feat: complete v1 implementation - all services, UI views, and tests

Adds remaining files from parallel development:
- Services: analysis, csv_reader, forecasting, normalizer, recurring
- UI: recurring_view, settings_view, sidebar, themes (dark/light)
- Tests: analysis, csv_reader, forecasting, import_categorize,
  normalizer, recurring, integration
- App entry point (main.py) and CLAUDE.md

52 tests passing across all modules.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

tests/services/test_csv_reader.py

@@ -0,0 +1,42 @@
+from pathlib import Path
+from src.services.csv_reader import read_csv, detect_format
+
+RAWDATA = Path(__file__).parent.parent.parent / "rawdata"
+
+
+def test_detect_chase_format():
+    result = detect_format(RAWDATA / "Chase0372_Activity20260101_20260210_20260210.CSV")
+    assert result["has_headers"] is True
+    assert "Transaction Date" in result["headers"]
+    assert result["delimiter"] == ","
+    assert result["row_count"] > 0
+
+
+def test_detect_checking_format():
+    result = detect_format(RAWDATA / "Checking1.csv")
+    assert result["has_headers"] is False
+    assert result["column_count"] == 5
+    assert result["row_count"] > 0
+
+
+def test_read_chase_csv():
+    rows = read_csv(RAWDATA / "Chase0372_Activity20260101_20260210_20260210.CSV")
+    assert len(rows) > 100
+    first = rows[0]
+    assert "Transaction Date" in first
+    assert "Amount" in first
+    assert "Description" in first
+
+
+def test_read_headerless_csv():
+    rows = read_csv(RAWDATA / "Checking1.csv")
+    assert len(rows) > 50
+    first = rows[0]
+    # Headerless: keys should be column indices (integers)
+    assert 0 in first or "0" in first
+
+
+def test_preview_rows():
+    result = detect_format(RAWDATA / "Chase0372_Activity20260101_20260210_20260210.CSV")
+    assert "preview" in result
+    assert len(result["preview"]) <= 5