natural-pdf 0.1.11__py3-none-any.whl → 0.1.13__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.
- natural_pdf/__init__.py +7 -2
- natural_pdf/analyzers/shape_detection_mixin.py +1092 -0
- natural_pdf/analyzers/text_options.py +9 -1
- natural_pdf/analyzers/text_structure.py +371 -58
- natural_pdf/classification/manager.py +3 -4
- natural_pdf/collections/pdf_collection.py +19 -39
- natural_pdf/core/element_manager.py +11 -1
- natural_pdf/core/highlighting_service.py +146 -75
- natural_pdf/core/page.py +287 -188
- natural_pdf/core/pdf.py +57 -42
- natural_pdf/elements/base.py +51 -0
- natural_pdf/elements/collections.py +362 -67
- natural_pdf/elements/line.py +5 -0
- natural_pdf/elements/region.py +396 -23
- natural_pdf/exporters/data/__init__.py +0 -0
- natural_pdf/exporters/data/pdf.ttf +0 -0
- natural_pdf/exporters/data/sRGB.icc +0 -0
- natural_pdf/exporters/hocr.py +40 -61
- natural_pdf/exporters/hocr_font.py +7 -13
- natural_pdf/exporters/original_pdf.py +10 -13
- natural_pdf/exporters/paddleocr.py +51 -11
- natural_pdf/exporters/searchable_pdf.py +0 -10
- natural_pdf/flows/__init__.py +12 -0
- natural_pdf/flows/collections.py +533 -0
- natural_pdf/flows/element.py +382 -0
- natural_pdf/flows/flow.py +216 -0
- natural_pdf/flows/region.py +458 -0
- natural_pdf/search/__init__.py +65 -52
- natural_pdf/search/lancedb_search_service.py +325 -0
- natural_pdf/search/numpy_search_service.py +255 -0
- natural_pdf/search/searchable_mixin.py +25 -71
- natural_pdf/selectors/parser.py +163 -8
- natural_pdf/widgets/viewer.py +22 -31
- {natural_pdf-0.1.11.dist-info → natural_pdf-0.1.13.dist-info}/METADATA +55 -49
- {natural_pdf-0.1.11.dist-info → natural_pdf-0.1.13.dist-info}/RECORD +38 -30
- {natural_pdf-0.1.11.dist-info → natural_pdf-0.1.13.dist-info}/WHEEL +1 -1
- natural_pdf/search/haystack_search_service.py +0 -687
- natural_pdf/search/haystack_utils.py +0 -474
- natural_pdf/utils/tqdm_utils.py +0 -51
- {natural_pdf-0.1.11.dist-info → natural_pdf-0.1.13.dist-info}/licenses/LICENSE +0 -0
- {natural_pdf-0.1.11.dist-info → natural_pdf-0.1.13.dist-info}/top_level.txt +0 -0
natural_pdf/__init__.py
CHANGED
|
@@ -51,10 +51,13 @@ ElementCollection = None
|
|
|
51
51
|
|
|
52
52
|
# Search options (if extras installed)
|
|
53
53
|
try:
|
|
54
|
-
from natural_pdf.search.search_options import
|
|
54
|
+
from natural_pdf.search.search_options import (
|
|
55
|
+
BaseSearchOptions,
|
|
56
|
+
MultiModalSearchOptions,
|
|
57
|
+
TextSearchOptions,
|
|
58
|
+
)
|
|
55
59
|
except ImportError:
|
|
56
60
|
# Define dummy classes if extras not installed, so imports don't break
|
|
57
|
-
# but using them will raise the ImportError from check_haystack_availability
|
|
58
61
|
class BaseSearchOptions:
|
|
59
62
|
def __init__(self, *args, **kwargs):
|
|
60
63
|
pass
|
|
@@ -67,9 +70,11 @@ except ImportError:
|
|
|
67
70
|
def __init__(self, *args, **kwargs):
|
|
68
71
|
pass
|
|
69
72
|
|
|
73
|
+
|
|
70
74
|
# Import QA module if available
|
|
71
75
|
try:
|
|
72
76
|
from natural_pdf.qa import DocumentQA, get_qa_engine
|
|
77
|
+
|
|
73
78
|
HAS_QA = True
|
|
74
79
|
except ImportError:
|
|
75
80
|
HAS_QA = False
|