From bca01dc1c608b08d39558c10cb6186dfca1aa7fb Mon Sep 17 00:00:00 2001
From: Davide Grilli <davide.grilli@outlook.com>
Date: Mon, 8 Jun 2026 16:20:59 +0200
Subject: [PATCH] feat(chunks): aggiunge ChunkerConfig dataclass

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 chunks/config.py            | 24 ++++++++++++++++++++++++
 tests/chunks/test_config.py | 29 +++++++++++++++++++++++++++++
 2 files changed, 53 insertions(+)
 create mode 100644 chunks/config.py
 create mode 100644 tests/chunks/test_config.py

diff --git a/chunks/config.py b/chunks/config.py
new file mode 100644
index 0000000..d0da27b
--- /dev/null
+++ b/chunks/config.py
@@ -0,0 +1,24 @@
+from __future__ import annotations
+from dataclasses import dataclass, field
+
+
+@dataclass
+class ChunkerConfig:
+    max_chars: int = 1200
+    min_chars: int = 80
+    target_chars: int = 800
+    context_depth: int = 3
+    skip_headings: set[str] = field(default_factory=lambda: {
+        "indice",
+        "sommario",
+        "bibliografia",
+        "ringraziamenti",
+        "abbreviazioni",
+    })
+    skip_pre_heading: bool = True
+    merge_short: bool = True
+    atomic_types: set[str] = field(default_factory=lambda: {
+        "table", "code", "list", "math", "html",
+    })
+    fail_on_broken_fence: bool = True
+    fail_on_content_loss: bool = False
diff --git a/tests/chunks/test_config.py b/tests/chunks/test_config.py
new file mode 100644
index 0000000..5f94370
--- /dev/null
+++ b/tests/chunks/test_config.py
@@ -0,0 +1,29 @@
+from chunks.config import ChunkerConfig
+
+
+def test_default_values():
+    c = ChunkerConfig()
+    assert c.max_chars == 1200
+    assert c.min_chars == 80
+    assert c.target_chars == 800
+    assert c.context_depth == 3
+    assert "indice" in c.skip_headings
+    assert c.skip_pre_heading is True
+    assert c.merge_short is True
+    assert "table" in c.atomic_types
+    assert "code" in c.atomic_types
+
+
+def test_custom_values():
+    c = ChunkerConfig(max_chars=2000, min_chars=100, context_depth=2)
+    assert c.max_chars == 2000
+    assert c.min_chars == 100
+    assert c.context_depth == 2
+
+
+def test_skip_headings_case_insensitive_check():
+    c = ChunkerConfig()
+    skip_lower = {h.lower() for h in c.skip_headings}
+    assert "indice" in skip_lower
+    assert "sommario" in skip_lower
+    assert "bibliografia" in skip_lower