refactor(devgen/providers): shared provider base class

Abstract out generate validation and error handling into BaseProvider, eliminating duplicated logic across Anthropic, OpenRouter, and OpenAI providers.

devgen/providers/gemini.py

@@ -1,31 +1,22 @@
 import time
+
 from google import genai
 from google.genai import types
 
-from devgen.utils import format_token_limit_error, is_token_limit_error
+from devgen.providers.base import BaseProvider
+from devgen.utils import is_token_limit_error
 
 
-class GeminiProvider:
+class GeminiProvider(BaseProvider):
     """Generates content using Google's Gemini models."""
 
-    def generate(
-        self, prompt: str, api_key: str, model: str = "gemini-2.5-flash", **kwargs
-    ) -> str:
-        """Generates a response using the Gemini API.
-
-        Default model is updated to gemini-2.5-flash which is the latest
-        recommended model.
-        """
-        if not api_key:
-            raise ValueError(
-                "Gemini API key is missing. "
-                "Set it via `devgen setup config` or pass --api-key."
-            )
-
-        client = genai.Client(api_key=api_key)
+    DISPLAY_NAME = "Gemini"
+    DEFAULT_MODEL = "gemini-2.5-flash"
 
+    def _generate(self, prompt, api_key, model, **kwargs):
         max_retries = kwargs.get("max_retries", 5)
-        base_delay = kwargs.get("retry_delay", 10)  # 10 seconds base delay for 429
+        base_delay = kwargs.get("retry_delay", 10)  # 10s base for 429 backoff
+        client = genai.Client(api_key=api_key)
 
         for attempt in range(max_retries + 1):
             try:
@@ -39,38 +30,34 @@ class GeminiProvider:
                         max_output_tokens=kwargs.get("max_output_tokens", 2048),
                     ),
                 )
-                return response.text.strip()
+                return response.text
             except Exception as e:
-                error_msg = str(e).upper()
-                # Token / context-window overflow is not retryable — no point
-                # backing off, the prompt is just too big.
                 if is_token_limit_error(e):
-                    raise RuntimeError(format_token_limit_error("Gemini", e)) from e
-                # Check for 429 or ResourceExhausted in common error formats
-                if any(
-                    x in error_msg
-                    for x in [
-                        "429",
-                        "RESOURCE_EXHAUSTED",
-                        "QUOTA_EXCEEDED",
-                        "THROTTLED",
-                    ]
-                ):
-                    if attempt < max_retries:
-                        delay = base_delay * (2**attempt)
-                        print(
-                            f"\n[Gemini] Quota limit hit (429). This is often a temporary rate limit (RPM). "
-                            f"Waiting {delay}s and retrying... (Attempt {attempt + 1}/{max_retries})"
-                        )
-                        time.sleep(delay)
-                        continue
-                    else:
-                        raise RuntimeError(
-                            "Gemini API quota exhausted (429) after multiple retries.\n\n"
-                            "POSSIBLE CAUSES:\n"
-                            "1. Rate Limit (RPM/RPD): You are sending requests too fast for the Free Tier.\n"
-                            "2. Daily Limit: You have reached the maximum free requests allowed for today (typically 1,500).\n"
-                            "3. Credit/Billing: If on a paid plan, your credit may be exhausted.\n\n"
-                            "Please check your usage metrics at: https://aistudio.google.com/app/usage"
-                        ) from e
-                raise RuntimeError(f"Gemini generation failed: {e}") from e
+                    raise  # base class handles with friendly message
+                if self._is_quota_error(e) and attempt < max_retries:
+                    delay = base_delay * (2**attempt)
+                    print(
+                        f"\n[Gemini] Quota limit hit (429). This is often a temporary "
+                        f"rate limit (RPM). Waiting {delay}s and retrying... "
+                        f"(Attempt {attempt + 1}/{max_retries})"
+                    )
+                    time.sleep(delay)
+                    continue
+                if self._is_quota_error(e):
+                    raise RuntimeError(
+                        "Gemini API quota exhausted (429) after multiple retries.\n\n"
+                        "POSSIBLE CAUSES:\n"
+                        "1. Rate Limit (RPM/RPD): You are sending requests too fast for the Free Tier.\n"
+                        "2. Daily Limit: You have reached the maximum free requests allowed for today (typically 1,500).\n"
+                        "3. Credit/Billing: If on a paid plan, your credit may be exhausted.\n\n"
+                        "Please check your usage metrics at: https://aistudio.google.com/app/usage"
+                    ) from e
+                raise
+
+    @staticmethod
+    def _is_quota_error(error: Exception) -> bool:
+        text = str(error).upper()
+        return any(
+            marker in text
+            for marker in ("429", "RESOURCE_EXHAUSTED", "QUOTA_EXCEEDED", "THROTTLED")
+        )