init commit

db/mysql_import.py

@@ -0,0 +1,49 @@
+import csv
+import os
+
+def import_csv_to_mysql(csv_path, host='localhost', port=3306, user='root', password='', database='crawler_tiktok', table='comments'):
+    try:
+        import pymysql
+    except Exception:
+        print('missing dependency: pip install pymysql', flush=True)
+        raise SystemExit(1)
+    if not os.path.exists(csv_path):
+        print('csv not found: ' + csv_path, flush=True)
+        raise SystemExit(1)
+    conn = pymysql.connect(host=host, port=int(port), user=user, password=password, database=database, charset='utf8mb4')
+    cur = conn.cursor()
+    cur.execute(f"CREATE TABLE IF NOT EXISTS `{table}` (\n        `id` BIGINT AUTO_INCREMENT PRIMARY KEY,\n        `username` VARCHAR(255),\n        `text` TEXT\n      ) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci")
+    rows = []
+    with open(csv_path, 'r', encoding='utf-8', newline='') as f:
+        r = csv.reader(f)
+        first = True
+        for row in r:
+            if first and row and row[0].lower() == 'username':
+                first = False
+                continue
+            first = False
+            if not row:
+                continue
+            username = row[0] if len(row) > 0 else ''
+            text = row[1] if len(row) > 1 else ''
+            rows.append((username, text))
+    if rows:
+        cur.executemany(f"INSERT INTO `{table}` (`username`,`text`) VALUES (%s,%s)", rows)
+        conn.commit()
+    cur.close()
+    conn.close()
+    print(f"inserted={len(rows)}", flush=True)
+
+def create_database_if_not_exists(host='localhost', port=3306, user='root', password='', database='yunque'):
+    try:
+        import pymysql
+    except Exception:
+        print('missing dependency: pip install pymysql', flush=True)
+        raise SystemExit(1)
+    conn = pymysql.connect(host=host, port=int(port), user=user, password=password, charset='utf8mb4')
+    cur = conn.cursor()
+    cur.execute(f"CREATE DATABASE IF NOT EXISTS `{database}` CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci")
+    conn.commit()
+    cur.close()
+    conn.close()
+    print(f"database_ready={database}", flush=True)