add sequential migration system with version tracking

Replace the simple schema.sql runner with a proper sequential
migration system that tracks applied versions in a _migrations table.
Absorb scripts/migrate_to_paddle.py as versions/0001.

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

padelnomics/scripts/migrate_to_paddle.py

@@ -1,19 +0,0 @@
-"""One-time migration: rename lemonsqueezy columns to paddle."""
-import sqlite3
-import sys
-
-db_path = sys.argv[1] if len(sys.argv) > 1 else "data/app.db"
-conn = sqlite3.connect(db_path)
-conn.execute(
-    "ALTER TABLE subscriptions RENAME COLUMN lemonsqueezy_customer_id TO paddle_customer_id"
-)
-conn.execute(
-    "ALTER TABLE subscriptions RENAME COLUMN lemonsqueezy_subscription_id TO paddle_subscription_id"
-)
-conn.execute("DROP INDEX IF EXISTS idx_subscriptions_provider")
-conn.execute(
-    "CREATE INDEX IF NOT EXISTS idx_subscriptions_provider ON subscriptions(paddle_subscription_id)"
-)
-conn.commit()
-conn.close()
-print(f"Migration complete: {db_path}")

padelnomics/src/padelnomics/migrations/migrate.py

@@ -1,51 +1,103 @@
 """
-Simple migration runner. Runs schema.sql against the database.
+Sequential migration runner.
+
+- Runs schema.sql (idempotent CREATE IF NOT EXISTS for fresh DBs)
+- Scans versions/ for NNNN_*.py files and runs unapplied ones in order
+- Fresh DBs: marks all versions as applied without running them
+  (schema.sql already contains the final schema)
 """
+
+import importlib
 import os
+import re
 import sqlite3
 import sys
 from pathlib import Path
 
-# Add parent to path for imports
 sys.path.insert(0, str(Path(__file__).parent.parent.parent.parent))
 
 from dotenv import load_dotenv
 
 load_dotenv()
 
+VERSIONS_DIR = Path(__file__).parent / "versions"
+VERSION_RE = re.compile(r"^(\d{4})_.+\.py$")
+
+
+def _discover_versions():
+    """Return sorted list of (name, module_path) for version files."""
+    if not VERSIONS_DIR.is_dir():
+        return []
+    versions = []
+    for f in sorted(VERSIONS_DIR.iterdir()):
+        if VERSION_RE.match(f.name):
+            versions.append(f.stem)  # e.g. "0001_rename_ls_to_paddle"
+    return versions
+
+
+def _is_fresh_db(conn):
+    """A DB is fresh if it has no application tables at all."""
+    row = conn.execute(
+        "SELECT 1 FROM sqlite_master WHERE type='table'"
+        " AND name NOT LIKE 'sqlite_%'"
+    ).fetchone()
+    return row is None
+
 
 def migrate():
-    """Run migrations."""
-    # Get database path from env or default
     db_path = os.getenv("DATABASE_PATH", "data/app.db")
-    
-    # Ensure directory exists
     Path(db_path).parent.mkdir(parents=True, exist_ok=True)
-    
-    # Read schema
-    schema_path = Path(__file__).parent / "schema.sql"
-    schema = schema_path.read_text()
-    
-    # Connect and execute
+
     conn = sqlite3.connect(db_path)
-    
-    # Enable WAL mode
     conn.execute("PRAGMA journal_mode=WAL")
     conn.execute("PRAGMA foreign_keys=ON")
-    
-    # Run schema
+
+    is_fresh = _is_fresh_db(conn)
+
+    # schema.sql is always idempotent (CREATE IF NOT EXISTS)
+    schema = (Path(__file__).parent / "schema.sql").read_text()
     conn.executescript(schema)
-    conn.commit()
-    
-    print(f"✓ Migrations complete: {db_path}")
-    
-    # Show tables
+
+    versions = _discover_versions()
+    applied = {
+        row[0]
+        for row in conn.execute("SELECT name FROM _migrations").fetchall()
+    }
+    pending = [v for v in versions if v not in applied]
+
+    if is_fresh:
+        # Fresh DB — schema.sql already created final schema.
+        # Record all versions as applied without executing them.
+        for name in pending:
+            conn.execute("INSERT INTO _migrations (name) VALUES (?)", (name,))
+        conn.commit()
+        print(f"✓ Fresh database initialised: {db_path}")
+        if pending:
+            print(f"  Recorded {len(pending)} migration(s) as already applied")
+    elif pending:
+        for name in pending:
+            print(f"  Applying {name}...")
+            mod = importlib.import_module(
+                f"padelnomics.migrations.versions.{name}"
+            )
+            mod.up(conn)
+            conn.execute(
+                "INSERT INTO _migrations (name) VALUES (?)", (name,)
+            )
+        conn.commit()
+        print(f"✓ Applied {len(pending)} migration(s): {db_path}")
+    else:
+        print(f"✓ All migrations already applied: {db_path}")
+
+    # Show tables (excluding internal sqlite/fts tables)
     cursor = conn.execute(
-        "SELECT name FROM sqlite_master WHERE type='table' ORDER BY name"
+        "SELECT name FROM sqlite_master WHERE type='table'"
+        " AND name NOT LIKE 'sqlite_%'"
+        " ORDER BY name"
     )
     tables = [row[0] for row in cursor.fetchall()]
     print(f"  Tables: {', '.join(tables)}")
-    
+
     conn.close()
 
 

padelnomics/src/padelnomics/migrations/schema.sql

@@ -1,6 +1,13 @@
 -- Padelnomics Database Schema
 -- Run with: python -m padelnomics.migrations.migrate
 
+-- Migration tracking
+CREATE TABLE IF NOT EXISTS _migrations (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    name TEXT UNIQUE NOT NULL,
+    applied_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
 -- Users
 CREATE TABLE IF NOT EXISTS users (
     id INTEGER PRIMARY KEY AUTOINCREMENT,

padelnomics/src/padelnomics/migrations/versions/0001_rename_ls_to_paddle.py

@@ -0,0 +1,17 @@
+"""Rename lemonsqueezy columns to paddle."""
+
+
+def up(conn):
+    conn.execute(
+        "ALTER TABLE subscriptions"
+        " RENAME COLUMN lemonsqueezy_customer_id TO paddle_customer_id"
+    )
+    conn.execute(
+        "ALTER TABLE subscriptions"
+        " RENAME COLUMN lemonsqueezy_subscription_id TO paddle_subscription_id"
+    )
+    conn.execute("DROP INDEX IF EXISTS idx_subscriptions_provider")
+    conn.execute(
+        "CREATE INDEX IF NOT EXISTS idx_subscriptions_provider"
+        " ON subscriptions(paddle_subscription_id)"
+    )