chore: auto-commit (95 archivos)

- cmd/fn/doctor.go - cmd/fn/main.go - cpp/apps/primitives_gallery/playground/tables/CMakeLists.txt - cpp/apps/primitives_gallery/playground/tables/data_table.cpp - cpp/apps/primitives_gallery/playground/tables/data_table_logic.cpp - cpp/apps/primitives_gallery/playground/tables/data_table_logic.h - cpp/apps/primitives_gallery/playground/tables/self_test.cpp - cpp/apps/primitives_gallery/playground/tables/tql.cpp - cpp/apps/primitives_gallery/playground/tables/viz.cpp - cpp/apps/primitives_gallery/playground/tables/viz.h - ... Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-13 00:50:34 +02:00
parent a2bbf23374
commit e3c8979e8d
189 changed files with 18964 additions and 330 deletions
@@ -0,0 +1,154 @@
+"""Tests para vault_dedupe_report."""
+
+from __future__ import annotations
+
+import os
+import sqlite3
+import sys
+from pathlib import Path
+
+import pytest
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), ".."))
+from vault_dedupe_report import vault_dedupe_report
+
+
+# ---------------------------------------------------------------------------
+# Helpers
+# ---------------------------------------------------------------------------
+
+def _make_db(tmp_path: Path, rows: list[tuple]) -> Path:
+    """Crea vault_index.db con la tabla files y las filas dadas.
+
+    rows: lista de (rel_path, size, sha256)
+    """
+    db_path = tmp_path / "vault_index.db"
+    conn = sqlite3.connect(str(db_path))
+    conn.execute(
+        """
+        CREATE TABLE files (
+            rel_path  TEXT PRIMARY KEY,
+            size      INTEGER,
+            mtime     REAL,
+            sha256    TEXT,
+            mime      TEXT,
+            ext       TEXT,
+            bucket    TEXT,
+            sub_bucket TEXT,
+            indexed_at REAL
+        );
+        """
+    )
+    conn.executemany(
+        "INSERT INTO files (rel_path, size, sha256) VALUES (?, ?, ?);",
+        rows,
+    )
+    conn.commit()
+    conn.close()
+    return db_path
+
+
+# ---------------------------------------------------------------------------
+# Tests
+# ---------------------------------------------------------------------------
+
+def test_no_duplicates(tmp_path):
+    """test_no_duplicates — 3 archivos con sha256 distintos -> groups=[]."""
+    _make_db(tmp_path, [
+        ("a/file1.txt", 100, "aaa111"),
+        ("a/file2.txt", 200, "bbb222"),
+        ("a/file3.txt", 300, "ccc333"),
+    ])
+    result = vault_dedupe_report(str(tmp_path), db_path=str(tmp_path / "vault_index.db"))
+
+    assert result["groups"] == []
+    assert result["total_groups"] == 0
+    assert result["total_duplicates"] == 0
+    assert result["total_reclaimable_bytes"] == 0
+    assert result["scanned_files"] == 3
+    assert result["vault_path"] == str(tmp_path)
+
+
+def test_basic_duplicates(tmp_path):
+    """test_basic_duplicates — 2 archivos mismo sha256 -> 1 group, count=2, reclaimable=size."""
+    _make_db(tmp_path, [
+        ("data/orig.jpg", 500, "deadbeef"),
+        ("backup/orig.jpg", 500, "deadbeef"),
+    ])
+    result = vault_dedupe_report(str(tmp_path), db_path=str(tmp_path / "vault_index.db"))
+
+    assert result["total_groups"] == 1
+    assert result["total_duplicates"] == 1
+    assert result["total_reclaimable_bytes"] == 500
+
+    g = result["groups"][0]
+    assert g["sha256"] == "deadbeef"
+    assert g["size"] == 500
+    assert g["count"] == 2
+    assert g["reclaimable_bytes"] == 500
+    assert sorted(g["files"]) == ["backup/orig.jpg", "data/orig.jpg"]
+
+
+def test_three_in_group(tmp_path):
+    """test_three_in_group — 3 archivos mismo sha256 -> count=3, reclaimable=size*2."""
+    size = 1000
+    _make_db(tmp_path, [
+        ("a/f1.bin", size, "cafebabe"),
+        ("b/f2.bin", size, "cafebabe"),
+        ("c/f3.bin", size, "cafebabe"),
+    ])
+    result = vault_dedupe_report(str(tmp_path), db_path=str(tmp_path / "vault_index.db"))
+
+    assert result["total_groups"] == 1
+    assert result["total_duplicates"] == 2
+    assert result["total_reclaimable_bytes"] == size * 2
+
+    g = result["groups"][0]
+    assert g["count"] == 3
+    assert g["reclaimable_bytes"] == size * 2
+    assert g["files"] == sorted(["a/f1.bin", "b/f2.bin", "c/f3.bin"])
+
+
+def test_min_size_filter(tmp_path):
+    """test_min_size_filter — duplicados de tamano 50, min_size=100 -> groups=[]."""
+    _make_db(tmp_path, [
+        ("x/small1.txt", 50, "tiny123"),
+        ("y/small2.txt", 50, "tiny123"),
+    ])
+    result = vault_dedupe_report(
+        str(tmp_path),
+        min_size=100,
+        db_path=str(tmp_path / "vault_index.db"),
+    )
+
+    assert result["groups"] == []
+    assert result["total_groups"] == 0
+    assert result["total_reclaimable_bytes"] == 0
+    assert result["scanned_files"] == 0
+
+
+def test_multiple_groups_ordered(tmp_path):
+    """test_multiple_groups_ordered — 2 grupos con distinto ahorro -> orden DESC."""
+    # grupo A: 2 copias de 200 bytes -> reclaimable=200
+    # grupo B: 3 copias de 500 bytes -> reclaimable=1000
+    # el grupo B debe salir primero
+    _make_db(tmp_path, [
+        ("p/a1.dat", 200, "groupA"),
+        ("q/a2.dat", 200, "groupA"),
+        ("r/b1.dat", 500, "groupB"),
+        ("s/b2.dat", 500, "groupB"),
+        ("t/b3.dat", 500, "groupB"),
+        ("u/uniq.dat", 999, "unique1"),
+    ])
+    result = vault_dedupe_report(str(tmp_path), db_path=str(tmp_path / "vault_index.db"))
+
+    assert result["total_groups"] == 2
+    assert result["total_duplicates"] == 3          # (2-1) + (3-1)
+    assert result["total_reclaimable_bytes"] == 1200  # 200 + 1000
+    assert result["scanned_files"] == 6              # 6 filas con sha256 != '' (incluye el unico)
+
+    # Primer grupo debe ser el de mayor ahorro (B: 1000)
+    assert result["groups"][0]["sha256"] == "groupB"
+    assert result["groups"][0]["reclaimable_bytes"] == 1000
+    assert result["groups"][1]["sha256"] == "groupA"
+    assert result["groups"][1]["reclaimable_bytes"] == 200
@@ -0,0 +1,153 @@
+"""Tests para vault_knowledge_parse."""
+
+from __future__ import annotations
+
+import os
+import sqlite3
+import sys
+from pathlib import Path
+
+import pytest
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), ".."))
+from vault_knowledge_parse import vault_knowledge_parse
+
+
+def _make_vault(tmp: Path) -> tuple[Path, Path]:
+    """Crea un vault mínimo con vault_index.db."""
+    db = tmp / "vault_index.db"
+    conn = sqlite3.connect(str(db))
+    conn.executescript(
+        """
+        CREATE TABLE IF NOT EXISTS files (
+            rowid INTEGER PRIMARY KEY AUTOINCREMENT,
+            rel_path TEXT UNIQUE NOT NULL,
+            size_bytes INTEGER,
+            ext TEXT
+        );
+        CREATE VIRTUAL TABLE IF NOT EXISTS files_fts
+            USING fts5(rel_path, content_text, content='', contentless_delete=1);
+        CREATE TABLE IF NOT EXISTS knowledge_docs (
+            rel_path TEXT PRIMARY KEY,
+            title TEXT,
+            frontmatter_json TEXT,
+            headings_json TEXT,
+            parsed_at INTEGER
+        );
+        """
+    )
+    conn.commit()
+    conn.close()
+    return tmp, db
+
+
+def _insert_file_entry(db: Path, rel_path: str):
+    conn = sqlite3.connect(str(db))
+    conn.execute(
+        "INSERT OR IGNORE INTO files(rel_path, size_bytes, ext) VALUES (?, 0, '.md')",
+        (rel_path,),
+    )
+    conn.commit()
+    conn.close()
+
+
+def test_md_with_frontmatter(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/guia.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text(
+        "---\ntitle: Mi Guía\nauthor: Lucas\n---\n\n# Mi Guía\n\nContenido del documento.\n",
+        encoding="utf-8",
+    )
+    _insert_file_entry(db, rel)
+
+    result = vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    assert result["title"] == "Mi Guía"
+    assert result["frontmatter"]["author"] == "Lucas"
+    assert "Contenido del documento" in result["content_text"]
+    assert result["persisted"] is True
+
+
+def test_md_no_frontmatter(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/sin_fm.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text("# Título\n\nCuerpo sin frontmatter.\n", encoding="utf-8")
+    _insert_file_entry(db, rel)
+
+    result = vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    assert result["frontmatter"] == {}
+    assert result["title"] == "Título"
+    assert "Cuerpo sin frontmatter" in result["content_text"]
+
+
+def test_md_title_from_h1(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/title_h1.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text("# Primer H1\n\nAlgún texto.\n", encoding="utf-8")
+    _insert_file_entry(db, rel)
+
+    result = vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    assert result["title"] == "Primer H1"
+
+
+def test_md_title_from_filename(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/nombre_archivo.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text("Solo texto sin headings ni frontmatter.\n", encoding="utf-8")
+    _insert_file_entry(db, rel)
+
+    result = vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    assert result["title"] == "nombre_archivo"
+
+
+def test_md_headings_levels(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/headings.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text(
+        "# H1 Título\n\nTexto.\n\n## H2 Sección\n\n### H3 Subsección\n\n## H2 Otra\n",
+        encoding="utf-8",
+    )
+    _insert_file_entry(db, rel)
+
+    result = vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    headings = result["headings"]
+    assert len(headings) == 4
+    levels = [h["level"] for h in headings]
+    assert levels == [1, 2, 3, 2]
+    texts = [h["text"] for h in headings]
+    assert "H1 Título" in texts
+    assert "H2 Sección" in texts
+    assert "H3 Subsección" in texts
+
+
+def test_md_persists_to_fts(tmp_path):
+    vault, db = _make_vault(tmp_path)
+    rel = "docs/fts_md.md"
+    md = vault / rel
+    md.parent.mkdir(parents=True, exist_ok=True)
+    md.write_text("# Documento FTS\n\nPalabra clave: xenolito.\n", encoding="utf-8")
+    _insert_file_entry(db, rel)
+
+    vault_knowledge_parse(str(vault), rel, db_path=str(db))
+
+    conn = sqlite3.connect(str(db))
+    # FTS5 contentless: no permite SELECT directo, usar MATCH
+    row = conn.execute(
+        "SELECT rowid FROM files_fts WHERE files_fts MATCH 'xenolito'",
+    ).fetchone()
+    conn.close()
+    assert row is not None, "FTS no encontró 'xenolito'"