Unstructured-IO / Unstructured-IO/unstructured-python-client

When splitting_pdf_page is started, only the last set of API requests can succeed.

Ouverte
#220 1 commentaire 1 réaction 0 personnes assignées Voir sur GitHub

Personne n'a encore pris cette issue.

Langage dominant
Python
Étoiles
119
Forks
22
Merge moyen
1 j 21 h
PR mergées (30 j)
1

Description

Describe the bug
When I set split_pdf_page=True,split_pdf_concurrency_level=15.
Assuming the pdf is divided into 10 sets, it will report an error:
ERROR: Failed to send request for page 1
...
WARNING: Failed to partition set Unstructured-IO/unstructured-api#1, its elements will be omitted in the final result.
...
WARNING: Failed to partition set Unstructured-IO/unstructured-api#9, its elements will be omitted in the final result.
INFO: Successfully partitioned set Unstructured-IO/unstructured-api#10, elements added to the final result.

To Reproduce
code:

import os, json

import requests
from unstructured_client.models.operations import PartitionRequest
from unstructured_client.models.shared import PartitionParameters, ChunkingStrategy

os.environ["UNSTRUCTURED_API_KEY"] = "EMPTY"
os.environ["UNSTRUCTURED_API_URL"] = ""

import unstructured_client
from unstructured_client.models import shared, operations

requests_client = requests.Session()
client = unstructured_client.UnstructuredClient(
    api_key_auth=os.getenv("UNSTRUCTURED_API_KEY"),
    server_url=os.getenv("UNSTRUCTURED_API_URL"),
    client=requests_client
)

filename = "./test_pdf.pdf"

file = open(filename, "rb")
req = operations.PartitionRequest(
    partition_parameters=shared.PartitionParameters(
        files=shared.Files(
            content=file.read(),
            file_name=filename,
        ),
        strategy=shared.Strategy.HI_RES,
        split_pdf_page=True,
        split_pdf_concurrency_level=15,
        chunking_strategy=ChunkingStrategy("by_title")
    )
)

try:
    res = client.general.partition(req)
    element_dicts = [element for element in res.elements]

    print(element_dicts)
    for e in element_dicts:
        print(e['text'])
except Exception as e:
    print(e)

Console Information:

INFO: Preparing to split document for partition.
INFO: Concurrency level set to 15
INFO: Splitting pages 1 to 23 (23 total)
INFO: Determined optimal split size of 2 pages.
INFO: Partitioning 11 files with 2 page(s) each.
INFO: Partitioning 1 file with 1 page(s).
INFO: Partitioning set Unstructured-IO/unstructured-api#1 (pages 1-2).
INFO: Partitioning set Unstructured-IO/unstructured-api#2 (pages 3-4).
INFO: Partitioning set Unstructured-IO/unstructured-api#3 (pages 5-6).
INFO: Partitioning set Unstructured-IO/unstructured-api#4 (pages 7-8).
INFO: Partitioning set Unstructured-IO/unstructured-api#5 (pages 9-10).
INFO: Partitioning set Unstructured-IO/unstructured-api#6 (pages 11-12).
INFO: Partitioning set Unstructured-IO/unstructured-api#7 (pages 13-14).
INFO: Partitioning set Unstructured-IO/unstructured-api#8 (pages 15-16).
INFO: Partitioning set Unstructured-IO/unstructured-api#9 (pages 17-18).
INFO: Partitioning set Unstructured-IO/unstructured-api#10 (pages 19-20).
INFO: Partitioning set Unstructured-IO/unstructured-api#11 (pages 21-22).
INFO: Partitioning set Unstructured-IO/unstructured-api#12 (pages 23-23).
ERROR: Failed to send request for page 1
ERROR: Failed to send request for page 3
ERROR: Failed to send request for page 5
ERROR: Failed to send request for page 7
ERROR: Failed to send request for page 9
ERROR: Failed to send request for page 11
ERROR: Failed to send request for page 13
ERROR: Failed to send request for page 15
ERROR: Failed to send request for page 17
ERROR: Failed to send request for page 19
ERROR: Failed to send request for page 21
WARNING: Failed to partition set Unstructured-IO/unstructured-api#1, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#2, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#3, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#4, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#5, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#6, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#7, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#8, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#9, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#10, its elements will be omitted in the final result.
WARNING: Failed to partition set Unstructured-IO/unstructured-api#11, its elements will be omitted in the final result.
INFO: Successfully partitioned set Unstructured-IO/unstructured-api#12, elements added to the final result.
INFO: Successfully partitioned the document.

Guide de contribution

Ouvrir le guide de contribution

Par où commencer

  1. Lisez l'issue en entier, puis le guide de contribution du projet.
  2. Signalez en commentaire que vous la prenez — cela évite que deux personnes fassent le même travail.
  3. Forkez le dépôt et travaillez sur une branche.
  4. Ouvrez une pull request qui référence le numéro de l'issue.

Piste de recherche

Commencez par client.general.partition et suivez le traitement de split_pdf_page, split_pdf_concurrency_level et des requêtes par page présentées dans la reproduction. Exécutez l’exemple fourni de 23 pages et examinez pourquoi les ensembles de requêtes précédents échouent alors que le dernier réussit. C’est terminé lorsque tous les ensembles partitionnés s’achèvent avec succès et que leurs éléments sont inclus dans le résultat final.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Évaluation

Stack technique
python
Domaine
api
Type d'issue
Bug
Difficulté
3/5
Temps estimé
1-2 jours
Activité
À l'abandon
Clarté
Plutôt claire
Accessibilité débutants
38/100

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.