yujunjun/unstructured
1
0
Fork 0
mirror of https://github.com/Unstructured-IO/unstructured.git synced 2025-10-11 16:13:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Adds data source properties to elasticsearch, wikipedia and google-drive (#1282)

Browse Source
This commit is contained in:
rvztz 2023-09-19 14:25:38 -06:00 committed by GitHub
parent 92e18c3f58
commit 9a3e24fcbb
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
18 changed files with 468 additions and 70 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
CHANGELOG.md
View File

@ -1,9 +1,9 @@
## 0.10.16-dev1
## 0.10.16-dev2
### Enhancements
* **Adds data source properties to Airtable, Confluence and Discord connectors** These properties (date_created, date_modified, version, source_url, record_locator) are written to element metadata during ingest, mapping elements to information about the document source from which they derive. This functionality enables downstream applications to reveal source document applications, e.g. a link to a GDrive doc, Salesforce record, etc.
* **Adds data source properties to Airtable, Confluence, Discord, Elasticsearch, Google Drive, and Wikipedia connectors** These properties (date_created, date_modified, version, source_url, record_locator) are written to element metadata during ingest, mapping elements to information about the document source from which they derive. This functionality enables downstream applications to reveal source document applications, e.g. a link to a GDrive doc, Salesforce record, etc.
### Features

45
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/0-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "0"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "0d15a8bf4961deb609a392a8444e3520",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "0"
}
},
"filetype": "text/plain"
},
"text": "Cecil Hepworth"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "a1515877c1c63770057b2615cce25c5d",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "0"
}
},
"filetype": "text/plain"
},
"text": "Alice follows a large white rabbit down a \"Rabbit-hole\". She finds a tiny door. When she finds a bottle labeled \"Drink me\", she does, and shrinks, but not enough to pass through the door. She then eats something labeled \"Eat me\" and grows larger. She finds a fan when enables her to shrink enough to get into the \"Garden\" and try to get a \"Dog\" to play with her. She enters the \"White Rabbit's tiny House,\" but suddenly resumes her normal size. In order to get out, she has to use the \"magic fan.\""
@ -30,7 +51,14 @@
"type": "NarrativeText",
"element_id": "58ec505c394f8af4fc5c62bad6973652",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "0"
}
},
"filetype": "text/plain"
},
"text": "She enters a kitchen, in which there is a cook and a woman holding a baby. She persuades the woman to give her the child and takes the infant outside after the cook starts throwing things around. The baby then turns into a pig and squirms out of her grip. \"The Duchess's Cheshire Cat\" appears and disappears a couple of times to Alice and directs her to the Mad Hatter's \"Mad Tea-Party.\" After a while, she leaves."
@ -39,7 +67,14 @@
"type": "NarrativeText",
"element_id": "fffac28d27f8cea00e96f1e876a1d1f8",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "0"
}
},
"filetype": "text/plain"
},
"text": "The Queen invites Alice to join the \"ROYAL PROCESSION\": a parade of marching playing cards and others headed by the White Rabbit. When Alice \"unintentionally offends the Queen\", the latter summons the \"Executioner\". Alice \"boxes the ears\", then flees when all the playing cards come for her. Then she wakes up and realizes it was all a dream."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/1-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "1"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "576608bb13aa67420e79d575e0e26071",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "1"
}
},
"filetype": "text/plain"
},
"text": "Wallace McCutcheon and Ediwin S. Porter"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "d30707943c5b8e45088e21b0a9ba6f1a",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "1"
}
},
"filetype": "text/plain"
},
"text": "Boone's daughter befriends an Indian maiden as Boone and his companion start out on a hunting expedition. While he is away, Boone's cabin is attacked by the Indians, who set it on fire and abduct Boone's daughter. Boone returns, swears vengeance, then heads out on the trail to the Indian camp. His daughter escapes but is chased. The Indians encounter Boone, which sets off a huge fight on the edge of a cliff. A burning arrow gets shot into the Indian camp. Boone gets tied to the stake and tortured. The burning arrow sets the Indian camp on fire, causing panic. Boone is rescued by his horse, and Boone has a knife fight in which he kills the Indian chief. [2]"

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/2-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "2"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "b764cdc0eab7137467211272fa539f12",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "2"
}
},
"filetype": "text/plain"
},
"text": "Unknown"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "2659129a67b301911027d0ea747109e4",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "2"
}
},
"filetype": "text/plain"
},
"text": "Before heading out to a baseball game at a nearby ballpark, sports fan Mr. Brown drinks several highball cocktails. He arrives at the ballpark to watch the game, but has become so inebriated that the game appears to him in reverse, with the players running the bases backwards and the baseball flying back into the pitcher's hand. After the game is over, Mr. Brown is escorted home by one of his friends. When they arrive at Brown's house, they encounter his wife who becomes furious with the friend and proceeds to physically assault him, believing he is responsible for her husband's severe intoxication. [1]"

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/3-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "3"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "9496aba3ea633310e2d669820269ad00",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "3"
}
},
"filetype": "text/plain"
},
"text": "Edwin Stanton Porter"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "6ead4aca7b509813147c41699dd1a7d4",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "3"
}
},
"filetype": "text/plain"
},
"text": "The plot is that of a black woman going to the dentist for a toothache and being given laughing gas. On her way walking home, and in other situations, she can't stop laughing, and everyone she meets \"catches\" the laughter from her, including a vendor and police officers."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/4-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "4"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "9abdc842ab4bacbaa4da45cec2ef7e0d",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "4"
}
},
"filetype": "text/plain"
},
"text": "D. W. Griffith"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "b8004022d0994669c5fdb4ec8a5088a9",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "4"
}
},
"filetype": "text/plain"
},
"text": "On a beautiful summer day a father and mother take their daughter Dollie on an outing to the river. The mother refuses to buy a gypsy's wares. The gypsy tries to rob the mother, but the father drives him off. The gypsy returns to the camp and devises a plan. They return and kidnap Dollie while her parents are distracted. A rescue crew is organized, but the gypsy takes Dollie to his camp. They gag Dollie and hide her in a barrel before the rescue party gets to the camp. Once they leave the gypsies and escapes in their wagon. As the wagon crosses the river, the barrel falls into the water. Still sealed in the barrel, Dollie is swept downstream in dangerous currents. A boy who is fishing in the river finds the barrel, and Dollie is reunited safely with her parents."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/5-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "5"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "9abdc842ab4bacbaa4da45cec2ef7e0d",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "5"
}
},
"filetype": "text/plain"
},
"text": "D. W. Griffith"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "8e16a508f3df737af12e84d9cba2c7d0",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "5"
}
},
"filetype": "text/plain"
},
"text": "A thug accosts a girl as she leaves her workplace but a man rescues her. The thug vows revenge and, with the help of two friends, attacks the girl and her rescuer again as they're going for a walk. This time they succeed in kidnapping the rescuer. He is bound and gagged and taken away in a cart. The girl runs home and gets help from several neighbors. They track the ruffians down to a cabin in the mountains where the gang has trapped their victim and set the cabin on fire. A thug and Rescuer fight on the roof of the house."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/6-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "6"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "2ac7798b427181278fb2b450e28f4902",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "6"
}
},
"filetype": "text/plain"
},
"text": "D.W. Griffith"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "9e92dee6e0d6ef246f51d7f8f4eb8c01",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "6"
}
},
"filetype": "text/plain"
},
"text": "A young couple decides to elope after being caught in the midst of a romantic moment by the woman's angry father. They make plans to leave, but a thief discovers their plans and hides in their trunk and waits for the right moment to steal their belongings."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/7-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "7"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "9abdc842ab4bacbaa4da45cec2ef7e0d",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "7"
}
},
"filetype": "text/plain"
},
"text": "D. W. Griffith"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "d366dfc3239f22e3c03ee629f6567a68",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "7"
}
},
"filetype": "text/plain"
},
"text": "A white girl (Florence Lawrence) rejects a proposal from an Indian brave (Charles Inslee) in this early one-reel Western melodrama. Despite the rejection, the Indian still comes to the girl's defense when she is abducted by his warring tribe. In her first year in films, Florence Lawrence was already the most popular among the Biograph Company's anonymous stock company players. By 1909, she was known the world over as \"The Biograph Girl.\""

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/8-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "8"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "b764cdc0eab7137467211272fa539f12",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "8"
}
},
"filetype": "text/plain"
},
"text": "Unknown"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "7ddfc82896f749f2c5b5c5baac5a93bf",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "8"
}
},
"filetype": "text/plain"
},
"text": "No prints of the first American film adaptation of A Christmas Carol are known to exist,[1] but The Moving Picture World magazine provided a scene-by-scene description before the film's release. [2] Scrooge goes into his office and begins working. His nephew, along with three women who wish for Scrooge to donate enter. However, Scrooge dismisses them. On the night of Christmas Eve, his long-dead partner Jacob Marley comes as a ghost, warning him of a horrible fate if he does not change his ways. Scrooge meets three spirits that show Scrooge the real meaning of Christmas, along with his grave, the result of his parsimonious ways. The next morning, he wakes and realizes the error of his ways. Scrooge was then euphoric and generous for the rest of his life."

27
test_unstructured_ingest/expected-structured-output/elasticsearch/movies/9-31261b6f.json
View File

@ -3,7 +3,14 @@
"type": "Title",
"element_id": "0baaf6f25d1c5da82d9e4a3229cd45bd",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "9"
}
},
"filetype": "text/plain"
},
"text": "American"
@ -12,7 +19,14 @@
"type": "Title",
"element_id": "9abdc842ab4bacbaa4da45cec2ef7e0d",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "9"
}
},
"filetype": "text/plain"
},
"text": "D. W. Griffith"
@ -21,7 +35,14 @@
"type": "NarrativeText",
"element_id": "b87d0bbbe5c735bca621fc172fc44605",
"metadata": {
"data_source": {},
"data_source": {
"version": 1,
"record_locator": {
"url": "http://localhost:9200",
"index_name": "movies",
"document_id": "9"
}
},
"filetype": "text/plain"
},
"text": "The film opens in a town on the Mexican border. A poker game is going on in the local saloon. One of the players cheats and is shot dead by another of the players, a Mexican named Pedro. In the uproar that follows Pedro is wounded as he escapes from the saloon. The sheriff is called, who tracks Pedro to his home but Pedro kills the sherriff too. While Pedro hides, his wife Juanita, is arrested on suspicion of murdering the sheriff. Pedro rescues her from the town jail and the two head for the Mexican border. Caught by the posse before they reach the border, Juanita is killed and the film ends with Pedro being arrested and taken back to town."

22
test_unstructured_ingest/expected-structured-output/google-drive/117qrVqiCoR5EjYMsDHGdy3UMkEtKr9Q8-test-drive-doc.docx.json
View File

@ -3,7 +3,16 @@
"type": "Title",
"element_id": "7e8cd2056da73a7fefb6cd91f4e5d199",
"metadata": {
"data_source": {},
"data_source": {
"url": "https://drive.google.com/uc?id=117qrVqiCoR5EjYMsDHGdy3UMkEtKr9Q8&export=download",
"version": "15",
"record_locator": {
"drive_id": "1OQZ66OHBE30rNsNa7dweGLfRmXvkT_jr",
"file_id": "117qrVqiCoR5EjYMsDHGdy3UMkEtKr9Q8"
},
"date_created": "2023-06-15T06:15:58.931000",
"date_modified": "2023-06-15T06:15:44"
},
"filetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
"emphasized_text_contents": [
"Title"
@ -18,7 +27,16 @@
"type": "NarrativeText",
"element_id": "9870998df89c1da4e01378d0fd085106",
"metadata": {
"data_source": {},
"data_source": {
"url": "https://drive.google.com/uc?id=117qrVqiCoR5EjYMsDHGdy3UMkEtKr9Q8&export=download",
"version": "15",
"record_locator": {
"drive_id": "1OQZ66OHBE30rNsNa7dweGLfRmXvkT_jr",
"file_id": "117qrVqiCoR5EjYMsDHGdy3UMkEtKr9Q8"
},
"date_created": "2023-06-15T06:15:58.931000",
"date_modified": "2023-06-15T06:15:44"
},
"filetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
"emphasized_text_contents": [
"This is a good reason to continue"

11
test_unstructured_ingest/expected-structured-output/google-drive/1SpQuE7jHz9nMt5hfQXsiok1SgIdRYX5o-fake.docx.json
View File

@ -3,7 +3,16 @@
"type": "Title",
"element_id": "dd14cbbf0e74909aac7f248a85d190af",
"metadata": {
"data_source": {},
"data_source": {
"url": "https://drive.google.com/uc?id=1SpQuE7jHz9nMt5hfQXsiok1SgIdRYX5o&export=download",
"version": "17",
"record_locator": {
"drive_id": "1OQZ66OHBE30rNsNa7dweGLfRmXvkT_jr",
"file_id": "1SpQuE7jHz9nMt5hfQXsiok1SgIdRYX5o"
},
"date_created": "2023-06-15T06:15:59.687000",
"date_modified": "2023-06-15T06:15:43"
},
"filetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
},
"text": "Lorem ipsum dolor sit amet."

11
test_unstructured_ingest/expected-structured-output/google-drive/1cTKXAreuj-wYmL38nFnqKvz3X8UKcaMC-foo.txt.json
View File

@ -3,7 +3,16 @@
"type": "Title",
"element_id": "8b5b9db0c13db24256c829aa364aa90c",
"metadata": {
"data_source": {},
"data_source": {
"url": "https://drive.google.com/uc?id=1cTKXAreuj-wYmL38nFnqKvz3X8UKcaMC&export=download",
"version": "9",
"record_locator": {
"drive_id": "1OQZ66OHBE30rNsNa7dweGLfRmXvkT_jr",
"file_id": "1cTKXAreuj-wYmL38nFnqKvz3X8UKcaMC"
},
"date_created": "2023-06-15T06:15:59.687000",
"date_modified": "2023-06-15T06:15:39"
},
"filetype": "text/plain"
},
"text": "three"

2
unstructured/__version__.py
View File

@ -1 +1 @@
__version__ = "0.10.16-dev1" # pragma: no cover
__version__ = "0.10.16-dev2" # pragma: no cover

80
unstructured/ingest/connector/elasticsearch.py
View File

@ -12,6 +12,7 @@ from unstructured.ingest.interfaces import (
BaseSourceConnector,
IngestDocCleanupMixin,
SourceConnectorCleanupMixin,
SourceMetadata,
)
from unstructured.ingest.logger import logger
from unstructured.utils import requires_dependencies
@ -33,7 +34,7 @@ class SimpleElasticsearchConfig(BaseConnectorConfig):
@dataclass
class ElasticsearchFileMeta:
class ElasticsearchDocumentMeta:
"""Metadata specifying:
name of the elasticsearch index that is being reached to,
and the id of document that is being reached to,
@ -53,7 +54,7 @@ class ElasticsearchIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
"""
connector_config: SimpleElasticsearchConfig
file_meta: ElasticsearchFileMeta
document_meta: ElasticsearchDocumentMeta
registry_name: str = "elasticsearch"
# TODO: remove one of filename or _tmp_download_file, using a wrapper
@ -61,8 +62,8 @@ class ElasticsearchIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
def filename(self):
return (
Path(self.read_config.download_dir)
/ self.file_meta.index_name
/ f"{self.file_meta.document_id}.txt"
/ self.document_meta.index_name
/ f"{self.document_meta.document_id}.txt"
).resolve()
@property
@ -71,7 +72,7 @@ class ElasticsearchIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
the output file."""
# Generate SHA256 hash and take the first 8 characters
query_hash = hashlib.sha256((self.connector_config.jq_query or "").encode()).hexdigest()[:8]
output_file = f"{self.file_meta.document_id}-{query_hash}.json"
output_file = f"{self.document_meta.document_id}-{query_hash}.json"
return (
Path(self.partition_config.output_dir) / self.connector_config.index_name / output_file
)
@ -104,21 +105,50 @@ class ElasticsearchIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
concatenated_values = seperator.join(values)
return concatenated_values
@requires_dependencies(["elasticsearch"], extras="elasticsearch")
def _get_document(self):
from elasticsearch import Elasticsearch, NotFoundError
try:
# TODO: instead of having a separate client for each doc,
# have a separate client for each process
es = Elasticsearch(self.connector_config.url)
document = es.get(
index=self.connector_config.index_name,
id=self.document_meta.document_id,
)
except NotFoundError:
logger.error("Couldn't find document with ID: %s", self.document_meta.document_id)
return None
return document
def update_source_metadata(self, **kwargs):
document = kwargs.get("document", self._get_document())
if document is None:
self.source_metadata = SourceMetadata(
exists=False,
)
return
self.source_metadata = SourceMetadata(
version=document["_version"],
exists=document["found"],
)
@SourceConnectionError.wrap
@requires_dependencies(["elasticsearch", "jq"], extras="elasticsearch")
@requires_dependencies(["jq"], extras="elasticsearch")
@BaseIngestDoc.skip_if_file_exists
def get_file(self):
import jq
from elasticsearch import Elasticsearch
logger.debug(f"Fetching {self} - PID: {os.getpid()}")
# TODO: instead of having a separate client for each doc,
# have a separate client for each process
es = Elasticsearch(self.connector_config.url)
document_dict = es.get(
index=self.connector_config.index_name,
id=self.file_meta.document_id,
).body["_source"]
document = self._get_document()
self.update_source_metadata(document=document)
if document is None:
raise ValueError(
f"Failed to get document {self.document_meta.document_id}",
)
document_dict = document.body["_source"]
if self.connector_config.jq_query:
document_dict = json.loads(
jq.compile(self.connector_config.jq_query).input(document_dict).text(),
@ -128,6 +158,26 @@ class ElasticsearchIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
with open(self.filename, "w", encoding="utf8") as f:
f.write(self.document)
@property
def date_created(self) -> t.Optional[str]:
return None
@property
def date_modified(self) -> t.Optional[str]:
return None
@property
def source_url(self) -> t.Optional[str]:
return None
@property
def record_locator(self) -> t.Optional[t.Dict[str, t.Any]]:
return {
"url": self.connector_config.url,
"index_name": self.connector_config.index_name,
"document_id": self.document_meta.document_id,
}
@dataclass
class ElasticsearchSourceConnector(SourceConnectorCleanupMixin, BaseSourceConnector):
@ -166,7 +216,7 @@ class ElasticsearchSourceConnector(SourceConnectorCleanupMixin, BaseSourceConnec
connector_config=self.connector_config,
partition_config=self.partition_config,
read_config=self.read_config,
file_meta=ElasticsearchFileMeta(self.connector_config.index_name, id),
document_meta=ElasticsearchDocumentMeta(self.connector_config.index_name, id),
)
for id in ids
]

85
unstructured/ingest/connector/google_drive.py
View File

@ -3,6 +3,7 @@ import json
import os
import typing as t
from dataclasses import dataclass
from datetime import datetime
from mimetypes import guess_extension
from pathlib import Path
@ -18,6 +19,7 @@ from unstructured.ingest.interfaces import (
IngestDocCleanupMixin,
IngestDocSessionHandleMixin,
SourceConnectorCleanupMixin,
SourceMetadata,
)
from unstructured.ingest.logger import logger
from unstructured.utils import requires_dependencies
@ -103,57 +105,109 @@ class SimpleGoogleDriveConfig(ConfigSessionHandleMixin, BaseConnectorConfig):
@dataclass
class GoogleDriveIngestDoc(IngestDocSessionHandleMixin, IngestDocCleanupMixin, BaseIngestDoc):
connector_config: SimpleGoogleDriveConfig
file_meta: t.Dict[str, str]
meta: t.Dict[str, str]
registry_name: str = "google_drive"
@property
def filename(self):
return Path(self.file_meta.get("download_filepath")).resolve() # type: ignore
return Path(self.meta.get("download_filepath")).resolve() # type: ignore
@property
def _output_filename(self):
return Path(f"{self.file_meta.get('output_filepath')}.json").resolve()
return Path(f"{self.meta.get('output_filepath')}.json").resolve()
@property
def record_locator(self) -> t.Optional[t.Dict[str, t.Any]]:
return {
"drive_id": self.connector_config.drive_id,
"file_id": self.meta["id"],
}
@requires_dependencies(["googleapiclient"], extras="google-drive")
def update_source_metadata(self):
from googleapiclient.errors import HttpError
try:
file_obj = (
self.session_handle.service.files()
.get(
fileId=self.meta["id"],
fields="id, createdTime, modifiedTime, version, webContentLink",
)
.execute()
)
except HttpError as e:
if e.status_code == 404:
logger.error(f"File {self.meta['name']} not found")
self.source_metadata = SourceMetadata(
exists=True,
)
return
raise
date_created = None
if dc := file_obj.get("createdTime", ""):
date_created = datetime.strptime(
dc,
"%Y-%m-%dT%H:%M:%S.%fZ",
).isoformat()
date_modified = None
if dm := file_obj.get("modifiedTime", ""):
date_modified = datetime.strptime(
dm,
"%Y-%m-%dT%H:%M:%S.%fZ",
).isoformat()
self.source_metadata = SourceMetadata(
date_created=date_created,
date_modified=date_modified,
version=file_obj.get("version", ""),
source_url=file_obj.get("webContentLink", ""),
exists=True,
)
@requires_dependencies(["googleapiclient"], extras="google-drive")
@SourceConnectionError.wrap
@BaseIngestDoc.skip_if_file_exists
@requires_dependencies(["googleapiclient"], extras="google-drive")
def get_file(self):
from googleapiclient.errors import HttpError
from googleapiclient.http import MediaIoBaseDownload
if self.file_meta.get("mimeType", "").startswith("application/vnd.google-apps"):
if self.meta.get("mimeType", "").startswith("application/vnd.google-apps"):
export_mime = GOOGLE_DRIVE_EXPORT_TYPES.get(
self.file_meta.get("mimeType"), # type: ignore
self.meta.get("mimeType"), # type: ignore
)
if not export_mime:
logger.info(
f"File not supported. Name: {self.file_meta.get('name')} "
f"ID: {self.file_meta.get('id')} "
f"MimeType: {self.file_meta.get('mimeType')}",
f"File not supported. Name: {self.meta.get('name')} "
f"ID: {self.meta.get('id')} "
f"MimeType: {self.meta.get('mimeType')}",
)
return
request = self.session_handle.service.files().export_media(
fileId=self.file_meta.get("id"),
fileId=self.meta.get("id"),
mimeType=export_mime,
)
else:
request = self.session_handle.service.files().get_media(fileId=self.file_meta.get("id"))
request = self.session_handle.service.files().get_media(fileId=self.meta.get("id"))
file = io.BytesIO()
downloader = MediaIoBaseDownload(file, request)
self.update_source_metadata()
downloaded = False
try:
while downloaded is False:
status, downloaded = downloader.next_chunk()
_, downloaded = downloader.next_chunk()
except HttpError:
pass
saved = False
if downloaded and file:
dir_ = Path(self.file_meta["download_dir"])
dir_ = Path(self.meta["download_dir"])
if dir_:
if not dir_.is_dir():
logger.debug(f"Creating directory: {self.file_meta.get('download_dir')}")
logger.debug(f"Creating directory: {self.meta.get('download_dir')}")
if dir_:
dir_.mkdir(parents=True, exist_ok=True)
@ -162,7 +216,6 @@ class GoogleDriveIngestDoc(IngestDocSessionHandleMixin, IngestDocCleanupMixin, B
handler.write(file.getbuffer())
saved = True
logger.debug(f"File downloaded: {self.filename}.")
if not saved:
logger.error(f"Error while downloading and saving file: {self.filename}.")
@ -267,7 +320,7 @@ class GoogleDriveSourceConnector(SourceConnectorCleanupMixin, BaseSourceConnecto
connector_config=self.connector_config,
partition_config=self.partition_config,
read_config=self.read_config,
file_meta=file,
meta=file,
)
for file in files
]

35
unstructured/ingest/connector/wikipedia.py
View File

@ -10,6 +10,7 @@ from unstructured.ingest.interfaces import (
BaseSourceConnector,
IngestDocCleanupMixin,
SourceConnectorCleanupMixin,
SourceMetadata,
)
from unstructured.ingest.logger import logger
from unstructured.utils import requires_dependencies
@ -50,15 +51,49 @@ class WikipediaIngestDoc(IngestDocCleanupMixin, BaseIngestDoc):
def _output_filename(self):
raise NotImplementedError()
@property
def date_created(self) -> t.Optional[str]:
return None
@property
def date_modified(self) -> t.Optional[str]:
return None
@property
def record_locator(self) -> t.Optional[t.Dict[str, t.Any]]:
return {
"page_title": self.connector_config.title,
"page_url": self.source_metadata.source_url, # type: ignore
}
def _create_full_tmp_dir_path(self):
self.filename.parent.mkdir(parents=True, exist_ok=True)
@requires_dependencies(["wikipedia"], extras="wikipedia")
def update_source_metadata(self):
from wikipedia.exceptions import PageError
try:
page = self.page
except PageError:
self.source_metadata = SourceMetadata(
exists=False,
)
return
self.source_metadata = SourceMetadata(
version=page.revision_id,
source_url=page.url,
exists=True,
)
@SourceConnectionError.wrap
@BaseIngestDoc.skip_if_file_exists
def get_file(self):
"""Fetches the "remote" doc and stores it locally on the filesystem."""
self._create_full_tmp_dir_path()
logger.debug(f"Fetching {self} - PID: {os.getpid()}")
self.update_source_metadata()
with open(self.filename, "w", encoding="utf8") as f:
f.write(self.text)