Add multilingual STT support and config UI/runtime updates

tests/test_aiprocess.py

@@ -1,3 +1,5 @@
+import json
+import os
 import sys
 import tempfile
 import unittest
@@ -12,7 +14,9 @@ if str(SRC) not in sys.path:
 
 import aiprocess
 from aiprocess import (
+    ExternalApiProcessor,
     _assert_expected_model_checksum,
+    _build_request_payload,
     _extract_cleaned_text,
     _profile_generation_kwargs,
     _supports_response_format,
@@ -120,6 +124,20 @@ class ModelChecksumTests(unittest.TestCase):
             _assert_expected_model_checksum("0" * 64)
 
 
+class RequestPayloadTests(unittest.TestCase):
+    def test_build_request_payload_with_dictionary(self):
+        payload = _build_request_payload("hello", lang="en", dictionary_context="Docker")
+        self.assertEqual(payload["language"], "en")
+        self.assertEqual(payload["transcript"], "hello")
+        self.assertEqual(payload["dictionary"], "Docker")
+
+    def test_build_request_payload_omits_empty_dictionary(self):
+        payload = _build_request_payload("hello", lang="en", dictionary_context="   ")
+        self.assertEqual(payload["language"], "en")
+        self.assertEqual(payload["transcript"], "hello")
+        self.assertNotIn("dictionary", payload)
+
+
 class _Response:
     def __init__(self, payload: bytes):
         self.payload = payload
@@ -136,9 +154,13 @@ class _Response:
             return str(len(self.payload))
         return None
 
-    def read(self, size: int) -> bytes:
+    def read(self, size: int = -1) -> bytes:
         if self.offset >= len(self.payload):
             return b""
+        if size < 0:
+            chunk = self.payload[self.offset :]
+            self.offset = len(self.payload)
+            return chunk
         chunk = self.payload[self.offset : self.offset + size]
         self.offset += len(chunk)
         return chunk
@@ -196,5 +218,42 @@ class EnsureModelTests(unittest.TestCase):
                     ensure_model()
 
 
+class ExternalApiProcessorTests(unittest.TestCase):
+    def test_requires_api_key_env_var(self):
+        with patch.dict(os.environ, {}, clear=True):
+            with self.assertRaisesRegex(RuntimeError, "missing external api key"):
+                ExternalApiProcessor(
+                    provider="openai",
+                    base_url="https://api.openai.com/v1",
+                    model="gpt-4o-mini",
+                    api_key_env_var="AMAN_EXTERNAL_API_KEY",
+                    timeout_ms=1000,
+                    max_retries=0,
+                )
+
+    def test_process_uses_chat_completion_endpoint(self):
+        response_payload = {
+            "choices": [{"message": {"content": '{"cleaned_text":"clean"}'}}],
+        }
+        response_body = json.dumps(response_payload).encode("utf-8")
+        with patch.dict(os.environ, {"AMAN_EXTERNAL_API_KEY": "test-key"}, clear=True), patch(
+            "aiprocess.urllib.request.urlopen",
+            return_value=_Response(response_body),
+        ) as urlopen:
+            processor = ExternalApiProcessor(
+                provider="openai",
+                base_url="https://api.openai.com/v1",
+                model="gpt-4o-mini",
+                api_key_env_var="AMAN_EXTERNAL_API_KEY",
+                timeout_ms=1000,
+                max_retries=0,
+            )
+            out = processor.process("raw text", dictionary_context="Docker")
+
+        self.assertEqual(out, "clean")
+        request = urlopen.call_args[0][0]
+        self.assertTrue(request.full_url.endswith("/chat/completions"))
+
+
 if __name__ == "__main__":
     unittest.main()