Implemented optional concurrency for the LLM backend

tests/test_new_config.py

@@ -6,6 +6,7 @@ from audita.core.config import (
     DEFAULT_GLOSSARY_CONFIDENCE_THRESHOLD,
     DEFAULT_GRAMMAR_CONFIDENCE_THRESHOLD,
     DEFAULT_HOMOPHONES_CONFIDENCE_THRESHOLD,
+    DEFAULT_LLM_CONCURRENCY,
     DEFAULT_NORMALIZE_MAX_SEGMENT_GAP,
     DEFAULT_SPOKEN_WORD_CONFIDENCE_THRESHOLD,
     DEFAULT_WORK_DIR_RETENTION,
@@ -18,6 +19,7 @@ def test_default_config_allows_missing_api_key():
     config = AuditaConfig.from_sources(env={})
 
     assert config.api_key is None
+    assert config.llm_concurrency == DEFAULT_LLM_CONCURRENCY
     assert config.module_keys == DEFAULT_MODULE_KEYS
     assert config.glossary_confidence_threshold == DEFAULT_GLOSSARY_CONFIDENCE_THRESHOLD
     assert config.grammar_confidence_threshold == DEFAULT_GRAMMAR_CONFIDENCE_THRESHOLD
@@ -36,6 +38,21 @@ def test_cli_overrides_take_precedence():
     assert config.max_section_tokens == 2000
 
 
+def test_llm_concurrency_cli_override_takes_precedence():
+    config = AuditaConfig.from_sources(
+        env={"AUDITA_LLM_CONCURRENCY": "2"},
+        overrides=ConfigOverrides(llm_concurrency=4),
+    )
+
+    assert config.llm_concurrency == 4
+
+
+def test_llm_concurrency_env_is_parsed():
+    config = AuditaConfig.from_sources(env={"AUDITA_LLM_CONCURRENCY": "3"})
+
+    assert config.llm_concurrency == 3
+
+
 def test_generic_llm_api_key_env_is_read():
     config = AuditaConfig.from_sources(env={"AUDITA_LLM_API_KEY": "generic-key"})
 
@@ -141,3 +158,9 @@ def test_invalid_module_sequences_are_rejected(value):
 def test_invalid_thresholds_are_rejected(env_name):
     with pytest.raises(AuditaConfigError):
         AuditaConfig.from_sources(env={env_name: "1.5"})
+
+
+@pytest.mark.parametrize("value", ["0", "-1", "many"])
+def test_invalid_llm_concurrency_is_rejected(value):
+    with pytest.raises(AuditaConfigError, match="AUDITA_LLM_CONCURRENCY"):
+        AuditaConfig.from_sources(env={"AUDITA_LLM_CONCURRENCY": value})