feature: Introduce default_encoding parameter to set/autodetect the encoding if the charset is missing from the headers (lexiforest#284)

* Add a `default_encoding` parameter to [set|autodetect] the encoding if no charset is found in the headers

* Update github workflow actions/*

Author: deedy5

.github/workflows/build-and-test.yaml

@@ -19,8 +19,8 @@ jobs:
     name: Lint
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
         with:
           python-version: '3.10'
       - run: |
@@ -31,15 +31,15 @@ jobs:
     name: Build sdist wheel
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - run: |
           make preprocess
           pipx run build --sdist
-      - uses: actions/upload-artifact@v3
+      - uses: actions/upload-artifact@v3  # https://github.com/actions/upload-artifact/issues/478
         with:
           path: ./dist/*.tar.gz
 
-      - uses: actions/setup-python@v4
+      - uses: actions/setup-python@v5
         with:
           python-version: '3.10'
       - run: |
@@ -52,51 +52,50 @@ jobs:
       matrix:
         os: [ubuntu-22.04, macos-12, macos-14, windows-2019]
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
 
-      - uses: actions/setup-python@v4
+      - uses: actions/setup-python@v5
         with:
           python-version: '3.10'
 
       - if: runner.os == 'Linux'
-        uses: docker/setup-qemu-action@v2
+        uses: docker/setup-qemu-action@v3
         with:
           platforms: all
 
       # macOS make is too old
       - if: runner.os == 'macOS'
         run: |
           brew install make automake libtool
-          which pipx || brew install pipx && pipx ensurepath
 
       - name: Build and test wheels
-        uses: pypa/cibuildwheel@v2.16.5
+        uses: pypa/cibuildwheel@v2.17.0
 
       # - name: Setup tmate session
       #   uses: mxschmitt/action-tmate@v3
 
-      - uses: actions/upload-artifact@v3
+      - uses: actions/upload-artifact@v3  # https://github.com/actions/upload-artifact/issues/478
         with:
           path: ./wheelhouse/*.whl
 
   upload_all:
     needs: [bdist, sdist]
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/download-artifact@v3
+    - uses: actions/download-artifact@v3  # https://github.com/actions/upload-artifact/issues/478
       if: startsWith(github.ref, 'refs/tags/')
       with:
         name: artifact
         path: dist
 
-    - uses: pypa/gh-action-pypi-publish@v1.5.0
+    - uses: pypa/gh-action-pypi-publish@v1.8.14
       if: startsWith(github.ref, 'refs/tags/')
       with:
         password: ${{ secrets.PYPI_TOKEN }}
 
     - name: Upload release files
       if: startsWith(github.ref, 'refs/tags/')
-      uses: softprops/action-gh-release@v1
+      uses: softprops/action-gh-release@v2
       with:
         files: |
           ./dist/*.whl

curl_cffi/requests/__init__.py

@@ -58,6 +58,7 @@ def request(
     impersonate: Optional[Union[str, BrowserType]] = None,
     thread: Optional[ThreadType] = None,
     default_headers: Optional[bool] = None,
+    default_encoding: Union[str, Callable[[bytes], str]] = "utf-8",
     curl_options: Optional[dict] = None,
     http_version: Optional[CurlHttpVersion] = None,
     debug: bool = False,
@@ -90,6 +91,8 @@ def request(
         impersonate: which browser version to impersonate.
         thread: work with other thread implementations. choices: eventlet, gevent.
         default_headers: whether to set default browser headers.
+        default_encoding: encoding for decoding response content if charset is not found in headers. 
+                Defaults to "utf-8". Can be set to a callable for automatic detection.
         curl_options: extra curl options to use.
         http_version: limiting http version, http2 will be tries by default.
         debug: print extra curl debug info.
@@ -122,6 +125,7 @@ def request(
             content_callback=content_callback,
             impersonate=impersonate,
             default_headers=default_headers,
+            default_encoding=default_encoding,
             http_version=http_version,
             interface=interface,
             multipart=multipart,

curl_cffi/requests/models.py

@@ -1,14 +1,18 @@
 import queue
+import re
 import warnings
 from concurrent.futures import Future
+from functools import cached_property
 from json import loads
-from typing import Any, Awaitable, Dict, List, Optional
+from typing import Any, Awaitable, Callable, Dict, List, Optional, Union
 
 from .. import Curl
 from .cookies import Cookies
 from .errors import RequestsError
 from .headers import Headers
 
+CHARSET_RE = re.compile(r"charset=([\w-]+)")
+
 
 def clear_queue(q: queue.Queue):
     with q.mutex:
@@ -41,6 +45,8 @@ class Response:
         elapsed: how many seconds the request cost.
         encoding: http body encoding.
         charset: alias for encoding.
+        charset_encoding: encoding specified by the Content-Type header.
+        default_encoding: user-defined encoding used for decoding content if charset is not found in headers.
         redirect_count: how many redirects happened.
         redirect_url: the final redirected url.
         http_version: http version used.
@@ -58,8 +64,7 @@ def __init__(self, curl: Optional[Curl] = None, request: Optional[Request] = Non
         self.headers = Headers()
         self.cookies = Cookies()
         self.elapsed = 0.0
-        self.encoding = "utf-8"
-        self.charset = self.encoding
+        self.default_encoding: Union[str, Callable[[bytes], str]] = "utf-8"
         self.redirect_count = 0
         self.redirect_url = ""
         self.http_version = 0
@@ -70,16 +75,62 @@ def __init__(self, curl: Optional[Curl] = None, request: Optional[Request] = Non
         self.astream_task: Optional[Awaitable] = None
         self.quit_now = None
 
+    @property
+    def charset(self) -> str:
+        """Alias for encoding."""
+        return self.encoding
+
+    @property
+    def encoding(self) -> str:
+        """
+        Determines the encoding to decode byte content into text.
+
+        The method follows a specific priority to decide the encoding:
+        1. If `.encoding` has been explicitly set, it is used.
+        2. The encoding specified by the `charset` parameter in the `Content-Type` header.
+        3. The encoding specified by the `default_encoding` attribute. This can either be
+           a string (e.g., "utf-8") or a callable for charset autodetection.
+        """
+        if not hasattr(self, "_encoding"):
+            encoding = self.charset_encoding
+            if encoding is None:
+                if isinstance(self.default_encoding, str):
+                    encoding = self.default_encoding
+                elif callable(self.default_encoding):
+                    encoding = self.default_encoding(self.content)
+            self._encoding = encoding or "utf-8"
+        return self._encoding
+
+    @encoding.setter
+    def encoding(self, value: str) -> None:
+        if hasattr(self, "_text"):
+            raise ValueError("Cannot set encoding after text has been accessed")
+        self._encoding = value
+
+    @property
+    def charset_encoding(self) -> Optional[str]:
+        """Return the encoding, as specified by the Content-Type header."""
+        content_type = self.headers.get("Content-Type")
+        if content_type:
+            charset_match = CHARSET_RE.search(content_type)
+            return charset_match.group(1) if charset_match else None
+        return None
+
+    @property
+    def text(self) -> str:
+        if not hasattr(self, "_text"):
+            if not self.content:
+                self._text = ""
+            else:
+                self._text = self._decode(self.content)
+        return self._text
+
     def _decode(self, content: bytes) -> str:
         try:
-            return content.decode(self.charset, errors="replace")
+            return content.decode(self.encoding, errors="replace")
         except (UnicodeDecodeError, LookupError):
             return content.decode("utf-8-sig")
 
-    @property
-    def text(self) -> str:
-        return self._decode(self.content)
-
     def raise_for_status(self):
         """Raise an error if status code is not in [200, 400)"""
         if not self.ok:

curl_cffi/requests/session.py

@@ -1,7 +1,6 @@
 import asyncio
 import math
 import queue
-import re
 import threading
 import warnings
 from concurrent.futures import ThreadPoolExecutor
@@ -55,7 +54,6 @@ class ProxySpec(TypedDict, total=False):
 else:
     ProxySpec = Dict[str, str]
 
-CHARSET_RE = re.compile(r"charset=([\w-]+)")
 ThreadType = Literal["eventlet", "gevent"]
 
 
@@ -205,6 +203,7 @@ def __init__(
         max_redirects: int = -1,
         impersonate: Optional[Union[str, BrowserType]] = None,
         default_headers: bool = True,
+        default_encoding: Union[str, Callable[[bytes], str]] = "utf-8",
         curl_options: Optional[dict] = None,
         curl_infos: Optional[list] = None,
         http_version: Optional[CurlHttpVersion] = None,
@@ -224,6 +223,7 @@ def __init__(
         self.max_redirects = max_redirects
         self.impersonate = impersonate
         self.default_headers = default_headers
+        self.default_encoding = default_encoding
         self.curl_options = curl_options or {}
         self.curl_infos = curl_infos or []
         self.http_version = http_version
@@ -547,7 +547,7 @@ def qput(chunk):
 
         return req, buffer, header_buffer, q, header_recved, quit_now
 
-    def _parse_response(self, curl, buffer, header_buffer):
+    def _parse_response(self, curl, buffer, header_buffer, default_encoding):
         c = curl
         rsp = Response(c)
         rsp.url = cast(bytes, c.getinfo(CurlInfo.EFFECTIVE_URL)).decode()
@@ -583,13 +583,7 @@ def _parse_response(self, curl, buffer, header_buffer):
         rsp.cookies = self.cookies
         # print("Cookies after extraction", self.cookies)
 
-        content_type = rsp.headers.get("Content-Type", default="")
-        charset_match = CHARSET_RE.search(content_type)
-        charset = charset_match.group(1) if charset_match else "utf-8"
-
-        rsp.charset = charset
-        rsp.encoding = charset  # TODO use chardet
-
+        rsp.default_encoding = default_encoding
         rsp.elapsed = cast(float, c.getinfo(CurlInfo.TOTAL_TIME))
         rsp.redirect_count = cast(int, c.getinfo(CurlInfo.REDIRECT_COUNT))
         rsp.redirect_url = cast(bytes, c.getinfo(CurlInfo.REDIRECT_URL)).decode()
@@ -639,6 +633,8 @@ def __init__(
             max_redirects: max redirect counts, default unlimited(-1).
             impersonate: which browser version to impersonate in the session.
             interface: which interface use in request to server.
+            default_encoding: encoding for decoding response content if charset is not found in headers. 
+                Defaults to "utf-8". Can be set to a callable for automatic detection.
 
         Notes:
             This class can be used as a context manager.
@@ -767,6 +763,7 @@ def request(
         content_callback: Optional[Callable] = None,
         impersonate: Optional[Union[str, BrowserType]] = None,
         default_headers: Optional[bool] = None,
+        default_encoding: Union[str, Callable[[bytes], str]] = "utf-8",
         http_version: Optional[CurlHttpVersion] = None,
         interface: Optional[str] = None,
         cert: Optional[Union[str, Tuple[str, str]]] = None,
@@ -825,7 +822,7 @@ def perform():
                 try:
                     c.perform()
                 except CurlError as e:
-                    rsp = self._parse_response(c, buffer, header_buffer)
+                    rsp = self._parse_response(c, buffer, header_buffer, default_encoding)
                     rsp.request = req
                     cast(queue.Queue, q).put_nowait(RequestsError(str(e), e.code, rsp))
                 finally:
@@ -843,7 +840,7 @@ def cleanup(fut):
 
             # Wait for the first chunk
             cast(threading.Event, header_recved).wait()
-            rsp = self._parse_response(c, buffer, header_buffer)
+            rsp = self._parse_response(c, buffer, header_buffer, default_encoding)
             header_parsed.set()
 
             # Raise the exception if something wrong happens when receiving the header.
@@ -868,11 +865,11 @@ def cleanup(fut):
                 else:
                     c.perform()
             except CurlError as e:
-                rsp = self._parse_response(c, buffer, header_buffer)
+                rsp = self._parse_response(c, buffer, header_buffer, default_encoding)
                 rsp.request = req
                 raise RequestsError(str(e), e.code, rsp) from e
             else:
-                rsp = self._parse_response(c, buffer, header_buffer)
+                rsp = self._parse_response(c, buffer, header_buffer, default_encoding)
                 rsp.request = req
                 return rsp
             finally:
@@ -919,6 +916,8 @@ def __init__(
             allow_redirects: whether to allow redirection.
             max_redirects: max redirect counts, default unlimited(-1).
             impersonate: which browser version to impersonate in the session.
+            default_encoding: encoding for decoding response content if charset is not found in headers. 
+                Defaults to "utf-8". Can be set to a callable for automatic detection.
 
         Notes:
             This class can be used as a context manager, and it's recommended to use via
@@ -1043,6 +1042,7 @@ async def request(
         content_callback: Optional[Callable] = None,
         impersonate: Optional[Union[str, BrowserType]] = None,
         default_headers: Optional[bool] = None,
+        default_encoding: Union[str, Callable[[bytes], str]] = "utf-8",
         http_version: Optional[CurlHttpVersion] = None,
         interface: Optional[str] = None,
         cert: Optional[Union[str, Tuple[str, str]]] = None,
@@ -1093,7 +1093,7 @@ async def perform():
                 try:
                     await task
                 except CurlError as e:
-                    rsp = self._parse_response(curl, buffer, header_buffer)
+                    rsp = self._parse_response(curl, buffer, header_buffer, default_encoding)
                     rsp.request = req
                     cast(asyncio.Queue, q).put_nowait(RequestsError(str(e), e.code, rsp))
                 finally:
@@ -1113,7 +1113,7 @@ def cleanup(fut):
             # Unlike threads, coroutines does not use preemptive scheduling.
             # For asyncio, there is no need for a header_parsed event, the
             # _parse_response will execute in the foreground, no background tasks running.
-            rsp = self._parse_response(curl, buffer, header_buffer)
+            rsp = self._parse_response(curl, buffer, header_buffer, default_encoding)
 
             first_element = _peek_aio_queue(cast(asyncio.Queue, q))
             if isinstance(first_element, RequestsError):
@@ -1132,11 +1132,11 @@ def cleanup(fut):
                 await task
                 # print(curl.getinfo(CurlInfo.CAINFO))
             except CurlError as e:
-                rsp = self._parse_response(curl, buffer, header_buffer)
+                rsp = self._parse_response(curl, buffer, header_buffer, default_encoding)
                 rsp.request = req
                 raise RequestsError(str(e), e.code, rsp) from e
             else:
-                rsp = self._parse_response(curl, buffer, header_buffer)
+                rsp = self._parse_response(curl, buffer, header_buffer, default_encoding)
                 rsp.request = req
                 return rsp
             finally: