fix: update docling dependency to use docling_core instead (#8714)

* chore: update dependencies in pyproject.toml and uv.lock

* Updated `docling` to `docling_core>=2.38.0` in `pyproject.toml`.
* Removed outdated `docling` package entries from `uv.lock`.
* Adjusted dependency markers for `gassist` and `triton` to improve compatibility across platforms.

* chore: downgrade docling_core dependency in pyproject.toml

* Changed `docling_core` version from `>=2.38.0` to `>=2.36.1` for compatibility.
* Added optional dependency group for `docling` with the updated version.

* Added traceloop.py and initialized in service.py, also updated pyproject.toml

* updated returned callbacks list length from 5 to 6. Formatted service.py

* [autofix.ci] apply automated fixes

* OpenInference span kind replaced with Opentelemetry Span kind

* Added traceloop.py and initialized in service.py, also updated pyproject.toml

* updated returned callbacks list length from 5 to 6. Formatted service.py

* [autofix.ci] apply automated fixes

* OpenInference span kind replaced with Opentelemetry Span kind

* Added traceloop.py and initialized in service.py, also updated pyproject.toml

* updated returned callbacks list length from 5 to 6. Formatted service.py

* [autofix.ci] apply automated fixes

* OpenInference span kind replaced with Opentelemetry Span kind

* Updated as per coderabbitai suggested and tested

* Fixed deprecated NoneType

* code formatted

* revert uv.lock

* add new line to uv.lock

* updated service.py

* [autofix.ci] apply automated fixes

* fix(DoclingInlineComponent): add import error handling for missing Docling module

* Revert "Merge branch '2getsandesh/main' into remove-torch-deps"

This reverts commit b95891f0e27e8620873fbe97116709ca5de41e51, reversing
changes made to 9e0f1ae30a6655a82948546035147e0e2fc924ae.

---------

Co-authored-by: Sandesh R <2getsandesh@gmail.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Sandesh R <Sandesh.R24@ibm.com>
Co-authored-by: Edwin Jose <edwin.jose@datastax.com>
This commit is contained in:
Gabriel Luiz Freitas Almeida 2025-06-25 20:55:50 -03:00 • committed by GitHub
commit e28104c852
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 40 additions and 26 deletions

View file

@ -1,13 +1,3 @@
from docling.datamodel.base_models import ConversionStatus, InputFormat
from docling.datamodel.pipeline_options import (
OcrOptions,
PdfPipelineOptions,
VlmPipelineOptions,
)
from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption
from docling.models.factories import get_ocr_factory
from docling.pipeline.vlm_pipeline import VlmPipeline
from langflow.base.data import BaseFileComponent
from langflow.inputs import DropdownInput
from langflow.schema import Data
@ -80,6 +70,23 @@ class DoclingInlineComponent(BaseFileComponent):
]
def process_files(self, file_list: list[BaseFileComponent.BaseFile]) -> list[BaseFileComponent.BaseFile]:
try:
from docling.datamodel.base_models import ConversionStatus, InputFormat
from docling.datamodel.pipeline_options import (
OcrOptions,
PdfPipelineOptions,
VlmPipelineOptions,
)
from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption
from docling.models.factories import get_ocr_factory
from docling.pipeline.vlm_pipeline import VlmPipeline
except ImportError as e:
msg = (
"Docling is not installed. Please install it with `uv pip install docling` or"
" `uv pip install langflow[docling]`."
)
raise ImportError(msg) from e
# Configure the standard PDF pipeline
def _get_standard_opts() -> PdfPipelineOptions:
pipeline_options = PdfPipelineOptions()