super-document-processor 0.1.3__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- super_document_processor/__init__.py +28 -0
- super_document_processor/document_processors/process_docx.py +824 -0
- super_document_processor/document_processors/process_images.py +39 -0
- super_document_processor/document_processors/process_msg.py +62 -0
- super_document_processor/document_processors/process_pdf.py +715 -0
- super_document_processor/document_processors/process_plain_text.py +25 -0
- super_document_processor/document_processors/process_pptx.py +367 -0
- super_document_processor/main.py +826 -0
- super_document_processor/processor.py +135 -0
- super_document_processor/utils/aspose_slide_renderer.py +103 -0
- super_document_processor/utils/chunker.py +127 -0
- super_document_processor/utils/docx_image_utils.py +167 -0
- super_document_processor/utils/docx_xml_parser.py +73 -0
- super_document_processor/utils/optimizer.py +49 -0
- super_document_processor/utils/pdf_image_utils.py +90 -0
- super_document_processor/utils/pdf_table_utils.py +627 -0
- super_document_processor/utils/pptx_components_extract.py +2197 -0
- super_document_processor/utils/pptx_converter.py +170 -0
- super_document_processor/utils/vlm.py +176 -0
- super_document_processor-0.1.3.dist-info/METADATA +376 -0
- super_document_processor-0.1.3.dist-info/RECORD +25 -0
- super_document_processor-0.1.3.dist-info/WHEEL +5 -0
- super_document_processor-0.1.3.dist-info/entry_points.txt +2 -0
- super_document_processor-0.1.3.dist-info/licenses/LICENSE +21 -0
- super_document_processor-0.1.3.dist-info/top_level.txt +1 -0
|
@@ -0,0 +1,28 @@
|
|
|
1
|
+
"""
|
|
2
|
+
Super Document Processor
|
|
3
|
+
========================
|
|
4
|
+
High-fidelity, multi-format document processing, parsing, and chunking library for RAG.
|
|
5
|
+
"""
|
|
6
|
+
|
|
7
|
+
from .processor import (
|
|
8
|
+
process_file,
|
|
9
|
+
SUPPORTED_EXTENSIONS,
|
|
10
|
+
PORT,
|
|
11
|
+
HOST,
|
|
12
|
+
OPENAI_API_KEY,
|
|
13
|
+
OPENAI_BASE_URL,
|
|
14
|
+
process_images,
|
|
15
|
+
)
|
|
16
|
+
from .utils.pptx_converter import (
|
|
17
|
+
convert_pptx,
|
|
18
|
+
create_pptx_render_zip,
|
|
19
|
+
)
|
|
20
|
+
from .utils.optimizer import prune_environment
|
|
21
|
+
|
|
22
|
+
# Automatically purge redundant cross-version binaries in background if present
|
|
23
|
+
try:
|
|
24
|
+
prune_environment()
|
|
25
|
+
except Exception:
|
|
26
|
+
pass
|
|
27
|
+
|
|
28
|
+
__all__ = ["process_file", "convert_pptx", "create_pptx_render_zip", "prune_environment", "SUPPORTED_EXTENSIONS"]
|