settings.py 3.2 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899
  1. import os
  2. from dotenv import load_dotenv
  3. load_dotenv()
  4. DATABASE_TYPE = os.getenv('DB_TYPE', 'sqlite').lower()
  5. DATABASE_CONFIG = {
  6. 'type': DATABASE_TYPE,
  7. 'sqlite': {
  8. 'path': os.getenv('SQLITE_PATH', 'data/price_crawler.db'),
  9. },
  10. 'mysql': {
  11. 'host': os.getenv('DB_HOST', 'localhost'),
  12. 'port': int(os.getenv('DB_PORT', 3306)),
  13. 'user': os.getenv('DB_USER', 'root'),
  14. 'password': os.getenv('DB_PASSWORD', ''),
  15. 'database': os.getenv('DB_NAME', 'price_crawler'),
  16. 'charset': 'utf8mb4'
  17. }
  18. }
  19. CRAWLER_CONFIG = {
  20. 'taobao': {
  21. 'base_url': 'https://s.taobao.com',
  22. 'search_url': 'https://s.taobao.com/search',
  23. 'headers': {
  24. 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
  25. 'Referer': 'https://www.taobao.com',
  26. 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
  27. 'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8',
  28. },
  29. 'timeout': 30,
  30. 'retry_times': 3,
  31. 'retry_delay': 2,
  32. 'delay_range': (1, 3),
  33. },
  34. 'jd': {
  35. 'base_url': 'https://search.jd.com',
  36. 'search_url': 'https://search.jd.com/Search',
  37. 'headers': {
  38. 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
  39. 'Referer': 'https://www.jd.com',
  40. 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
  41. 'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8',
  42. },
  43. 'timeout': 30,
  44. 'retry_times': 3,
  45. 'retry_delay': 2,
  46. 'delay_range': (1, 3),
  47. },
  48. 'alibaba1688': {
  49. 'base_url': 'https://s.1688.com',
  50. 'search_url': 'https://s.1688.com/selloffer/offer_search.htm',
  51. 'headers': {
  52. 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
  53. 'Referer': 'https://www.1688.com',
  54. 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
  55. 'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8',
  56. },
  57. 'timeout': 30,
  58. 'retry_times': 3,
  59. 'retry_delay': 2,
  60. 'delay_range': (1, 3),
  61. }
  62. }
  63. LOGGING_CONFIG = {
  64. 'version': 1,
  65. 'disable_existing_loggers': False,
  66. 'formatters': {
  67. 'standard': {
  68. 'format': '%(asctime)s - %(name)s - %(levelname)s - %(message)s'
  69. },
  70. },
  71. 'handlers': {
  72. 'console': {
  73. 'class': 'logging.StreamHandler',
  74. 'level': 'INFO',
  75. 'formatter': 'standard'
  76. },
  77. 'file': {
  78. 'class': 'logging.handlers.RotatingFileHandler',
  79. 'level': 'INFO',
  80. 'formatter': 'standard',
  81. 'filename': 'logs/crawler.log',
  82. 'maxBytes': 1024*1024*10,
  83. 'backupCount': 5,
  84. 'encoding': 'utf-8'
  85. }
  86. },
  87. 'loggers': {
  88. '': {
  89. 'handlers': ['console', 'file'],
  90. 'level': 'INFO',
  91. 'propagate': True
  92. }
  93. }
  94. }