feat: file parsing (#102)

Co-authored-by: mengqian <[email protected]>

Author: code-c-light

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "zhipuai"
-version = "2.1.5.20250801"
+version = "2.1.5.20250825"
 description = "A SDK library for accessing big model apis from ZhipuAI"
 authors = ["Zhipu AI"]
 readme = "README.md"

tests/integration_tests/test_file_parser.py

@@ -0,0 +1,40 @@
+from __future__ import annotations
+
+import logging
+import logging.config
+import os
+
+import pytest
+
+import zhipuai
+from zhipuai import ZhipuAI
+
+
+def test_file_parser_create(logging_conf):
+	logging.config.dictConfig(logging_conf)  # type: ignore
+	client = ZhipuAI()  # 填写您自己的APIKey
+	try:
+		response = client.file_parser.create(file=open('hitsuyoushorui-cn.pdf', 'rb'), file_type='pdf', tool_type='zhipu_pro')
+		print(response)
+
+	except zhipuai.core._errors.APIRequestFailedError as err:
+		print(err)
+	except zhipuai.core._errors.APIInternalError as err:
+		print(err)
+	except zhipuai.core._errors.APIStatusError as err:
+		print(err)
+
+def test_file_parser_content(logging_conf):
+	logging.config.dictConfig(logging_conf)  # type: ignore
+	client = ZhipuAI()  # 填写您自己的APIKey
+	try:
+		response = client.file_parser.content(task_id="66e8f7ab884448c8b4190f251f6c2982-1", format_type="text")
+		print(response.content.decode('utf-8'))
+
+	except zhipuai.core._errors.APIRequestFailedError as err:
+		print(err)
+	except zhipuai.core._errors.APIInternalError as err:
+		print(err)
+	except zhipuai.core._errors.APIStatusError as err:
+		print(err)
+

zhipuai/_client.py

@@ -69,6 +69,7 @@ def __init__(
         self.audio = api_resource.audio.Audio(self)
         self.moderations = api_resource.moderation.Moderations(self)
         self.agents = api_resource.agents.Agents(self)
+        self.file_parser = api_resource.file_parser.FileParser(self)
 
     @property
     @override

zhipuai/api_resource/__init__.py

@@ -53,6 +53,10 @@
     Audio
 )
 
+from .file_parser import (
+    FileParser
+)
+
 __all__ = [
     'Videos',
     'AsyncCompletions',
@@ -68,5 +72,6 @@
     'Tools',
     'Assistant',
     'Audio',
-    'Moderation'
+    'Moderation',
+    'FileParser'
 ]

zhipuai/api_resource/file_parser/__init__.py

@@ -0,0 +1,3 @@
+from .file_parser import FileParser
+
+__all__ = ['FileParser']

zhipuai/api_resource/file_parser/file_parser.py

@@ -0,0 +1,95 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Mapping, cast
+
+import httpx
+from typing_extensions import Literal
+
+from ...core import BaseAPI, maybe_transform
+from ...core import NOT_GIVEN, Body, Headers, NotGiven, FileTypes
+from ...core import _legacy_binary_response
+from ...core import _legacy_response
+from ...core import deepcopy_minimal, extract_files
+from ...core import (
+    make_request_options,
+)
+from ...types.file_parser.file_parser_create_params import FileParserCreateParams
+from ...types.file_parser.file_parser_resp import FileParserTaskCreateResp
+
+if TYPE_CHECKING:
+    from ..._client import ZhipuAI
+
+__all__ = ["FileParser"]
+
+
+class FileParser(BaseAPI):
+
+    def __init__(self, client: "ZhipuAI") -> None:
+        super().__init__(client)
+
+    def create(
+            self,
+            *,
+            file: FileTypes = None,
+            file_type: str = None,
+            tool_type: Literal["simple", "doc2x", "tencent", "zhipu-pro"],
+            extra_headers: Headers | None = None,
+            extra_body: Body | None = None,
+            timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> FileParserTaskCreateResp:
+
+        if not file:
+            raise ValueError("At least one `file` must be provided.")
+        body = deepcopy_minimal(
+            {
+                "file": file,
+                "file_type": file_type,
+                "tool_type": tool_type,
+            }
+        )
+
+        files = extract_files(cast(Mapping[str, object], body), paths=[["file"]])
+        if files:
+            # It should be noted that the actual Content-Type header that will be
+            # sent to the server will contain a `boundary` parameter, e.g.
+            # multipart/form-data; boundary=---abc--
+            extra_headers = {"Content-Type": "multipart/form-data", **(extra_headers or {})}
+        return self._post(
+            "/files/parser/create",
+            body=maybe_transform(body, FileParserCreateParams),
+            files=files,
+            options=make_request_options(
+                extra_headers=extra_headers, extra_body=extra_body, timeout=timeout
+            ),
+            cast_type=FileParserTaskCreateResp,
+        )
+
+    def content(
+            self,
+            task_id: str,
+            *,
+            format_type: Literal["text", "download_link"],
+            extra_headers: Headers | None = None,
+            extra_body: Body | None = None,
+            timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> _legacy_response.HttpxBinaryResponseContent:
+        """
+        Returns the contents of the specified file.
+
+        Args:
+          extra_headers: Send extra headers
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        if not task_id:
+            raise ValueError(f"Expected a non-empty value for `task_id` but received {task_id!r}")
+        extra_headers = {"Accept": "application/binary", **(extra_headers or {})}
+        return self._get(
+            f"/files/parser/result/{task_id}/{format_type}",
+            options=make_request_options(
+                extra_headers=extra_headers, extra_body=extra_body, timeout=timeout
+            ),
+            cast_type=_legacy_binary_response.HttpxBinaryResponseContent,
+        )

zhipuai/types/file_parser/__init__.py

@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+from typing_extensions import Literal, Required, TypedDict
+from ...core import NOT_GIVEN, Body, Headers, NotGiven, FileTypes
+
+
+__all__ = ["FileParserCreateParams", "FileParserDownloadParams"]
+
+
+class FileParserCreateParams(TypedDict):
+    file: FileTypes
+    """上传的文件"""
+    file_type: str
+    """文件类型"""
+    tool_type: Literal["simple", "doc2x", "tencent", "zhipu-pro"]
+    """工具类型"""
+
+
+class FileParserDownloadParams(TypedDict):
+    task_id: str
+    """解析任务id"""
+    format_type: Literal["text", "download_link"]
+    """结果返回类型"""
+

zhipuai/types/file_parser/file_parser_create_params.py

@@ -0,0 +1,18 @@
+from typing import List, Optional
+
+from zhipuai.core import BaseModel
+
+__all__ = [
+    "FileParserTaskCreateResp"
+]
+
+
+class FileParserTaskCreateResp(BaseModel):
+    task_id: str
+    # 任务id
+    message: str
+    # message
+    success: bool
+    # 是否成功
+
+