22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572 | class OneDriveReader(BasePydanticReader):
"""
Microsoft OneDrive reader.
Initializes a new instance of the OneDriveReader.
:param client_id: The Application (client) ID for the app registered in the Azure Entra (formerly Azure Active directory) portal with MS Graph permission "Files.Read.All".
:param tenant_id: The Directory (tenant) ID of the Azure Active Directory (AAD) tenant the app is registered with.
Defaults to "consumers" for multi-tenant applications and OneDrive personal.
:param client_secret: The Application Secret for the app registered in the Azure portal.
If provided, the MSAL client credential flow will be used for authentication (ConfidentialClientApplication).
If not provided, interactive authentication will be used (Not recommended for CI/CD or scenarios where manual interaction for authentication is not feasible).
Required for App authentication.
:param userprinciplename: The user principal name (normally organization provided email) whose OneDrive will be accessed. Required for App authentication. Will be used if the
parameter is not provided when calling load_data().
:param folder_id: The folder ID of the folder to fetch from OneDrive. Will be used if the parameter is not provided when calling load_data().
:param file_ids: A list of file IDs of files to fetch from OneDrive. Will be used if the parameter is not provided when calling load_data().
:param folder_path (str, optional): The relative path of the OneDrive folder to download. If provided, files within the folder are downloaded. Will be used if the parameter is
not provided when calling load_data().
:param file_paths (List[str], optional): List of specific file paths to download. Will be used if the parameter is not provided when calling load_data().
:param file_extractor (Optional[Dict[str, BaseReader]]): A mapping of file extension to a BaseReader class that specifies how to convert that file to text.
See `SimpleDirectoryReader` for more details.
For interactive authentication to work, a browser is used to authenticate, hence the registered application should have a redirect URI set to 'https://localhost'
for mobile and native applications.
"""
client_id: str
client_secret: Optional[str] = None
tenant_id: Optional[str] = None
userprincipalname: Optional[str] = None
folder_id: Optional[str] = None
file_ids: Optional[List[str]] = None
folder_path: Optional[str] = None
file_paths: Optional[List[str]] = None
file_extractor: Optional[Dict[str, Union[str, BaseReader]]] = Field(
default=None, exclude=True
)
_is_interactive_auth = PrivateAttr(False)
_authority = PrivateAttr()
_downloaded_files_metadata = PrivateAttr({})
def __init__(
self,
client_id: str,
client_secret: Optional[str] = None,
tenant_id: Optional[str] = "consumers",
userprincipalname: Optional[str] = None,
folder_id: Optional[str] = None,
file_ids: Optional[List[str]] = None,
folder_path: Optional[str] = None,
file_paths: Optional[List[str]] = None,
file_extractor: Optional[Dict[str, Union[str, BaseReader]]] = None,
**kwargs,
) -> None:
self._is_interactive_auth = not client_secret
self._authority = f"https://login.microsoftonline.com/{tenant_id}/"
super().__init__(
client_id=client_id,
client_secret=client_secret,
tenant_id=tenant_id,
userprincipalname=userprincipalname,
folder_id=folder_id,
file_ids=file_ids,
folder_path=folder_path,
file_paths=file_paths,
file_extractor=file_extractor,
**kwargs,
)
def _authenticate_with_msal(self) -> Any:
"""Authenticate with MSAL.
For interactive authentication to work, a browser is used to authenticate, hence the registered application should have a redirect URI set to 'localhost'
for mobile and native applications.
"""
import msal
result = None
if self._is_interactive_auth:
logger.debug("Starting user authentication...")
app = msal.PublicClientApplication(
self.client_id, authority=self._authority
)
# The acquire_token_interactive method will open the default web browser
# for the interactive part of the OAuth2 flow. The registered application should have a redirect URI set to 'https://localhost'
# under mobile and native applications.
result = app.acquire_token_interactive(SCOPES)
else:
logger.debug("Starting app authentication...")
app = msal.ConfidentialClientApplication(
self.client_id,
authority=self._authority,
client_credential=self.client_secret,
)
result = app.acquire_token_for_client(scopes=CLIENTCREDENTIALSCOPES)
if "access_token" in result:
logger.debug("Authentication is successful...")
return result["access_token"]
else:
logger.error(result.get("error"))
logger.error(result.get("error_description"))
logger.error(result.get("correlation_id"))
raise Exception(result.get("error"))
def _construct_endpoint(
self,
item_ref: str,
isRelativePath: bool,
isFile: bool,
userprincipalname: Optional[str] = None,
) -> str:
"""
Constructs the appropriate OneDrive API endpoint based on the provided parameters.
Parameters:
item_ref (str): The reference to the item; could be an item ID or a relative path.
isRelativePath (bool): A boolean indicating whether the item_ref is a relative path.
isFile (bool): A boolean indicating whether the target is a file.
userprincipalname (str, optional): The user principal name; used if authentication is not interactive. Defaults to None.
Returns:
str: A string representing the constructed endpoint.
"""
if not self._is_interactive_auth and not userprincipalname:
raise Exception(
"userprincipalname cannot be empty for App authentication. Provide the userprincipalname (usually email) of the user whose OneDrive will be accessed."
)
endpoint = "https://graph.microsoft.com/v1.0/"
# Update the base endpoint based on the authentication method
if self._is_interactive_auth:
endpoint += "me/drive"
else:
endpoint += f"users/{userprincipalname}/drive"
# Update the endpoint for relative paths or item IDs
if isRelativePath:
endpoint += f"/root:/{item_ref}"
else:
endpoint += f"/items/{item_ref}"
# If the target is not a file, adjust the endpoint to retrieve children of a folder
if not isFile:
endpoint += ":/children" if isRelativePath else "/children"
logger.info(f"API Endpoint determined: {endpoint}")
return endpoint
def _get_items_in_drive_with_maxretries(
self,
access_token: str,
item_ref: Optional[str] = "root",
max_retries: int = 3,
userprincipalname: Optional[str] = None,
isFile: bool = False,
isRelativePath=False,
) -> Any:
"""
Retrieves items from a drive using Microsoft Graph API.
Parameters:
access_token (str): Access token for API calls.
item_ref (Optional[str]): Specific item ID/path or root for root folder.
max_retries (int): Max number of retries on rate limit or server errors.
userprincipalname: str value indicating the userprincipalname (usually organization-provided email) whose OneDrive will be accessed. Required for App authentication.
isFile: bool value to indicate if to query file or folder
isRelativePath: bool value to indicate if to query file or folder using relative path
Returns:
dict/None: JSON response or None after max retries.
Raises:
Exception: On non-retriable status code.
"""
endpoint = self._construct_endpoint(
item_ref, isRelativePath, isFile, userprincipalname
)
headers = {"Authorization": f"Bearer {access_token}"}
retries = 0
while retries < max_retries:
response = requests.get(endpoint, headers=headers)
if response.status_code == 200:
return response.json()
# Check for Ratelimit error, this can happen if you query endpoint recursively
# very frequently for large amount of file
elif response.status_code in (429, *range(500, 600)):
logger.warning(
f"Retrying {retries+1} in {retries+1} secs. Status code: {response.status_code}"
)
retries += 1
time.sleep(retries) # Exponential back-off
else:
raise Exception(
f"API request to download {item_ref} failed with status code: {response.status_code}, message: {response.content}"
)
logger.error(f"Failed after {max_retries} attempts.")
return None
def _download_file_by_url(self, item: Dict[str, Any], local_dir: str) -> str:
"""
Downloads a file from OneDrive using the provided item's download URL.
Parameters:
- item (Dict[str, str]): Dictionary containing file metadata and download URL.
- local_dir (str): Local directory where the file should be saved.
Returns:
- str: The file path of the download file
"""
# Extract download URL and filename from the provided item.
file_download_url = item["@microsoft.graph.downloadUrl"]
file_name = item["name"]
# Download the file.
file_data = requests.get(file_download_url)
# Save the downloaded file to the specified local directory.
file_path = os.path.join(local_dir, file_name)
with open(file_path, "wb") as f:
f.write(file_data.content)
return file_path
def _extract_metadata_for_file(self, item: Dict[str, Any]) -> Dict[str, str]:
"""
Extracts metadata related to the file.
Parameters:
- item (Dict[str, str]): Dictionary containing file metadata.
Returns:
- Dict[str, str]: A dictionary containing the extracted metadata.
"""
# Extract the required metadata for file.
created_by = item.get("createdBy", {})
modified_by = item.get("lastModifiedBy", {})
return {
"file_id": item.get("id"),
"file_name": item.get("name"),
"created_by_user": created_by.get("user", {}).get("displayName"),
"created_by_app": created_by.get("application", {}).get("displayName"),
"created_dateTime": item.get("createdDateTime"),
"last_modified_by_user": modified_by.get("user", {}).get("displayName"),
"last_modified_by_app": modified_by.get("application", {}).get(
"displayName"
),
"last_modified_datetime": item.get("lastModifiedDateTime"),
}
def _check_approved_mimetype_and_download_file(
self,
item: Dict[str, Any],
local_dir: str,
mime_types: Optional[List[str]] = None,
):
"""
Checks files based on MIME types and download the accepted files.
:param item: dict, a dictionary representing a file item, must contain 'file' and 'mimeType' keys.
:param local_dir: str, the local directory to download files to.
:param mime_types: list, a list of accepted MIME types. If None or empty, all file types are accepted.
:return: dict, a dictionary containing metadata of downloaded files.
"""
metadata = {}
# Convert accepted MIME types to lowercase for case-insensitive comparison
accepted_mimetypes = (
[mimetype.lower() for mimetype in mime_types] if mime_types else ["*"]
)
# Check if the item's MIME type is among the accepted MIME types
is_accepted_mimetype = (
"*" in accepted_mimetypes
or item["file"]["mimeType"].lower() in accepted_mimetypes
)
if is_accepted_mimetype:
# It's a file with an accepted MIME type; download and extract metadata
file_path = self._download_file_by_url(
item, local_dir
) # Assuming this method is implemented
metadata[file_path] = self._extract_metadata_for_file(
item
) # Assuming this method is implemented
else:
# Log a debug message for files that are ignored due to an invalid MIME type
logger.debug(
f"Ignoring file '{item['name']}' as its MIME type does not match the accepted types."
)
return metadata
def _connect_download_and_return_metadata(
self,
access_token: str,
local_dir: str,
item_id: str = None,
include_subfolders: bool = True,
mime_types: Optional[List[str]] = None,
userprincipalname: Optional[str] = None,
isRelativePath=False,
) -> Any:
"""
Recursively download files from OneDrive, starting from the specified item_id or the root.
Parameters:
- access_token (str): Token for authorization.
- local_dir (str): Local directory to store downloaded files.
- item_id (str, optional): ID of the specific item (folder/file) to start from. If None, starts from the root.
- include_subfolders (bool, optional): Whether to include subfolders. Defaults to True.
- mime_types(List[str], optional): the mimeTypes you want to allow e.g.: "application/pdf", default is None which loads all files
- userprincipalname (str): The userprincipalname(normally organization provided email id) whose ondrive needs to be accessed. Mandatory for App authentication scenarios.
- isRelativePath (bool): Value to indicate if to query file/folder using relative path
Returns:
- dict: Dictionary of file paths and their corresponding metadata.
Raises:
- Exception: If items can't be retrieved for the current item.
"""
data = self._get_items_in_drive_with_maxretries(
access_token,
item_id,
userprincipalname=userprincipalname,
isRelativePath=isRelativePath,
)
if data:
metadata = {}
for item in data["value"]:
if (
"folder" in item and include_subfolders
): # It's a folder; traverse if flag is set
subfolder_metadata = self._connect_download_and_return_metadata(
access_token,
local_dir,
item["id"],
include_subfolders,
mime_types=mime_types,
userprincipalname=userprincipalname,
)
metadata.update(subfolder_metadata) # Merge metadata
elif "file" in item:
file_metadata = self._check_approved_mimetype_and_download_file(
item, local_dir, mime_types
)
metadata.update(file_metadata)
return metadata
# No data received; raise exception
current_item = item_id if item_id else "RootFolder"
raise Exception(f"Unable to retrieve items for: {current_item}")
def _init_download_and_get_metadata(
self,
temp_dir: str,
folder_id: Optional[str] = None,
file_ids: Optional[List[str]] = None,
folder_path: Optional[str] = None,
file_paths: Optional[List[str]] = None,
recursive: bool = False,
mime_types: Optional[List[str]] = None,
userprincipalname: Optional[str] = None,
) -> None:
"""
Download files from OneDrive based on specified folder or file IDs/Paths.
Parameters:
- temp_dir (str): The temporary directory where files will be downloaded.
- folder_id (str, optional): The ID of the OneDrive folder to download. If provided, files within the folder are downloaded.
- file_ids (List[str], optional): List of specific file IDs to download.
- folder_path (str, optional): The relative path of the OneDrive folder to download. If provided, files within the folder are downloaded.
- file_paths (List[str], optional): List of specific file paths to download.
- recursive (bool): Flag indicating whether to download files from subfolders if a folder_id is provided.
- mime_types(List[str], optional): the mimeTypes you want to allow e.g.: "application/pdf", default is None which loads all files
- userprincipalname (str): The userprincipalname (normally organization-provided email) whose OneDrive will be accessed. Required for App authentication.
"""
access_token = self._authenticate_with_msal()
is_download_from_root = True
downloaded_files_metadata = {}
# If a folder_id is provided, download files from the folder
if folder_id:
is_download_from_root = False
folder_metadata = self._connect_download_and_return_metadata(
access_token,
temp_dir,
folder_id,
recursive,
mime_types=mime_types,
userprincipalname=userprincipalname,
)
downloaded_files_metadata.update(folder_metadata)
# Download files using the provided file IDs
if file_ids:
is_download_from_root = False
for file_id in file_ids or []:
item = self._get_items_in_drive_with_maxretries(
access_token,
file_id,
userprincipalname=userprincipalname,
isFile=True,
)
file_metadata = self._check_approved_mimetype_and_download_file(
item, temp_dir, mime_types
)
downloaded_files_metadata.update(file_metadata)
# If a folder_path is provided, download files from the folder
if folder_path:
is_download_from_root = False
folder_metadata = self._connect_download_and_return_metadata(
access_token,
temp_dir,
folder_path,
recursive,
mime_types=mime_types,
userprincipalname=userprincipalname,
isRelativePath=True,
)
downloaded_files_metadata.update(folder_metadata)
# Download files using the provided file paths
if file_paths:
is_download_from_root = False
for file_path in file_paths or []:
item = self._get_items_in_drive_with_maxretries(
access_token,
file_path,
userprincipalname=userprincipalname,
isFile=True,
isRelativePath=True,
)
file_metadata = self._check_approved_mimetype_and_download_file(
item, temp_dir, mime_types
)
downloaded_files_metadata.update(file_metadata)
if is_download_from_root:
# download files from root folder
root_folder_metadata = self._connect_download_and_return_metadata(
access_token,
temp_dir,
"root",
recursive,
mime_types=mime_types,
userprincipalname=userprincipalname,
)
downloaded_files_metadata.update(root_folder_metadata)
return downloaded_files_metadata
def _load_documents_with_metadata(
self, directory: str, recursive: bool = True
) -> List[Document]:
"""
Load documents from a specified directory using the SimpleDirectoryReader
and associate them with their respective metadata.
Parameters:
- directory (str): The directory from which to load the documents.
- recursive (bool, optional): Whether to perform a recursive search through the directory. Defaults to True.
Returns:
- List[Document]: Loaded documents from the specified directory with associated metadata.
"""
def get_metadata(filename: str) -> Any:
return self._downloaded_files_metadata[filename]
simple_loader = SimpleDirectoryReader(
directory,
file_extractor=self.file_extractor,
file_metadata=get_metadata,
recursive=recursive,
)
return simple_loader.load_data()
def load_data(
self,
folder_id: Optional[str] = None,
file_ids: Optional[List[str]] = None,
folder_path: Optional[str] = None,
file_paths: Optional[List[str]] = None,
mime_types: Optional[List[str]] = None,
recursive: bool = True,
userprincipalname: Optional[str] = None,
) -> List[Document]:
"""Load data from the folder id / file ids, f both are not provided download from the root.
Args:
folder_id (str, optional): folder id of the folder in OneDrive.
file_ids (List[str], optional): file ids of the files in OneDrive.
folder_path (str, optional): The relative path of the OneDrive folder to download. If provided, files within the folder are downloaded.
file_paths (List[str], optional): List of specific file paths to download.
mime_types: the mimeTypes you want to allow e.g.: "application/pdf", default is none, which loads all files found
recursive: boolean value to traverse and read subfolder, default is True
userprincipalname: str value indicating the userprincipalname (normally organization-provided email) whose OneDrive will be accessed. Required for App authentication scenarios.
Returns:
List[Document]: A list of documents.
"""
# If arguments are not provided to load_data(), initialize them from the object's attributes
if not userprincipalname:
userprincipalname = self.userprincipalname
if not folder_id:
folder_id = self.folder_id
if not file_ids:
file_ids = self.file_ids
if not folder_path:
folder_path = self.folder_path
if not file_paths:
file_paths = self.file_paths
try:
with tempfile.TemporaryDirectory() as temp_dir:
self._downloaded_files_metadata = self._init_download_and_get_metadata(
temp_dir=temp_dir,
folder_id=folder_id,
file_ids=file_ids,
folder_path=folder_path,
file_paths=file_paths,
recursive=recursive,
mime_types=mime_types,
userprincipalname=userprincipalname,
)
return self._load_documents_with_metadata(temp_dir, recursive=recursive)
except Exception as e:
logger.error(
f"An error occurred while loading the data: {e}", exc_info=True
)
|