langchain-xparse
LangChain integration with xParse Parse API for intelligent document parsing. Converts unstructured documents (PDF, images, Word, Excel, PPT, etc.) into AI-friendly structured data (JSON, Markdown) with rich metadata.
Installation
From PyPI:
pip install langchain-xparse
Configuration
Set your TextIn credentials (from Textin Workspace):
export XPARSE_APP_ID="your-app-id"
export XPARSE_SECRET_CODE="your-secret-code"
Or pass them when creating the loader:
loader = XParseLoader(
file_path="doc.pdf",
app_id="your-app-id",
secret_code="your-secret-code",
)
Usage
Basic Usage
from langchain_xparse import XParseLoader
loader = XParseLoader(file_path="example.pdf")
docs = loader.load()
print(docs[0].page_content[:200])
print(docs[0].metadata) # source, category, element_id, filename, page_number
Lazy Load
for doc in loader.lazy_load():
# process each document
print(doc.page_content[:100])
Async Load
async for doc in loader.alazy_load():
# process each document asynchronously
print(doc.page_content[:100])
Custom Parse Configuration
Customize parsing behavior using the config parameter. See Parse Config Documentation for details.
loader = XParseLoader(
file_path="doc.pdf",
config={
"document": {
"password": "pdf-password" # For encrypted PDFs
},
"capabilities": {
"include_hierarchy": True, # Include parent-child relationships
"include_inline_objects": True, # Extract formulas, handwriting, etc.
"include_table_structure": True, # Detailed table structure
"include_char_details": True, # Character-level details
"include_image_data": True, # Image URLs and data
"pages": True, # Page metadata
"title_tree": True, # Document outline/TOC
"table_view": "html" # Table format: "html" or "markdown"
},
"scope": {
"page_range": "1-10" # Process specific pages
},
"config": {
"force_engine": "textin", # Engine selection (expert mode)
"engine_params": {
"formula_level": 0,
"image_output_type": "url"
}
}
}
)
docs = loader.load()
Multiple Files
loader = XParseLoader(file_path=["a.pdf", "b.pdf", "c.docx"])
for doc in loader.lazy_load():
print(f"{doc.metadata.get('source')}: {doc.page_content[:50]}")
File-like Object
When passing a file-like object instead of a path, you must set metadata_filename:
with open("doc.pdf", "rb") as f:
loader = XParseLoader(file=f, metadata_filename="doc.pdf")
docs = loader.load()
Document Metadata
Each loaded document includes rich metadata:
source: File path or filenamecategory: Element type (Title, NarrativeText, Table, Image, Formula, etc.)element_id: Unique element identifierfilename: Original filenamepage_number: Page number (if available)parent_id: Parent element ID (withinclude_hierarchy)children_ids: Child element IDs (withinclude_hierarchy)- Additional element-specific metadata
References
- xParse Parse API - API endpoint documentation
- Parse Config - Configuration parameters
- Parse Response - Response structure and fields
Metadata
Release files for langchain-xparse 1.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| langchain_xparse-1.2.0.tar.gz | 6.9 kB | Without provenance | Details for langchain_xparse-1.2.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| langchain_xparse-1.2.0-py3-none-any.whl | Python 3 | none | any | Details for langchain_xparse-1.2.0-py3-none-any.whl |
Total release size: 14.4 kB
Release files / langchain_xparse-1.2.0.tar.gz
| Download URL | langchain_xparse-1.2.0.tar.gz |
|---|---|
| Size | 6.9 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
4dd99c93f1dcb004b64e01a9eac40632cd5149c98ce21c0e3ab05972b950fddc
|
|
BLAKE2b-256 checksum How to use checksums |
c85fdf3bd5bc3651e14f00048c37dd8d9189e2a0bd89fb1f7d6ed8be5d09baad
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.19
|
Release files / langchain_xparse-1.2.0-py3-none-any.whl
| Download URL | langchain_xparse-1.2.0-py3-none-any.whl |
|---|---|
| Size | 7.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
83031d7e7fb4eeb49fa7b4e17e1c0e2418f1ce7abb3e4560148dc0be98f4ad83
|
|
BLAKE2b-256 checksum How to use checksums |
2d117010217e589958fa78987995c5690d62d67d4cdc76d33a1d87bbd432ee76
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.19
|
Security for langchain-xparse 1.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.