private__ = {}
237
237
 
238
238
  self._model = model
@@ -320,7 +320,7 @@ class BaseAsyncPage(BasePage[_T], Generic[_T]):
320
320
  client: AsyncAPIClient,
321
321
  options: FinalRequestOptions,
322
322
  ) -> None:
323
- if PYDANTIC_V2 and getattr(self, "__pydantic_private__", None) is None:
323
+ if (not PYDANTIC_V1) and getattr(self, "__pydantic_private__", None) is None:
324
324
  self.__pydantic_private__ = {}
325
325
 
326
326
  self._model = model
chunkr_ai/_client.py CHANGED
@@ -21,7 +21,7 @@ from ._types import (
21
21
  )
22
22
  from ._utils import is_given, get_async_library
23
23
  from ._version import __version__
24
- from .resources import files, health
24
+ from .resources import files, health, webhooks
25
25
  from ._streaming import Stream as Stream, AsyncStream as AsyncStream
26
26
  from ._exceptions import ChunkrError, APIStatusError
27
27
  from ._base_client import (
@@ -38,16 +38,19 @@ class Chunkr(SyncAPIClient):
38
38
  tasks: tasks.TasksResource
39
39
  files: files.FilesResource
40
40
  health: health.HealthResource
41
+ webhooks: webhooks.WebhooksResource
41
42
  with_raw_response: ChunkrWithRawResponse
42
43
  with_streaming_response: ChunkrWithStreamedResponse
43
44
 
44
45
  # client options
45
46
  api_key: str
47
+ webhook_key: str | None
46
48
 
47
49
  def __init__(
48
50
  self,
49
51
  *,
50
52
  api_key: str | None = None,
53
+ webhook_key: str | None = None,
51
54
  base_url: str | httpx.URL | None = None,
52
55
  timeout: Union[float, Timeout, None, NotGiven] = NOT_GIVEN,
53
56
  max_retries: int = DEFAULT_MAX_RETRIES,
@@ -69,7 +72,9 @@ class Chunkr(SyncAPIClient):
69
72
  ) -> None:
70
73
  """Construct a new synchronous Chunkr client instance.
71
74
 
72
- This automatically infers the `api_key` argument from the `CHUNKR_API_KEY` environment variable if it is not provided.
75
+ This automatically infers the following arguments from their corresponding environment variables if they are not provided:
76
+ - `api_key` from `CHUNKR_API_KEY`
77
+ - `webhook_key` from `CHUNKR_WEBHOOK_KEY`
73
78
  """
74
79
  if api_key is None:
75
80
  api_key = os.environ.get("CHUNKR_API_KEY")
@@ -79,6 +84,10 @@ class Chunkr(SyncAPIClient):
79
84
  )
80
85
  self.api_key = api_key
81
86
 
87
+ if webhook_key is None:
88
+ webhook_key = os.environ.get("CHUNKR_WEBHOOK_KEY")
89
+ self.webhook_key = webhook_key
90
+
82
91
  if base_url is None:
83
92
  base_url = os.environ.get("CHUNKR_BASE_URL")
84
93
  if base_url is None:
@@ -100,6 +109,7 @@ class Chunkr(SyncAPIClient):
100
109
  self.tasks = tasks.TasksResource(self)
101
110
  self.files = files.FilesResource(self)
102
111
  self.health = health.HealthResource(self)
112
+ self.webhooks = webhooks.WebhooksResource(self)
103
113
  self.with_raw_response = ChunkrWithRawResponse(self)
104
114
  self.with_streaming_response = ChunkrWithStreamedResponse(self)
105
115
 
@@ -127,6 +137,7 @@ class Chunkr(SyncAPIClient):
127
137
  self,
128
138
  *,
129
139
  api_key: str | None = None,
140
+ webhook_key: str | None = None,
130
141
  base_url: str | httpx.URL | None = None,
131
142
  timeout: float | Timeout | None | NotGiven = NOT_GIVEN,
132
143
  http_client: httpx.Client | None = None,
@@ -161,6 +172,7 @@ class Chunkr(SyncAPIClient):
161
172
  http_client = http_client or self._client
162
173
  return self.__class__(
163
174
  api_key=api_key or self.api_key,
175
+ webhook_key=webhook_key or self.webhook_key,
164
176
  base_url=base_url or self.base_url,
165
177
  timeout=self.timeout if isinstance(timeout, NotGiven) else timeout,
166
178
  http_client=http_client,
@@ -212,16 +224,19 @@ class AsyncChunkr(AsyncAPIClient):
212
224
  tasks: tasks.AsyncTasksResource
213
225
  files: files.AsyncFilesResource
214
226
  health: health.AsyncHealthResource
227
+ webhooks: webhooks.AsyncWebhooksResource
215
228
  with_raw_response: AsyncChunkrWithRawResponse
216
229
  with_streaming_response: AsyncChunkrWithStreamedResponse
217
230
 
218
231
  # client options
219
232
  api_key: str
233
+ webhook_key: str | None
220
234
 
221
235
  def __init__(
222
236
  self,
223
237
  *,
224
238
  api_key: str | None = None,
239
+ webhook_key: str | None = None,
225
240
  base_url: str | httpx.URL | None = None,
226
241
  timeout: Union[float, Timeout, None, NotGiven] = NOT_GIVEN,
227
242
  max_retries: int = DEFAULT_MAX_RETRIES,
@@ -243,7 +258,9 @@ class AsyncChunkr(AsyncAPIClient):
243
258
  ) -> None:
244
259
  """Construct a new async AsyncChunkr client instance.
245
260
 
246
- This automatically infers the `api_key` argument from the `CHUNKR_API_KEY` environment variable if it is not provided.
261
+ This automatically infers the following arguments from their corresponding environment variables if they are not provided:
262
+ - `api_key` from `CHUNKR_API_KEY`
263
+ - `webhook_key` from `CHUNKR_WEBHOOK_KEY`
247
264
  """
248
265
  if api_key is None:
249
266
  api_key = os.environ.get("CHUNKR_API_KEY")
@@ -253,6 +270,10 @@ class AsyncChunkr(AsyncAPIClient):
253
270
  )
254
271
  self.api_key = api_key
255
272
 
273
+ if webhook_key is None:
274
+ webhook_key = os.environ.get("CHUNKR_WEBHOOK_KEY")
275
+ self.webhook_key = webhook_key
276
+
256
277
  if base_url is None:
257
278
  base_url = os.environ.get("CHUNKR_BASE_URL")
258
279
  if base_url is None:
@@ -274,6 +295,7 @@ class AsyncChunkr(AsyncAPIClient):
274
295
  self.tasks = tasks.AsyncTasksResource(self)
275
296
  self.files = files.AsyncFilesResource(self)
276
297
  self.health = health.AsyncHealthResource(self)
298
+ self.webhooks = webhooks.AsyncWebhooksResource(self)
277
299
  self.with_raw_response = AsyncChunkrWithRawResponse(self)
278
300
  self.with_streaming_response = AsyncChunkrWithStreamedResponse(self)
279
301
 
@@ -301,6 +323,7 @@ class AsyncChunkr(AsyncAPIClient):
301
323
  self,
302
324
  *,
303
325
  api_key: str | None = None,
326
+ webhook_key: str | None = None,
304
327
  base_url: str | httpx.URL | None = None,
305
328
  timeout: float | Timeout | None | NotGiven = NOT_GIVEN,
306
329
  http_client: httpx.AsyncClient | None = None,
@@ -335,6 +358,7 @@ class AsyncChunkr(AsyncAPIClient):
335
358
  http_client = http_client or self._client
336
359
  return self.__class__(
337
360
  api_key=api_key or self.api_key,
361
+ webhook_key=webhook_key or self.webhook_key,
338
362
  base_url=base_url or self.base_url,
339
363
  timeout=self.timeout if isinstance(timeout, NotGiven) else timeout,
340
364
  http_client=http_client,
@@ -387,6 +411,7 @@ class ChunkrWithRawResponse:
387
411
  self.tasks = tasks.TasksResourceWithRawResponse(client.tasks)
388
412
  self.files = files.FilesResourceWithRawResponse(client.files)
389
413
  self.health = health.HealthResourceWithRawResponse(client.health)
414
+ self.webhooks = webhooks.WebhooksResourceWithRawResponse(client.webhooks)
390
415
 
391
416
 
392
417
  class AsyncChunkrWithRawResponse:
@@ -394,6 +419,7 @@ class AsyncChunkrWithRawResponse:
394
419
  self.tasks = tasks.AsyncTasksResourceWithRawResponse(client.tasks)
395
420
  self.files = files.AsyncFilesResourceWithRawResponse(client.files)
396
421
  self.health = health.AsyncHealthResourceWithRawResponse(client.health)
422
+ self.webhooks = webhooks.AsyncWebhooksResourceWithRawResponse(client.webhooks)
397
423
 
398
424
 
399
425
  class ChunkrWithStreamedResponse:
@@ -401,6 +427,7 @@ class ChunkrWithStreamedResponse:
401
427
  self.tasks = tasks.TasksResourceWithStreamingResponse(client.tasks)
402
428
  self.files = files.FilesResourceWithStreamingResponse(client.files)
403
429
  self.health = health.HealthResourceWithStreamingResponse(client.health)
430
+ self.webhooks = webhooks.WebhooksResourceWithStreamingResponse(client.webhooks)
404
431
 
405
432
 
406
433
  class AsyncChunkrWithStreamedResponse:
@@ -408,6 +435,7 @@ class AsyncChunkrWithStreamedResponse:
408
435
  self.tasks = tasks.AsyncTasksResourceWithStreamingResponse(client.tasks)
409
436
  self.files = files.AsyncFilesResourceWithStreamingResponse(client.files)
410
437
  self.health = health.AsyncHealthResourceWithStreamingResponse(client.health)
438
+ self.webhooks = webhooks.AsyncWebhooksResourceWithStreamingResponse(client.webhooks)
411
439
 
412
440
 
413
441
  Client = Chunkr
chunkr_ai/_compat.py CHANGED
@@ -12,14 +12,13 @@ from ._types import IncEx, StrBytesIntFloat
12
12
  _T = TypeVar("_T")
13
13
  _ModelT = TypeVar("_ModelT", bound=pydantic.BaseModel)
14
14
 
15
- # --------------- Pydantic v2 compatibility ---------------
15
+ # --------------- Pydantic v2, v3 compatibility ---------------
16
16
 
17
17
  # Pyright incorrectly reports some of our functions as overriding a method when they don't
18
18
  # pyright: reportIncompatibleMethodOverride=false
19
19
 
20
- PYDANTIC_V2 = pydantic.VERSION.startswith("2.")
20
+ PYDANTIC_V1 = pydantic.VERSION.startswith("1.")
21
21
 
22
- # v1 re-exports
23
22
  if TYPE_CHECKING:
24
23
 
25
24
  def parse_date(value: date | StrBytesIntFloat) -> date: # noqa: ARG001
@@ -44,90 +43,92 @@ if TYPE_CHECKING:
44
43
  ...
45
44
 
46
45
  else:
47
- if PYDANTIC_V2:
48
- from pydantic.v1.typing import (
46
+ # v1 re-exports
47
+ if PYDANTIC_V1:
48
+ from pydantic.typing import (
49
49
  get_args as get_args,
50
50
  is_union as is_union,
51
51
  get_origin as get_origin,
52
52
  is_typeddict as is_typeddict,
53
53
  is_literal_type as is_literal_type,
54
54
  )
55
- from pydantic.v1.datetime_parse import parse_date as parse_date, parse_datetime as parse_datetime
55
+ from pydantic.datetime_parse import parse_date as parse_date, parse_datetime as parse_datetime
56
56
  else:
57
- from pydantic.typing import (
57
+ from ._utils import (
58
58
  get_args as get_args,
59
59
  is_union as is_union,
60
60
  get_origin as get_origin,
61
+ parse_date as parse_date,
61
62
  is_typeddict as is_typeddict,
63
+ parse_datetime as parse_datetime,
62
64
  is_literal_type as is_literal_type,
63
65
  )
64
- from pydantic.datetime_parse import parse_date as parse_date, parse_datetime as parse_datetime
65
66
 
66
67
 
67
68
  # refactored config
68
69
  if TYPE_CHECKING:
69
70
  from pydantic import ConfigDict as ConfigDict
70
71
  else:
71
- if PYDANTIC_V2:
72
- from pydantic import ConfigDict
73
- else:
72
+ if PYDANTIC_V1:
74
73
  # TODO: provide an error message here?
75
74
  ConfigDict = None
75
+ else:
76
+ from pydantic import ConfigDict as ConfigDict
76
77
 
77
78
 
78
79
  # renamed methods / properties
79
80
  def parse_obj(model: type[_ModelT], value: object) -> _ModelT:
80
- if PYDANTIC_V2:
81
- return model.model_validate(value)
82
- else:
81
+ if PYDANTIC_V1:
83
82
  return cast(_ModelT, model.parse_obj(value)) # pyright: ignore[reportDeprecated, reportUnnecessaryCast]
83
+ else:
84
+ return model.model_validate(value)
84
85
 
85
86
 
86
87
  def field_is_required(field: FieldInfo) -> bool:
87
- if PYDANTIC_V2:
88
- return field.is_required()
89
- return field.required # type: ignore
88
+ if PYDANTIC_V1:
89
+ return field.required # type: ignore
90
+ return field.is_required()
90
91
 
91
92
 
92
93
  def field_get_default(field: FieldInfo) -> Any:
93
94
  value = field.get_default()
94
- if PYDANTIC_V2:
95
- from pydantic_core import PydanticUndefined
96
-
97
- if value == PydanticUndefined:
98
- return None
95
+ if PYDANTIC_V1:
99
96
  return value
97
+ from pydantic_core import PydanticUndefined
98
+
99
+ if value == PydanticUndefined:
100
+ return None
100
101
  return value
101
102
 
102
103
 
103
104
  def field_outer_type(field: FieldInfo) -> Any:
104
- if PYDANTIC_V2:
105
- return field.annotation
106
- return field.outer_type_ # type: ignore
105
+ if PYDANTIC_V1:
106
+ return field.outer_type_ # type: ignore
107
+ return field.annotation
107
108
 
108
109
 
109
110
  def get_model_config(model: type[pydantic.BaseModel]) -> Any:
110
- if PYDANTIC_V2:
111
- return model.model_config
112
- return model.__config__ # type: ignore
111
+ if PYDANTIC_V1:
112
+ return model.__config__ # type: ignore
113
+ return model.model_config
113
114
 
114
115
 
115
116
  def get_model_fields(model: type[pydantic.BaseModel]) -> dict[str, FieldInfo]:
116
- if PYDANTIC_V2:
117
- return model.model_fields
118
- return model.__fields__ # type: ignore
117
+ if PYDANTIC_V1:
118
+ return model.__fields__ # type: ignore
119
+ return model.model_fields
119
120
 
120
121
 
121
122
  def model_copy(model: _ModelT, *, deep: bool = False) -> _ModelT:
122
- if PYDANTIC_V2:
123
- return model.model_copy(deep=deep)
124
- return model.copy(deep=deep) # type: ignore
123
+ if PYDANTIC_V1:
124
+ return model.copy(deep=deep) # type: ignore
125
+ return model.model_copy(deep=deep)
125
126
 
126
127
 
127
128
  def model_json(model: pydantic.BaseModel, *, indent: int | None = None) -> str:
128
- if PYDANTIC_V2:
129
- return model.model_dump_json(indent=indent)
130
- return model.json(indent=indent) # type: ignore
129
+ if PYDANTIC_V1:
130
+ return model.json(indent=indent) # type: ignore
131
+ return model.model_dump_json(indent=indent)
131
132
 
132
133
 
133
134
  def model_dump(
@@ -139,14 +140,14 @@ def model_dump(
139
140
  warnings: bool = True,
140
141
  mode: Literal["json", "python"] = "python",
141
142
  ) -> dict[str, Any]:
142
- if PYDANTIC_V2 or hasattr(model, "model_dump"):
143
+ if (not PYDANTIC_V1) or hasattr(model, "model_dump"):
143
144
  return model.model_dump(
144
145
  mode=mode,
145
146
  exclude=exclude,
146
147
  exclude_unset=exclude_unset,
147
148
  exclude_defaults=exclude_defaults,
148
149
  # warnings are not supported in Pydantic v1
149
- warnings=warnings if PYDANTIC_V2 else True,
150
+ warnings=True if PYDANTIC_V1 else warnings,
150
151
  )
151
152
  return cast(
152
153
  "dict[str, Any]",
@@ -159,9 +160,9 @@ def model_dump(
159
160
 
160
161
 
161
162
  def model_parse(model: type[_ModelT], data: Any) -> _ModelT:
162
- if PYDANTIC_V2:
163
- return model.model_validate(data)
164
- return model.parse_obj(data) # pyright: ignore[reportDeprecated]
163
+ if PYDANTIC_V1:
164
+ return model.parse_obj(data) # pyright: ignore[reportDeprecated]
165
+ return model.model_validate(data)
165
166
 
166
167
 
167
168
  # generic models
@@ -170,17 +171,16 @@ if TYPE_CHECKING:
170
171
  class GenericModel(pydantic.BaseModel): ...
171
172
 
172
173
  else:
173
- if PYDANTIC_V2:
174
+ if PYDANTIC_V1:
175
+ import pydantic.generics
176
+
177
+ class GenericModel(pydantic.generics.GenericModel, pydantic.BaseModel): ...
178
+ else:
174
179
  # there no longer needs to be a distinction in v2 but
175
180
  # we still have to create our own subclass to avoid
176
181
  # inconsistent MRO ordering errors
177
182
  class GenericModel(pydantic.BaseModel): ...
178
183
 
179
- else:
180
- import pydantic.generics
181
-
182
- class GenericModel(pydantic.generics.GenericModel, pydantic.BaseModel): ...
183
-
184
184
 
185
185
  # cached properties
186
186
  if TYPE_CHECKING:
chunkr_ai/_constants.py CHANGED
@@ -5,10 +5,10 @@ import httpx
5
5
  RAW_RESPONSE_HEADER = "X-Stainless-Raw-Response"
6
6
  OVERRIDE_CAST_TO_HEADER = "____stainless_override_cast_to"
7
7
 
8
- # default timeout is 30 seconds
9
- DEFAULT_TIMEOUT = httpx.Timeout(timeout=30, connect=5.0)
10
- DEFAULT_MAX_RETRIES = 50
8
+ # default timeout is 1 minute
9
+ DEFAULT_TIMEOUT = httpx.Timeout(timeout=60, connect=5.0)
10
+ DEFAULT_MAX_RETRIES = 2
11
11
  DEFAULT_CONNECTION_LIMITS = httpx.Limits(max_connections=100, max_keepalive_connections=20)
12
12
 
13
- INITIAL_RETRY_DELAY = 1.0
14
- MAX_RETRY_DELAY = 10.0
13
+ INITIAL_RETRY_DELAY = 0.5
14
+ MAX_RETRY_DELAY = 8.0
chunkr_ai/_exceptions.py CHANGED
@@ -54,6 +54,10 @@ class APIResponseValidationError(APIError):
54
54
  self.status_code = response.status_code
55
55
 
56
56
 
57
+ class APIWebhookValidationError(APIError):
58
+ pass
59
+
60
+
57
61
  class APIStatusError(APIError):
58
62
  """Raised when an API response has a status code of 4xx or 5xx."""
59
63
 
chunkr_ai/_models.py CHANGED
@@ -50,7 +50,7 @@ from ._utils import (
50
50
  strip_annotated_type,
51
51
  )
52
52
  from ._compat import (
53
- PYDANTIC_V2,
53
+ PYDANTIC_V1,
54
54
  ConfigDict,
55
55
  GenericModel as BaseGenericModel,
56
56
  get_args,
@@ -81,11 +81,7 @@ class _ConfigProtocol(Protocol):
81
81
 
82
82
 
83
83
  class BaseModel(pydantic.BaseModel):
84
- if PYDANTIC_V2:
85
- model_config: ClassVar[ConfigDict] = ConfigDict(
86
- extra="allow", defer_build=coerce_boolean(os.environ.get("DEFER_PYDANTIC_BUILD", "true"))
87
- )
88
- else:
84
+ if PYDANTIC_V1:
89
85
 
90
86
  @property
91
87
  @override
@@ -95,6 +91,10 @@ class BaseModel(pydantic.BaseModel):
95
91
 
96
92
  class Config(pydantic.BaseConfig): # pyright: ignore[reportDeprecated]
97
93
  extra: Any = pydantic.Extra.allow # type: ignore
94
+ else:
95
+ model_config: ClassVar[ConfigDict] = ConfigDict(
96
+ extra="allow", defer_build=coerce_boolean(os.environ.get("DEFER_PYDANTIC_BUILD", "true"))
97
+ )
98
98
 
99
99
  def to_dict(
100
100
  self,
@@ -215,25 +215,25 @@ class BaseModel(pydantic.BaseModel):
215
215
  if key not in model_fields:
216
216
  parsed = construct_type(value=value, type_=extra_field_type) if extra_field_type is not None else value
217
217
 
218
- if PYDANTIC_V2:
219
- _extra[key] = parsed
220
- else:
218
+ if PYDANTIC_V1:
221
219
  _fields_set.add(key)
222
220
  fields_values[key] = parsed
221
+ else:
222
+ _extra[key] = parsed
223
223
 
224
224
  object.__setattr__(m, "__dict__", fields_values)
225
225
 
226
- if PYDANTIC_V2:
227
- # these properties are copied from Pydantic's `model_construct()` method
228
- object.__setattr__(m, "__pydantic_private__", None)
229
- object.__setattr__(m, "__pydantic_extra__", _extra)
230
- object.__setattr__(m, "__pydantic_fields_set__", _fields_set)
231
- else:
226
+ if PYDANTIC_V1:
232
227
  # init_private_attributes() does not exist in v2
233
228
  m._init_private_attributes() # type: ignore
234
229
 
235
230
  # copied from Pydantic v1's `construct()` method
236
231
  object.__setattr__(m, "__fields_set__", _fields_set)
232
+ else:
233
+ # these properties are copied from Pydantic's `model_construct()` method
234
+ object.__setattr__(m, "__pydantic_private__", None)
235
+ object.__setattr__(m, "__pydantic_extra__", _extra)
236
+ object.__setattr__(m, "__pydantic_fields_set__", _fields_set)
237
237
 
238
238
  return m
239
239
 
@@ -243,7 +243,7 @@ class BaseModel(pydantic.BaseModel):
243
243
  # although not in practice
244
244
  model_construct = construct
245
245
 
246
- if not PYDANTIC_V2:
246
+ if PYDANTIC_V1:
247
247
  # we define aliases for some of the new pydantic v2 methods so
248
248
  # that we can just document these methods without having to specify
249
249
  # a specific pydantic version as some users may not know which
@@ -304,7 +304,7 @@ class BaseModel(pydantic.BaseModel):
304
304
  exclude_none=exclude_none,
305
305
  )
306
306
 
307
- return cast(dict[str, Any], json_safe(dumped)) if mode == "json" else dumped
307
+ return cast("dict[str, Any]", json_safe(dumped)) if mode == "json" else dumped
308
308
 
309
309
  @override
310
310
  def model_dump_json(
@@ -363,10 +363,10 @@ def _construct_field(value: object, field: FieldInfo, key: str) -> object:
363
363
  if value is None:
364
364
  return field_get_default(field)
365
365
 
366
- if PYDANTIC_V2:
367
- type_ = field.annotation
368
- else:
366
+ if PYDANTIC_V1:
369
367
  type_ = cast(type, field.outer_type_) # type: ignore
368
+ else:
369
+ type_ = field.annotation # type: ignore
370
370
 
371
371
  if type_ is None:
372
372
  raise RuntimeError(f"Unexpected field type is None for {key}")
@@ -375,7 +375,7 @@ def _construct_field(value: object, field: FieldInfo, key: str) -> object:
375
375
 
376
376
 
377
377
  def _get_extra_fields_type(cls: type[pydantic.BaseModel]) -> type | None:
378
- if not PYDANTIC_V2:
378
+ if PYDANTIC_V1:
379
379
  # TODO
380
380
  return None
381
381
 
@@ -628,30 +628,30 @@ def _build_discriminated_union_meta(*, union: type, meta_annotations: tuple[Any,
628
628
  for variant in get_args(union):
629
629
  variant = strip_annotated_type(variant)
630
630
  if is_basemodel_type(variant):
631
- if PYDANTIC_V2:
632
- field = _extract_field_schema_pv2(variant, discriminator_field_name)
633
- if not field:
631
+ if PYDANTIC_V1:
632
+ field_info = cast("dict[str, FieldInfo]", variant.__fields__).get(discriminator_field_name) # pyright: ignore[reportDeprecated, reportUnnecessaryCast]
633
+ if not field_info:
634
634
  continue
635
635
 
636
636
  # Note: if one variant defines an alias then they all should
637
- discriminator_alias = field.get("serialization_alias")
638
-
639
- field_schema = field["schema"]
637
+ discriminator_alias = field_info.alias
640
638
 
641
- if field_schema["type"] == "literal":
642
- for entry in cast("LiteralSchema", field_schema)["expected"]:
639
+ if (annotation := getattr(field_info, "annotation", None)) and is_literal_type(annotation):
640
+ for entry in get_args(annotation):
643
641
  if isinstance(entry, str):
644
642
  mapping[entry] = variant
645
643
  else:
646
- field_info = cast("dict[str, FieldInfo]", variant.__fields__).get(discriminator_field_name) # pyright: ignore[reportDeprecated, reportUnnecessaryCast]
647
- if not field_info:
644
+ field = _extract_field_schema_pv2(variant, discriminator_field_name)
645
+ if not field:
648
646
  continue
649
647
 
650
648
  # Note: if one variant defines an alias then they all should
651
- discriminator_alias = field_info.alias
649
+ discriminator_alias = field.get("serialization_alias")
652
650
 
653
- if (annotation := getattr(field_info, "annotation", None)) and is_literal_type(annotation):
654
- for entry in get_args(annotation):
651
+ field_schema = field["schema"]
652
+
653
+ if field_schema["type"] == "literal":
654
+ for entry in cast("LiteralSchema", field_schema)["expected"]:
655
655
  if isinstance(entry, str):
656
656
  mapping[entry] = variant
657
657
 
@@ -714,7 +714,7 @@ else:
714
714
  pass
715
715
 
716
716
 
717
- if PYDANTIC_V2:
717
+ if not PYDANTIC_V1:
718
718
  from pydantic import TypeAdapter as _TypeAdapter
719
719
 
720
720
  _CachedTypeAdapter = cast("TypeAdapter[object]", lru_cache(maxsize=None)(_TypeAdapter))
@@ -782,12 +782,12 @@ class FinalRequestOptions(pydantic.BaseModel):
782
782
  json_data: Union[Body, None] = None
783
783
  extra_json: Union[AnyMapping, None] = None
784
784
 
785
- if PYDANTIC_V2:
786
- model_config: ClassVar[ConfigDict] = ConfigDict(arbitrary_types_allowed=True)
787
- else:
785
+ if PYDANTIC_V1:
788
786
 
789
787
  class Config(pydantic.BaseConfig): # pyright: ignore[reportDeprecated]
790
788
  arbitrary_types_allowed: bool = True
789
+ else:
790
+ model_config: ClassVar[ConfigDict] = ConfigDict(arbitrary_types_allowed=True)
791
791
 
792
792
  def get_max_retries(self, max_retries: int) -> int:
793
793
  if isinstance(self.max_retries, NotGiven):
@@ -820,9 +820,9 @@ class FinalRequestOptions(pydantic.BaseModel):
820
820
  key: strip_not_given(value)
821
821
  for key, value in values.items()
822
822
  }
823
- if PYDANTIC_V2:
824
- return super().model_construct(_fields_set, **kwargs)
825
- return cast(FinalRequestOptions, super().construct(_fields_set, **kwargs)) # pyright: ignore[reportDeprecated]
823
+ if PYDANTIC_V1:
824
+ return cast(FinalRequestOptions, super().construct(_fields_set, **kwargs)) # pyright: ignore[reportDeprecated]
825
+ return super().model_construct(_fields_set, **kwargs)
826
826
 
827
827
  if not TYPE_CHECKING:
828
828
  # type checkers incorrectly complain about this assignment
chunkr_ai/_types.py CHANGED
@@ -13,10 +13,21 @@ from typing import (
13
13
  Mapping,
14
14
  TypeVar,
15
15
  Callable,
16
+ Iterator,
16
17
  Optional,
17
18
  Sequence,
18
19
  )
19
- from typing_extensions import Set, Literal, Protocol, TypeAlias, TypedDict, override, runtime_checkable
20
+ from typing_extensions import (
21
+ Set,
22
+ Literal,
23
+ Protocol,
24
+ TypeAlias,
25
+ TypedDict,
26
+ SupportsIndex,
27
+ overload,
28
+ override,
29
+ runtime_checkable,
30
+ )
20
31
 
21
32
  import httpx
22
33
  import pydantic
@@ -217,3 +228,26 @@ class _GenericAlias(Protocol):
217
228
  class HttpxSendArgs(TypedDict, total=False):
218
229
  auth: httpx.Auth
219
230
  follow_redirects: bool
231
+
232
+
233
+ _T_co = TypeVar("_T_co", covariant=True)
234
+
235
+
236
+ if TYPE_CHECKING:
237
+ # This works because str.__contains__ does not accept object (either in typeshed or at runtime)
238
+ # https://github.com/hauntsaninja/useful_types/blob/5e9710f3875107d068e7679fd7fec9cfab0eff3b/useful_types/__init__.py#L285
239
+ class SequenceNotStr(Protocol[_T_co]):
240
+ @overload
241
+ def __getitem__(self, index: SupportsIndex, /) -> _T_co: ...
242
+ @overload
243
+ def __getitem__(self, index: slice, /) -> Sequence[_T_co]: ...
244
+ def __contains__(self, value: object, /) -> bool: ...
245
+ def __len__(self) -> int: ...
246
+ def __iter__(self) -> Iterator[_T_co]: ...
247
+ def index(self, value: Any, start: int = 0, stop: int = ..., /) -> int: ...
248
+ def count(self, value: Any, /) -> int: ...
249
+ def __reversed__(self) -> Iterator[_T_co]: ...
250
+ else:
251
+ # just point this to a normal `Sequence` at runtime to avoid having to special case
252
+ # deserializing our custom sequence type
253
+ SequenceNotStr = Sequence