From 023e7eaecff492d817a0f496e1b0641f2e29927b Mon Sep 17 00:00:00 2001
From: Jon Dufresne <jon.dufresne@gmail.com>
Date: Wed, 13 Jul 2022 08:04:50 -0700
Subject: [PATCH] Distinguish between final and intermediate ResultDict (#266)

The IntermediateResultDict has optional fields where as the final, all
fields can't be None.

Carried over from the typeshed types:
https://github.com/python/typeshed/blob/ee09a67c5cf61effaae1bdb894571eac4fc9ac6b/stubs/chardet/chardet/__init__.pyi
---
 chardet/__init__.py          |  8 ++++----
 chardet/resultdict.py        | 10 ++++++++--
 chardet/universaldetector.py | 12 ++++++------
 3 files changed, 18 insertions(+), 12 deletions(-)

diff --git a/chardet/__init__.py b/chardet/__init__.py
index 2112ee14..69afe7ef 100644
--- a/chardet/__init__.py
+++ b/chardet/__init__.py
@@ -20,14 +20,14 @@
 from .charsetgroupprober import CharSetGroupProber
 from .charsetprober import CharSetProber
 from .enums import InputState
-from .resultdict import ResultDict
+from .resultdict import FinalResultDict, IntermediateResultDict
 from .universaldetector import UniversalDetector
 from .version import VERSION, __version__
 
 __all__ = ["UniversalDetector", "detect", "detect_all", "__version__", "VERSION"]
 
 
-def detect(byte_str: Union[bytes, bytearray]) -> ResultDict:
+def detect(byte_str: Union[bytes, bytearray]) -> FinalResultDict:
     """
     Detect the encoding of the given byte string.
 
@@ -47,7 +47,7 @@ def detect(byte_str: Union[bytes, bytearray]) -> ResultDict:
 
 def detect_all(
     byte_str: Union[bytes, bytearray], ignore_threshold: bool = False
-) -> List[ResultDict]:
+) -> List[IntermediateResultDict]:
     """
     Detect all the possible encodings of the given byte string.
 
@@ -70,7 +70,7 @@ def detect_all(
     detector.close()
 
     if detector.input_state == InputState.HIGH_BYTE:
-        results: List[ResultDict] = []
+        results: List[IntermediateResultDict] = []
         probers: List[CharSetProber] = []
         for prober in detector.charset_probers:
             if isinstance(prober, CharSetGroupProber):
diff --git a/chardet/resultdict.py b/chardet/resultdict.py
index 7d36e64c..bf889be3 100644
--- a/chardet/resultdict.py
+++ b/chardet/resultdict.py
@@ -7,10 +7,16 @@
     # for Python 3.7.
     from typing import TypedDict
 
-    class ResultDict(TypedDict):
+    class FinalResultDict(TypedDict):
+        encoding: str
+        confidence: float
+        language: str
+
+    class IntermediateResultDict(TypedDict):
         encoding: Optional[str]
         confidence: float
         language: Optional[str]
 
 else:
-    ResultDict = dict
+    FinalResultDict = dict
+    IntermediateResultDict = dict
diff --git a/chardet/universaldetector.py b/chardet/universaldetector.py
index f94961e6..b3658474 100644
--- a/chardet/universaldetector.py
+++ b/chardet/universaldetector.py
@@ -39,7 +39,7 @@ class a user of ``chardet`` should use.
 import codecs
 import logging
 import re
-from typing import List, Optional, Union
+from typing import List, Optional, Union, cast
 
 from .charsetgroupprober import CharSetGroupProber
 from .charsetprober import CharSetProber
@@ -48,7 +48,7 @@ class a user of ``chardet`` should use.
 from .latin1prober import Latin1Prober
 from .macromanprober import MacRomanProber
 from .mbcsgroupprober import MBCSGroupProber
-from .resultdict import ResultDict
+from .resultdict import FinalResultDict, IntermediateResultDict
 from .sbcsgroupprober import SBCSGroupProber
 from .utf1632prober import UTF1632Prober
 
@@ -89,7 +89,7 @@ def __init__(self, lang_filter: LanguageFilter = LanguageFilter.ALL) -> None:
         self._esc_charset_prober: Optional[EscCharSetProber] = None
         self._utf1632_prober: Optional[UTF1632Prober] = None
         self._charset_probers: List[CharSetProber] = []
-        self.result: ResultDict = {
+        self.result: IntermediateResultDict = {
             "encoding": None,
             "confidence": 0.0,
             "language": None,
@@ -263,7 +263,7 @@ def feed(self, byte_str: Union[bytes, bytearray]) -> None:
             if self.WIN_BYTE_DETECTOR.search(byte_str):
                 self._has_win_bytes = True
 
-    def close(self) -> ResultDict:
+    def close(self) -> FinalResultDict:
         """
         Stop analyzing the current document and come up with a final
         prediction.
@@ -273,7 +273,7 @@ def close(self) -> ResultDict:
         """
         # Don't bother with checks if we're already done
         if self.done:
-            return self.result
+            return cast(FinalResultDict, self.result)
         self.done = True
 
         if not self._got_data:
@@ -335,4 +335,4 @@ def close(self) -> ResultDict:
                             group_prober.language,
                             group_prober.get_confidence(),
                         )
-        return self.result
+        return cast(FinalResultDict, self.result)