models.py 2.7 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667
  1. from django.db import models
  2. from abid_utils.models import ABIDModel, ABIDField, AutoDateTimeField, ModelWithHealthStats
  3. from django.conf import settings
  4. class Persona(ABIDModel, ModelWithHealthStats):
  5. """Aka a "SessionType", its a template for a crawler browsing session containing some config."""
  6. abid_prefix = 'prs_'
  7. abid_ts_src = 'self.created_at'
  8. abid_uri_src = 'self.name'
  9. abid_subtype_src = 'self.created_by'
  10. abid_rand_src = 'self.id'
  11. id = models.UUIDField(primary_key=True, default=None, null=False, editable=False, unique=True, verbose_name='ID')
  12. abid = ABIDField(prefix=abid_prefix)
  13. created_by = models.ForeignKey(settings.AUTH_USER_MODEL, on_delete=models.CASCADE, default=None, null=False)
  14. created_at = AutoDateTimeField(default=None, null=False, db_index=True)
  15. modified_at = models.DateTimeField(auto_now=True)
  16. name = models.CharField(max_length=100, blank=False, null=False, editable=False)
  17. persona_dir = models.FilePathField(path=settings.PERSONAS_DIR, allow_files=False, allow_folders=True, blank=True, null=False, editable=False)
  18. config = models.JSONField(default=dict)
  19. # e.g. {
  20. # USER_AGENT: 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36',
  21. # COOKIES_TXT_FILE: '/path/to/cookies.txt',
  22. # CHROME_USER_DATA_DIR: '/path/to/chrome/user/data/dir',
  23. # CHECK_SSL_VALIDITY: False,
  24. # SAVE_ARCHIVE_DOT_ORG: True,
  25. # CHROME_BINARY: 'chromium'
  26. # ...
  27. # }
  28. # domain_allowlist = models.CharField(max_length=1024, blank=True, null=False, default='')
  29. # domain_denylist = models.CharField(max_length=1024, blank=True, null=False, default='')
  30. class Meta:
  31. verbose_name = 'Session Type'
  32. verbose_name_plural = 'Session Types'
  33. unique_together = (('created_by', 'name'),)
  34. def clean(self):
  35. self.persona_dir = settings.PERSONAS_DIR / self.name
  36. assert self.persona_dir == settings.PERSONAS_DIR / self.name, f'Persona dir {self.persona_dir} must match settings.PERSONAS_DIR / self.name'
  37. # make sure config keys all exist in FLAT_CONFIG
  38. # make sure config values all match expected types
  39. pass
  40. def save(self, *args, **kwargs):
  41. self.full_clean()
  42. # make sure basic file structure is present in persona_dir:
  43. # - PERSONAS_DIR / self.name /
  44. # - chrome_profile/
  45. # - chrome_downloads/
  46. # - chrome_extensions/
  47. # - cookies.txt
  48. # - auth.json
  49. # - config.json # json dump of the model
  50. super().save(*args, **kwargs)