feat(mcp-docugen): Task 1-3 config, models, auth middleware

- Settings con Pydantic Settings, validazione env obbligatori - Shared models: TemplateVariable/Frontmatter, ImageVariable, TokenUsage, GenerationResult, TemplateSummary, TemplateAsset - ApiKeyAuthMiddleware Bearer token con exempt paths 19 test, tutti passed. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-21 12:18:43 +02:00
parent c5e84a578b
commit d5c645bf17
6 changed files with 324 additions and 0 deletions
@@ -0,0 +1,34 @@
+from starlette.middleware.base import BaseHTTPMiddleware
+from starlette.requests import Request
+from starlette.responses import JSONResponse
+from starlette.types import ASGIApp
+
+
+class ApiKeyAuthMiddleware(BaseHTTPMiddleware):
+    def __init__(
+        self,
+        app: ASGIApp,
+        api_key: str,
+        exempt_paths: set[str] | None = None,
+    ) -> None:
+        super().__init__(app)
+        self.api_key = api_key
+        self.exempt_paths = exempt_paths or set()
+
+    async def dispatch(self, request: Request, call_next):
+        if request.url.path in self.exempt_paths:
+            return await call_next(request)
+
+        header = request.headers.get("Authorization", "")
+        if not header.startswith("Bearer "):
+            return JSONResponse(
+                status_code=401, content={"error": "invalid_api_key"}
+            )
+
+        provided = header.removeprefix("Bearer ").strip()
+        if provided != self.api_key:
+            return JSONResponse(
+                status_code=401, content={"error": "invalid_api_key"}
+            )
+
+        return await call_next(request)
@@ -0,0 +1,25 @@
+from pathlib import Path
+
+from pydantic import Field, field_validator
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=".env", env_file_encoding="utf-8", extra="ignore"
+    )
+
+    api_key: str = Field(..., min_length=8)
+    openrouter_api_key: str = Field(..., min_length=8)
+    openrouter_base_url: str = "https://openrouter.ai/api/v1"
+    llm_model_default: str = "anthropic/claude-sonnet-4"
+    public_base_url: str = Field(...)
+    data_dir: Path = Path("/data")
+    asset_ttl_days: int = 30
+    max_image_size_mb: int = 10
+    llm_timeout_seconds: int = 60
+
+    @field_validator("public_base_url")
+    @classmethod
+    def strip_trailing_slash(cls, v: str) -> str:
+        return v.rstrip("/")
@@ -0,0 +1,72 @@
+import re
+from datetime import datetime
+from typing import Literal
+
+from pydantic import BaseModel, Field, field_validator
+
+_SLUG_RE = re.compile(r"^[a-z0-9][a-z0-9-]*$")
+_SUPPORTED_MIMES = {"image/png", "image/jpeg", "image/webp"}
+
+
+class TemplateVariable(BaseModel):
+    name: str = Field(..., min_length=1)
+    type: Literal["string", "image"]
+
+
+class TemplateFrontmatter(BaseModel):
+    name: str
+    description: str = ""
+    model: str | None = None
+    required_variables: list[TemplateVariable] = Field(default_factory=list)
+    instructions_hint: str | None = None
+
+    @field_validator("name")
+    @classmethod
+    def validate_slug(cls, v: str) -> str:
+        if not _SLUG_RE.match(v):
+            raise ValueError(
+                f"invalid template name '{v}': must match {_SLUG_RE.pattern}"
+            )
+        return v
+
+
+class ImageVariable(BaseModel):
+    kind: Literal["image"]
+    data_b64: str
+    mime: str
+
+    @field_validator("mime")
+    @classmethod
+    def validate_mime(cls, v: str) -> str:
+        if v not in _SUPPORTED_MIMES:
+            raise ValueError(
+                f"unsupported mime '{v}'; allowed: {sorted(_SUPPORTED_MIMES)}"
+            )
+        return v
+
+
+class TokenUsage(BaseModel):
+    input: int = 0
+    output: int = 0
+
+
+class GenerationResult(BaseModel):
+    generation_id: str
+    markdown: str
+    model: str
+    tokens: TokenUsage
+    cost_usd: float
+    ephemeral_assets_urls: list[str] = Field(default_factory=list)
+    ephemeral_expires_at: datetime | None = None
+
+
+class TemplateSummary(BaseModel):
+    name: str
+    description: str
+    updated_at: datetime
+
+
+class TemplateAsset(BaseModel):
+    filename: str
+    mime: str
+    size_bytes: int
@@ -0,0 +1,57 @@
+import pytest
+from fastapi import FastAPI
+from fastapi.testclient import TestClient
+
+from mcp_docugen.auth import ApiKeyAuthMiddleware
+
+
+@pytest.fixture
+def app_with_auth():
+    app = FastAPI()
+    app.add_middleware(
+        ApiKeyAuthMiddleware, api_key="secret", exempt_paths={"/health"}
+    )
+
+    @app.get("/protected")
+    async def protected():
+        return {"ok": True}
+
+    @app.get("/health")
+    async def health():
+        return {"status": "ok"}
+
+    return app
+
+
+def test_protected_without_header_returns_401(app_with_auth):
+    client = TestClient(app_with_auth)
+    response = client.get("/protected")
+    assert response.status_code == 401
+    assert response.json() == {"error": "invalid_api_key"}
+
+
+def test_protected_with_wrong_key_returns_401(app_with_auth):
+    client = TestClient(app_with_auth)
+    response = client.get("/protected", headers={"Authorization": "Bearer wrong"})
+    assert response.status_code == 401
+
+
+def test_protected_with_correct_key_passes(app_with_auth):
+    client = TestClient(app_with_auth)
+    response = client.get("/protected", headers={"Authorization": "Bearer secret"})
+    assert response.status_code == 200
+    assert response.json() == {"ok": True}
+
+
+def test_health_bypasses_auth(app_with_auth):
+    client = TestClient(app_with_auth)
+    response = client.get("/health")
+    assert response.status_code == 200
+
+
+def test_malformed_auth_header_returns_401(app_with_auth):
+    client = TestClient(app_with_auth)
+    response = client.get("/protected", headers={"Authorization": "secret"})
+    assert response.status_code == 401
+    response = client.get("/protected", headers={"Authorization": "Basic secret"})
+    assert response.status_code == 401
@@ -0,0 +1,44 @@
+import pytest
+from pydantic import ValidationError
+
+from mcp_docugen.config import Settings
+
+
+def test_settings_loads_from_env(monkeypatch, tmp_path):
+    monkeypatch.setenv("API_KEY", "test-api-key")
+    monkeypatch.setenv("OPENROUTER_API_KEY", "sk-or-test")
+    monkeypatch.setenv("PUBLIC_BASE_URL", "https://mcp.example.com")
+    monkeypatch.setenv("DATA_DIR", str(tmp_path))
+
+    settings = Settings()
+
+    assert settings.api_key == "test-api-key"
+    assert settings.openrouter_api_key == "sk-or-test"
+    assert settings.public_base_url == "https://mcp.example.com"
+    assert settings.data_dir == tmp_path
+    assert settings.llm_model_default == "anthropic/claude-sonnet-4"
+    assert settings.asset_ttl_days == 30
+    assert settings.max_image_size_mb == 10
+    assert settings.llm_timeout_seconds == 60
+    assert settings.openrouter_base_url == "https://openrouter.ai/api/v1"
+
+
+def test_settings_missing_required_fails(monkeypatch, tmp_path):
+    monkeypatch.delenv("API_KEY", raising=False)
+    monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
+    monkeypatch.delenv("PUBLIC_BASE_URL", raising=False)
+    monkeypatch.setenv("DATA_DIR", str(tmp_path))
+
+    with pytest.raises(ValidationError):
+        Settings(_env_file=None)
+
+
+def test_public_base_url_strips_trailing_slash(monkeypatch, tmp_path):
+    monkeypatch.setenv("API_KEY", "k-with-long-enough")
+    monkeypatch.setenv("OPENROUTER_API_KEY", "sk-long-enough")
+    monkeypatch.setenv("PUBLIC_BASE_URL", "https://mcp.example.com/")
+    monkeypatch.setenv("DATA_DIR", str(tmp_path))
+
+    settings = Settings()
+
+    assert settings.public_base_url == "https://mcp.example.com"
@@ -0,0 +1,92 @@
+import pytest
+from pydantic import ValidationError
+
+from mcp_docugen.models import (
+    GenerationResult,
+    ImageVariable,
+    TemplateFrontmatter,
+    TemplateVariable,
+    TokenUsage,
+)
+
+
+def test_template_variable_string():
+    v = TemplateVariable(name="cliente", type="string")
+    assert v.name == "cliente"
+    assert v.type == "string"
+
+
+def test_template_variable_image():
+    v = TemplateVariable(name="foto", type="image")
+    assert v.type == "image"
+
+
+def test_template_variable_rejects_unknown_type():
+    with pytest.raises(ValidationError):
+        TemplateVariable(name="x", type="number")
+
+
+def test_template_variable_rejects_empty_name():
+    with pytest.raises(ValidationError):
+        TemplateVariable(name="", type="string")
+
+
+def test_frontmatter_minimal():
+    fm = TemplateFrontmatter(name="fattura", description="Fattura commerciale")
+    assert fm.name == "fattura"
+    assert fm.required_variables == []
+    assert fm.model is None
+    assert fm.instructions_hint is None
+
+
+def test_frontmatter_with_variables():
+    fm = TemplateFrontmatter(
+        name="fattura",
+        description="...",
+        model="anthropic/claude-sonnet-4",
+        required_variables=[
+            {"name": "cliente", "type": "string"},
+            {"name": "foto", "type": "image"},
+        ],
+        instructions_hint="tono formale",
+    )
+    assert len(fm.required_variables) == 2
+    assert fm.required_variables[1].type == "image"
+
+
+def test_frontmatter_name_slug_rules():
+    with pytest.raises(ValidationError):
+        TemplateFrontmatter(name="Fattura Commerciale", description="x")
+    with pytest.raises(ValidationError):
+        TemplateFrontmatter(name="../escape", description="x")
+    with pytest.raises(ValidationError):
+        TemplateFrontmatter(name="-starts-with-dash", description="x")
+
+
+def test_image_variable():
+    img = ImageVariable(kind="image", data_b64="aGVsbG8=", mime="image/png")
+    assert img.mime == "image/png"
+
+
+def test_image_variable_rejects_unsupported_mime():
+    with pytest.raises(ValidationError):
+        ImageVariable(kind="image", data_b64="aGVsbG8=", mime="image/bmp")
+
+
+def test_image_variable_rejects_wrong_kind():
+    with pytest.raises(ValidationError):
+        ImageVariable(kind="file", data_b64="aGVsbG8=", mime="image/png")
+
+
+def test_generation_result_shape():
+    result = GenerationResult(
+        generation_id="abc-123",
+        markdown="# hello",
+        model="anthropic/claude-sonnet-4",
+        tokens=TokenUsage(input=100, output=200),
+        cost_usd=0.01,
+        ephemeral_assets_urls=[],
+        ephemeral_expires_at=None,
+    )
+    assert result.generation_id == "abc-123"
+    assert result.tokens.input == 100