Resolve linter errors

Author: dippatel1994

paperbanana/agents/visualizer.py

@@ -107,6 +107,7 @@ async def _generate_plot(
         full_description = description
         if raw_data:
             import json
+
             full_description += f"\n\n## Raw Data\n```json\n{json.dumps(raw_data, indent=2)}\n```"
 
         # Load and format the plot visualizer prompt template

paperbanana/cli.py

@@ -25,33 +25,23 @@
 
 @app.command()
 def generate(
-    input: str = typer.Option(
-        ..., "--input", "-i", help="Path to methodology text file"
-    ),
+    input: str = typer.Option(..., "--input", "-i", help="Path to methodology text file"),
     caption: str = typer.Option(
         ..., "--caption", "-c", help="Figure caption / communicative intent"
     ),
-    output: Optional[str] = typer.Option(
-        None, "--output", "-o", help="Output image path"
-    ),
+    output: Optional[str] = typer.Option(None, "--output", "-o", help="Output image path"),
     vlm_provider: Optional[str] = typer.Option(
         None, "--vlm-provider", help="VLM provider (gemini)"
     ),
-    vlm_model: Optional[str] = typer.Option(
-        None, "--vlm-model", help="VLM model name"
-    ),
+    vlm_model: Optional[str] = typer.Option(None, "--vlm-model", help="VLM model name"),
     image_provider: Optional[str] = typer.Option(
         None, "--image-provider", help="Image gen provider"
     ),
-    image_model: Optional[str] = typer.Option(
-        None, "--image-model", help="Image gen model name"
-    ),
+    image_model: Optional[str] = typer.Option(None, "--image-model", help="Image gen model name"),
     iterations: Optional[int] = typer.Option(
         None, "--iterations", "-n", help="Refinement iterations"
     ),
-    config: Optional[str] = typer.Option(
-        None, "--config", help="Path to config YAML file"
-    ),
+    config: Optional[str] = typer.Option(None, "--config", help="Path to config YAML file"),
 ):
     """Generate a methodology diagram from a text description."""
     # Load source text
@@ -81,6 +71,7 @@ def generate(
         settings = Settings.from_yaml(config, **overrides)
     else:
         from dotenv import load_dotenv
+
         load_dotenv()
         settings = Settings(**overrides)
 
@@ -91,13 +82,15 @@ def generate(
         diagram_type=DiagramType.METHODOLOGY,
     )
 
-    console.print(Panel.fit(
-        f"[bold]PaperBanana[/bold] - Generating Methodology Diagram\n\n"
-        f"VLM: {settings.vlm_provider} / {settings.vlm_model}\n"
-        f"Image: {settings.image_provider} / {settings.image_model}\n"
-        f"Iterations: {settings.refinement_iterations}",
-        border_style="blue",
-    ))
+    console.print(
+        Panel.fit(
+            f"[bold]PaperBanana[/bold] - Generating Methodology Diagram\n\n"
+            f"VLM: {settings.vlm_provider} / {settings.vlm_model}\n"
+            f"Image: {settings.image_provider} / {settings.image_model}\n"
+            f"Iterations: {settings.refinement_iterations}",
+            border_style="blue",
+        )
+    )
 
     # Run pipeline
     from paperbanana.core.pipeline import PaperBananaPipeline
@@ -135,8 +128,10 @@ def plot(
 
     # Load data
     import json as json_mod
+
     if data_path.suffix == ".csv":
         import pandas as pd
+
         df = pd.read_csv(data_path)
         raw_data = df.to_dict(orient="records")
         source_context = (
@@ -149,6 +144,7 @@ def plot(
         source_context = f"JSON data:\n{json_mod.dumps(raw_data, indent=2)[:2000]}"
 
     from dotenv import load_dotenv
+
     load_dotenv()
 
     settings = Settings(
@@ -163,12 +159,14 @@ def plot(
         raw_data={"data": raw_data},
     )
 
-    console.print(Panel.fit(
-        f"[bold]PaperBanana[/bold] - Generating Statistical Plot\n\n"
-        f"Data: {data_path.name}\n"
-        f"Intent: {intent}",
-        border_style="green",
-    ))
+    console.print(
+        Panel.fit(
+            f"[bold]PaperBanana[/bold] - Generating Statistical Plot\n\n"
+            f"Data: {data_path.name}\n"
+            f"Intent: {intent}",
+            border_style="green",
+        )
+    )
 
     from paperbanana.core.pipeline import PaperBananaPipeline
 
@@ -183,16 +181,19 @@ async def _run():
 @app.command()
 def setup():
     """Interactive setup wizard — get generating in 2 minutes with FREE APIs."""
-    console.print(Panel.fit(
-        "[bold]Welcome to PaperBanana Setup[/bold]\n\n"
-        "We'll set up FREE API keys so you can start generating diagrams.",
-        border_style="yellow",
-    ))
+    console.print(
+        Panel.fit(
+            "[bold]Welcome to PaperBanana Setup[/bold]\n\n"
+            "We'll set up FREE API keys so you can start generating diagrams.",
+            border_style="yellow",
+        )
+    )
 
     console.print("\n[bold]Step 1: Google Gemini API Key[/bold] (FREE, no credit card)")
     console.print("This powers the AI agents that plan and critique your diagrams.\n")
 
     import webbrowser
+
     open_browser = Prompt.ask(
         "Open browser to get a free Gemini API key?",
         choices=["y", "n"],
@@ -220,18 +221,10 @@ def setup():
 
 @app.command()
 def evaluate(
-    generated: str = typer.Option(
-        ..., "--generated", "-g", help="Path to generated image"
-    ),
-    context: str = typer.Option(
-        ..., "--context", help="Path to source context text file"
-    ),
-    caption: str = typer.Option(
-        ..., "--caption", "-c", help="Figure caption"
-    ),
-    reference: str = typer.Option(
-        ..., "--reference", "-r", help="Path to human reference image"
-    ),
+    generated: str = typer.Option(..., "--generated", "-g", help="Path to generated image"),
+    context: str = typer.Option(..., "--context", help="Path to source context text file"),
+    caption: str = typer.Option(..., "--caption", "-c", help="Figure caption"),
+    reference: str = typer.Option(..., "--reference", "-r", help="Path to human reference image"),
     vlm_provider: str = typer.Option(
         "gemini", "--vlm-provider", help="VLM provider for evaluation"
     ),
@@ -252,10 +245,12 @@ def evaluate(
     context_text = Path(context).read_text(encoding="utf-8")
 
     from dotenv import load_dotenv
+
     load_dotenv()
 
     settings = Settings(vlm_provider=vlm_provider)
     from paperbanana.providers.registry import ProviderRegistry
+
     vlm = ProviderRegistry.create_vlm(settings)
 
     judge = VLMJudge(vlm)
@@ -276,12 +271,14 @@ async def _run():
         result = getattr(scores, dim)
         dim_lines.append(f"{dim.capitalize():14s} {result.winner}")
 
-    console.print(Panel.fit(
-        "[bold]Evaluation Results (Comparative)[/bold]\n\n"
-        + "\n".join(dim_lines)
-        + f"\n[bold]{'Overall':14s} {scores.overall_winner}[/bold]",
-        border_style="cyan",
-    ))
+    console.print(
+        Panel.fit(
+            "[bold]Evaluation Results (Comparative)[/bold]\n\n"
+            + "\n".join(dim_lines)
+            + f"\n[bold]{'Overall':14s} {scores.overall_winner}[/bold]",
+            border_style="cyan",
+        )
+    )
 
     for dim in dims:
         result = getattr(scores, dim)

paperbanana/core/pipeline.py

@@ -293,6 +293,7 @@ async def generate(self, input: GenerationInput) -> GenerationOutput:
 
         # Copy final image to output location
         import shutil
+
         shutil.copy2(final_image, final_output_path)
 
         # Build metadata

paperbanana/core/types.py

@@ -90,7 +90,8 @@ class DimensionResult(BaseModel):
 
     winner: str = Field(description="Model | Human | Both are good | Both are bad")
     score: float = Field(
-        ge=0.0, le=100.0,
+        ge=0.0,
+        le=100.0,
         description="100 (Model wins), 0 (Human wins), 50 (Tie)",
     )
     reasoning: str = Field(default="", description="Comparison reasoning")
@@ -109,11 +110,10 @@ class EvaluationScore(BaseModel):
     conciseness: DimensionResult
     readability: DimensionResult
     aesthetics: DimensionResult
-    overall_winner: str = Field(
-        description="Hierarchical aggregation result"
-    )
+    overall_winner: str = Field(description="Hierarchical aggregation result")
     overall_score: float = Field(
-        ge=0.0, le=100.0,
+        ge=0.0,
+        le=100.0,
         description="100 (Model wins), 0 (Human wins), 50 (Tie)",
     )
 

paperbanana/evaluation/judge.py

@@ -94,9 +94,7 @@ async def evaluate(
             overall_score=overall_score,
         )
 
-    def _load_eval_prompt(
-        self, dimension: str, source_context: str, caption: str
-    ) -> str:
+    def _load_eval_prompt(self, dimension: str, source_context: str, caption: str) -> str:
         """Load evaluation prompt for a specific dimension."""
         prompt_path = self.prompt_dir / "evaluation" / f"{dimension}.txt"
         if not prompt_path.exists():
@@ -122,9 +120,7 @@ def _parse_result(self, response: str, dimension: str) -> DimensionResult:
                 winner = "Both are good"
 
             score = WINNER_SCORE_MAP.get(winner, 50.0)
-            return DimensionResult(
-                winner=winner, score=score, reasoning=reasoning
-            )
+            return DimensionResult(winner=winner, score=score, reasoning=reasoning)
         except (json.JSONDecodeError, ValueError, TypeError) as e:
             logger.warning(
                 "Failed to parse evaluation response",
@@ -137,9 +133,7 @@ def _parse_result(self, response: str, dimension: str) -> DimensionResult:
                 reasoning="Could not parse evaluation response.",
             )
 
-    def _hierarchical_aggregate(
-        self, results: dict[str, DimensionResult]
-    ) -> str:
+    def _hierarchical_aggregate(self, results: dict[str, DimensionResult]) -> str:
         """Apply hierarchical aggregation per paper Section 4.2.
 
         Primary dimensions (Faithfulness + Readability) take precedence.

paperbanana/providers/registry.py

@@ -27,18 +27,13 @@ def create_vlm(settings: Settings) -> VLMProvider:
                 model=settings.vlm_model,
             )
         else:
-            raise ValueError(
-                f"Unknown VLM provider: {provider}. "
-                f"Available: gemini"
-            )
+            raise ValueError(f"Unknown VLM provider: {provider}. Available: gemini")
 
     @staticmethod
     def create_image_gen(settings: Settings) -> ImageGenProvider:
         """Create an image generation provider based on settings."""
         provider = settings.image_provider.lower()
-        logger.info(
-            "Creating image gen provider", provider=provider, model=settings.image_model
-        )
+        logger.info("Creating image gen provider", provider=provider, model=settings.image_model)
 
         if provider == "google_imagen":
             from paperbanana.providers.image_gen.google_imagen import GoogleImagenGen
@@ -48,7 +43,4 @@ def create_image_gen(settings: Settings) -> ImageGenProvider:
                 model=settings.image_model,
             )
         else:
-            raise ValueError(
-                f"Unknown image provider: {provider}. "
-                f"Available: google_imagen"
-            )
+            raise ValueError(f"Unknown image provider: {provider}. Available: google_imagen")

tests/test_agents/test_retriever.py

@@ -19,8 +19,15 @@ class MockVLM:
     def __init__(self, response: str = ""):
         self._response = response
 
-    async def generate(self, prompt, images=None, system_prompt=None,
-                       temperature=1.0, max_tokens=4096, response_format=None):
+    async def generate(
+        self,
+        prompt,
+        images=None,
+        system_prompt=None,
+        temperature=1.0,
+        max_tokens=4096,
+        response_format=None,
+    ):
         return self._response
 
     def is_available(self):
@@ -76,13 +83,15 @@ async def test_retriever_empty_candidates():
 @pytest.mark.asyncio
 async def test_retriever_parses_vlm_response():
     """Test that retriever correctly parses VLM JSON response."""
-    response = json.dumps({
-        "selected_ids": ["ref_001", "ref_003"],
-        "reasoning": {
-            "ref_001": "Relevant because...",
-            "ref_003": "Relevant because...",
+    response = json.dumps(
+        {
+            "selected_ids": ["ref_001", "ref_003"],
+            "reasoning": {
+                "ref_001": "Relevant because...",
+                "ref_003": "Relevant because...",
+            },
         }
-    })
+    )
 
     vlm = MockVLM(response=response)
     agent = RetrieverAgent(vlm)