natural-pdf 0.1.21__py3-none-any.whl → 0.1.23__py3-none-any.whl

natural_pdf/selectors/parser.py

@@ -698,6 +698,28 @@ def _build_filter_list(selector: Dict[str, Any], **kwargs) -> List[Dict[str, Any
 
             filter_lambda = contains_check
 
+        # --- Handle :startswith and :starts-with (alias) --- #
+        elif name in ("starts-with", "startswith") and args is not None:
+            filter_name = f"pseudo-class :{name}({args!r})"
+
+            def startswith_check(element, arg=args):
+                if not hasattr(element, "text") or not element.text:
+                    return False
+                return str(element.text).startswith(str(arg))
+
+            filter_lambda = startswith_check
+
+        # --- Handle :endswith and :ends-with (alias) --- #
+        elif name in ("ends-with", "endswith") and args is not None:
+            filter_name = f"pseudo-class :{name}({args!r})"
+
+            def endswith_check(element, arg=args):
+                if not hasattr(element, "text") or not element.text:
+                    return False
+                return str(element.text).endswith(str(arg))
+
+            filter_lambda = endswith_check
+
         elif name == "starts-with" and args is not None:
             filter_lambda = (
                 lambda el, arg=args: hasattr(el, "text")

{natural_pdf-0.1.21.dist-info → natural_pdf-0.1.23.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: natural-pdf
-Version: 0.1.21
+Version: 0.1.23
 Summary: A more intuitive interface for working with PDFs
 Author-email: Jonathan Soma <jonathan.soma@gmail.com>
 License-Expression: MIT
@@ -21,14 +21,7 @@ Requires-Dist: urllib3
 Requires-Dist: tqdm
 Requires-Dist: pydantic
 Requires-Dist: jenkspy
-Requires-Dist: pikepdf
 Requires-Dist: scipy
-Requires-Dist: torch
-Requires-Dist: torchvision
-Requires-Dist: transformers[sentencepiece]
-Requires-Dist: huggingface_hub>=0.29.3
-Requires-Dist: sentence-transformers
-Requires-Dist: timm
 Requires-Dist: ipywidgets>=7.0.0
 Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
@@ -58,6 +51,7 @@ Requires-Dist: natural-pdf[test]; extra == "all"
 Requires-Dist: natural-pdf[search]; extra == "all"
 Requires-Dist: natural-pdf[favorites]; extra == "all"
 Requires-Dist: natural-pdf[export-extras]; extra == "all"
+Requires-Dist: natural-pdf[ai]; extra == "all"
 Provides-Extra: deskew
 Requires-Dist: deskew>=1.5; extra == "deskew"
 Requires-Dist: img2pdf; extra == "deskew"
@@ -69,6 +63,15 @@ Requires-Dist: pikepdf; extra == "ocr-export"
 Provides-Extra: export-extras
 Requires-Dist: jupytext; extra == "export-extras"
 Requires-Dist: nbformat; extra == "export-extras"
+Provides-Extra: ai
+Requires-Dist: sentence-transformers; extra == "ai"
+Requires-Dist: torch; extra == "ai"
+Requires-Dist: torchvision; extra == "ai"
+Requires-Dist: transformers[sentencepiece]; extra == "ai"
+Requires-Dist: huggingface_hub>=0.29.3; extra == "ai"
+Requires-Dist: timm; extra == "ai"
+Requires-Dist: doclayout_yolo; extra == "ai"
+Requires-Dist: easyocr; extra == "ai"
 Dynamic: license-file
 
 # Natural PDF
@@ -88,25 +91,29 @@ Natural PDF lets you find and extract content from PDFs using simple code that m
 pip install natural-pdf
 ```
 
-Need OCR engines, layout models, or other heavy add-ons? Install the **core** once, then use the helper CLI to pull in exactly what you need:
+Need OCR engines, layout models, or other heavy add-ons? Install the **core** once, then use the helper `npdf` command to pull in exactly what you need:
 
 ```bash
-# add PaddleOCR (+paddlex) after the fact
-npdf install paddle
+# Everything you need for classification, document-QA, semantic search, etc.
+npdf install ai
 
 # Surya OCR and the YOLO Doc-Layout detector in one go
 npdf install surya yolo
 
+# add PaddleOCR (+paddlex) after the fact
+npdf install paddle
+
 # see what's already on your machine
 npdf list
 ```
 
-Light-weight extras such as `deskew` or `search` can still be added with
-classic PEP-508 markers if you prefer:
+Lightweight extras such as `deskew` or `search` can still be added with
+classic `pip install`:
 
 ```bash
 pip install "natural-pdf[deskew]"
 pip install "natural-pdf[search]"
+pip install "natural-pdf[ai]"
 ```
 
 More details in the [installation guide](https://jsoma.github.io/natural-pdf/installation/).
@@ -117,7 +124,7 @@ More details in the [installation guide](https://jsoma.github.io/natural-pdf/ins
 from natural_pdf import PDF
 
 # Open a PDF
-pdf = PDF('document.pdf')
+pdf = PDF('https://github.com/jsoma/natural-pdf/raw/refs/heads/main/pdfs/01-practice.pdf')
 page = pdf.pages[0]
 
 # Extract all of the text on the page

{natural_pdf-0.1.21.dist-info → natural_pdf-0.1.23.dist-info}/RECORD

@@ -1,7 +1,7 @@
 natural_pdf/__init__.py,sha256=qDFJNF8sbEDO-2WSFAxoWEM8updOUP6dB-ckya0kxfs,3275
-natural_pdf/cli.py,sha256=0nAGVO2f_40E3G9c3Q0bfK5mhROyUJH5W25-YJVLMIo,4749
+natural_pdf/cli.py,sha256=IXrP2lCHihr-ed-CFiDbMTnSsutQa1j1PYALOLGbpsc,4019
 natural_pdf/analyzers/__init__.py,sha256=dIXjsMqoxKmd9OOnSBzn12wvdIz7D7YNQRAnXslpJSM,142
-natural_pdf/analyzers/shape_detection_mixin.py,sha256=HHefg-v7CJMxYiJHxdGOdqdtbWe9yk4OBoW3a_aRrjM,81798
+natural_pdf/analyzers/shape_detection_mixin.py,sha256=blpeHMWl6nXlutAByfdi6zjfmcyaDpdv2S7IR4l0WO0,81783
 natural_pdf/analyzers/text_options.py,sha256=qEkDaYWla0rIM_gszEOsu52q7C_dAfV81P2HLJZM2sw,3333
 natural_pdf/analyzers/text_structure.py,sha256=VfKTsTFrK877sC0grsis9jK3rrgp0Mbp13VWEbukTcs,28437
 natural_pdf/analyzers/utils.py,sha256=PYbzJzSAHZ7JsMes84WIrSbA0zkjJGs0CLvIeINsf_k,2100
@@ -18,27 +18,27 @@ natural_pdf/analyzers/layout/surya.py,sha256=4RdnhRxSS3i3Ns5mFhOA9-P0xd7Ms19uZuK
 natural_pdf/analyzers/layout/table_structure_utils.py,sha256=nISZDBd46RPYkFHxbQyIHwg9WweG4DslpoYJ31OMJYA,2768
 natural_pdf/analyzers/layout/tatr.py,sha256=cVr0ZyhY2mNLAKZ4DGMm-b7XNJpILKh8x8ZpyDeUhLk,15032
 natural_pdf/analyzers/layout/yolo.py,sha256=Iw8qsIOHg2lUP7z9GsmkOm3c9kJ-Ywk01Oej50kZgDw,8360
-natural_pdf/classification/manager.py,sha256=pzuTP-34W9N3im1ZFhCfQpOu37VSHEx4JHoHNxyy6o0,18894
-natural_pdf/classification/mixin.py,sha256=_XtoqCMqj1nxZYskIV2RbVYiVVcEWzFwae4s5vpzC74,6566
-natural_pdf/classification/results.py,sha256=El1dY7cBQVOB5lP-uj52dWgH6Y7TeQgJOVcZD-OLjes,2778
+natural_pdf/classification/manager.py,sha256=7HOyHdjMJtC9DfzI8OXAREnGDpXaAgSfTFVC42n3tVQ,18889
+natural_pdf/classification/mixin.py,sha256=nYpmHQ4BlrealdPtIJt-_idME5o-xKLKNuAdIHzWL6c,7580
+natural_pdf/classification/results.py,sha256=Mcay-xLBHbYoZ8U7f4gMj2IhhH_yORNEkZHWdWJzmsU,3239
 natural_pdf/collections/mixins.py,sha256=sj76Cn6EdBtb5f-bdAV-1qpdixX8tI4BzPccPiYLI1w,5117
 natural_pdf/collections/pdf_collection.py,sha256=HLlyakM--23ZOeHDPucoM6Tw3yUyMXm0SSoqJwxRc2E,30744
 natural_pdf/core/__init__.py,sha256=QC8H4M3KbXwMFiQORZ0pdPlzx1Ix6oKKQSS7Ib2KEaA,38
 natural_pdf/core/element_manager.py,sha256=_UdXu51sLi6STzc8Pj4k8R721G3yJixXDLuRHn3hmr8,25731
-natural_pdf/core/highlighting_service.py,sha256=_kQUS6_BBvsLBuSZloFrVag6jN90KzHa0ULyGBjufSs,36955
-natural_pdf/core/page.py,sha256=i3DriIQwoO4RuSrkrCXv44Dz8OL9KXPa2y4GhsD1y18,118324
-natural_pdf/core/pdf.py,sha256=yBvb1iGw9gwVPJ3Rm1EBaZ8_g60TuW_Elhg2EOcJMzc,73871
+natural_pdf/core/highlighting_service.py,sha256=wWoU2kJ_JBbxKV3NWEjqU6DLvmlwME9sTntk-TDqOfs,38223
+natural_pdf/core/page.py,sha256=U4GRy_zdoTB4sx4EPrAIKg4beIQ8atJsY5HX_jWfDjg,118953
+natural_pdf/core/pdf.py,sha256=qsSW4RxOJRmCnweLPMs0NhzkRfiAVdghTgnh4D_wuO4,74295
 natural_pdf/describe/__init__.py,sha256=B3zjuHjFI_dFuBLgXR1Q4v7c72fVDyk84d2hs0H4KV8,561
-natural_pdf/describe/base.py,sha256=7USCFIl4mI5b15LTVkwvhAn_mngMwhwxCnVYaZz5Vdc,16842
-natural_pdf/describe/elements.py,sha256=BOkz2wDhGh6P8NOm6pSNxitgmVokLTISztaFhrxMcdw,12717
+natural_pdf/describe/base.py,sha256=LAZLc_thK2u2surgGd0Pk7CN2uVaZK9AbMOE3-1RmQ4,16842
+natural_pdf/describe/elements.py,sha256=xD8wwR1z5IKat7RIwoAwQRUEL6zJTEwcOKorF4F-xPg,12717
 natural_pdf/describe/mixin.py,sha256=U0x6v8r57KQb8qC3VVo64hvhfXQWsti8vdKBM7AXnMo,3116
 natural_pdf/describe/summary.py,sha256=dPtjrn6fQ8nL0F74RITX2vXlDX7ZgaX9JQPnJB-S_XQ,6735
 natural_pdf/elements/__init__.py,sha256=S8XeiNWJ1WcgnyYKdYV1yxQlAxCCO3FfITT8MQwNbyk,41
-natural_pdf/elements/base.py,sha256=IlAeyzV66xMrxVx9U3ocGPekzGUBJgKkAiJ5kpvCSAg,39675
-natural_pdf/elements/collections.py,sha256=vgVZsVC3xxRF2S5KW7L0JKa-NSUFnqURk50NtvlwbcM,122113
+natural_pdf/elements/base.py,sha256=iTIy6FfQj48llZkm7wERnTky3VTmUgkYfQytRuyueZo,43304
+natural_pdf/elements/collections.py,sha256=zaqJ8pr0dmYwv1gPBs24oXfZExpSIX4URDRox-QLj98,123173
 natural_pdf/elements/line.py,sha256=300kSFBDUBIudfeQtH_tzW9gTYRgRKUDPiTABw6J-BE,4782
 natural_pdf/elements/rect.py,sha256=kiVa3e377ZnqIOXc89d9ZSY4EcmDxtccdtUw-HOQzpw,3796
-natural_pdf/elements/region.py,sha256=hBklYKcXJWyxayu9todYQOZ-d9KVDtqeV-CIt9IcSn8,123400
+natural_pdf/elements/region.py,sha256=BAOriJuQYovppV0S5xI6tq5YEuzffiMQneDvHuT22Uo,118562
 natural_pdf/elements/text.py,sha256=13HvVZGinj2Vm_fFCAnqi7hohtoKvnpCp3VCfkpeAbc,11146
 natural_pdf/export/mixin.py,sha256=L1q3MIEFWuvie4j4_EmW7GT3NerbZ1as0XMUoqTS7gM,5083
 natural_pdf/exporters/__init__.py,sha256=g1WRPCDVzceaUUsm8dchPhzdHFSjYM0NfRyc8iN0mtE,644
@@ -46,31 +46,32 @@ natural_pdf/exporters/base.py,sha256=XhR1xlkHOh7suOuX7mWbsj1h2o1pZNet-OAS5YCJyeI
 natural_pdf/exporters/hocr.py,sha256=wksvJvWLSxuAfhYzg_0T2_W8eqDoMgAVC-gwZ9FoO_k,19969
 natural_pdf/exporters/hocr_font.py,sha256=1wsGOMj6zoaRN2rxCwrv4MMLGawpNz984WgXpmWekgw,4574
 natural_pdf/exporters/original_pdf.py,sha256=dtvC4er6TWOfqq-n24Pejw3mlAuPd8IVyihggJtcf0s,6634
-natural_pdf/exporters/paddleocr.py,sha256=IAG2p9YeImYcsIvb6a_L5mMrKarvaMaDvRrvdlY6bX4,19489
+natural_pdf/exporters/paddleocr.py,sha256=RBP03GCk0mLeC7tWtuti8AIUHlpOrtvbWkE2n7Ja7k8,19484
 natural_pdf/exporters/searchable_pdf.py,sha256=G2Tc4tpDXSYIufXJlkA8ppW_3DuzHAaweYKae33pI_c,16290
 natural_pdf/exporters/data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 natural_pdf/exporters/data/pdf.ttf,sha256=x4RUIJJaI9iO2DCmOVe4r4Wmao2vjZ_JDoQ2c7LvGlk,572
 natural_pdf/exporters/data/sRGB.icc,sha256=KpLUuuRQt22LCqQhk9-XTXX2Jzjs6_dPAcXnWxKpV5Y,6922
-natural_pdf/extraction/manager.py,sha256=mUBbfgLG5Pl31wmajXwyipdEJb_dZ5I-y8GnWw7IzGo,4969
-natural_pdf/extraction/mixin.py,sha256=eKbr70VibpbtfjvCE80lTFuYHzq_BoVtOHjznL_GMRA,11719
-natural_pdf/extraction/result.py,sha256=c1vLguCR6l95cvg-BJJmZvL_MPg2McJaczge55bKZMg,934
+natural_pdf/extraction/manager.py,sha256=sASPJZ5cWFsl8A4PyTjg2yqkyC00tRl6glfoFA6HcsM,4979
+natural_pdf/extraction/mixin.py,sha256=_5wGnzOCEuRWhqdSUV1Lqo9HIi56YC4MWzbBxOkOEKU,23160
+natural_pdf/extraction/result.py,sha256=D5DhjxLW7IvhEkvsAP7Zs2YA8K4hyuoTg681CSn5qA0,1825
 natural_pdf/flows/__init__.py,sha256=cUN4A8hTDLZSRr4PO2W_lR4z6hWpbNG8Seox-IIcrLU,277
 natural_pdf/flows/collections.py,sha256=qGuSPFSPQF-wiYquG6STiSzg_o951MSsFEq_B44Jef8,28441
 natural_pdf/flows/element.py,sha256=mKzk3B7A7sWNvu4CDvAjLr3_ZFLt--ktrSNoLfLpFxU,23940
 natural_pdf/flows/flow.py,sha256=I61BpFVDQyo6ORsmoqoYiOEP1DBRp0vgDJjm_V8frhc,10562
-natural_pdf/flows/region.py,sha256=OJNt1nVt_w89kCw4sQmZRZE5SHgiFWYb8ebcX49jq2A,22192
+natural_pdf/flows/region.py,sha256=hucKKmjjmLt__x-RiX6S1Amsp88yweyjcgWJ7PQtTgY,22187
 natural_pdf/ocr/__init__.py,sha256=VY8hhvDPf7Gh2lB-d2QRmghLLyTy6ydxlgo1cS4dOSk,2482
 natural_pdf/ocr/engine.py,sha256=ZBC1tZNM5EDbGDJJmZI9mNHr4nCMLEZvUFhiJq8GdF4,8741
 natural_pdf/ocr/engine_doctr.py,sha256=ptKrupMWoulZb-R93zr9btoe94JPWU7vlJuN7OBJEIM,17740
 natural_pdf/ocr/engine_easyocr.py,sha256=bWz6kHUgAJfe3rqdnZBAF-IPvw3B35DlvX5KDdFUtzo,9888
 natural_pdf/ocr/engine_paddle.py,sha256=0vobobjnsM1G3zihYL7f1roLlKKZWRwioxkGkgIxEUA,16159
-natural_pdf/ocr/engine_surya.py,sha256=PNjvpsHnBghAoa-df52HEyvXzfNI-gTFgKvs2LxHgKo,5051
+natural_pdf/ocr/engine_surya.py,sha256=lOvSbZk53VKFVxRmqcQzM_0dHVdwTkRGiDZ9AWCgL1Q,5951
 natural_pdf/ocr/ocr_factory.py,sha256=GkODuBmqNVECg4u1-KW6ZMfBgVndLkK1W5GM15faf8M,5318
 natural_pdf/ocr/ocr_manager.py,sha256=K2gpFo3e6RB1ouXOstlEAAYd14DbjBNt5RH6J7ZdDQY,14263
 natural_pdf/ocr/ocr_options.py,sha256=l33QKu_93r-uwi3t_v8UH8pEgHo6HTVzP4tfmQFRF1w,5488
 natural_pdf/ocr/utils.py,sha256=OxuHwDbHWj6setvnC0QYwMHrAjxGkhmLzWHpMqqGupA,4397
-natural_pdf/qa/__init__.py,sha256=Pjo62JTnUNEjGNsC437mvsS5KQ5m7X_BibGvavR9AW0,108
-natural_pdf/qa/document_qa.py,sha256=Jw4yyq3Vifn57D0ANmOfUlZeG8CJjBkItZBV-8ZAmos,15111
+natural_pdf/qa/__init__.py,sha256=2u2KJcA71g1I0HnLD-j6yvDw1moAjo9kkLhhfoYRURM,166
+natural_pdf/qa/document_qa.py,sha256=bwOrO_bq_9wEaLu7j7h8EkN3ya5xMxDoE7oNurEb6-E,14889
+natural_pdf/qa/qa_result.py,sha256=_q4dlSqsjtgomcI8-pqbOT69lqQKnEMkhZNydoxEkkE,2227
 natural_pdf/search/__init__.py,sha256=0Xa7tT_2q57wHObFMQLQLd4gd9AV0oyS-svV6BmmdMI,4276
 natural_pdf/search/lancedb_search_service.py,sha256=6dz2IEZUWk3hFW28C-LF_85pWohd7Sr5k44bM0pBdm4,14472
 natural_pdf/search/numpy_search_service.py,sha256=MoPBlyHTDqah1IrwBzyglEyiXlF4wqaU_5mml_ngvGc,10328
@@ -78,7 +79,7 @@ natural_pdf/search/search_options.py,sha256=sq_e8_jSROicD94b_xtDtLnjEr_Zsy4icjzP
 natural_pdf/search/search_service_protocol.py,sha256=Dl-Q-CrutkhZwI69scbW9EWPeYM63qxB60_EA7YqIYo,6699
 natural_pdf/search/searchable_mixin.py,sha256=hqQ_AuID5eTGRCtKYdFLZ1zF35y73uk3x1M1VW9Il8U,23514
 natural_pdf/selectors/__init__.py,sha256=oZGeqSv53EqmIZOhcnawuaGGlRg1h79vArXuZCWKm4A,123
-natural_pdf/selectors/parser.py,sha256=0EJskGjRfwXiqjedY6siIso4TA4dACHb-_DOyogW9TY,31970
+natural_pdf/selectors/parser.py,sha256=nJJPACzFKBMomAZMXzI_L3EMz9ji1agkmxJ5eaij310,32880
 natural_pdf/templates/__init__.py,sha256=jYBxzfi73vew0f6yhIh1MlRxw4F_TVN2hKQR0YXOFe0,20
 natural_pdf/utils/__init__.py,sha256=s3M8FggaK1P3EBYn6R_-HgSDjNc9C73gyKe1hihtNWg,43
 natural_pdf/utils/debug.py,sha256=RN7H3E6ph-GtxubCW6psW7TO8o2BxcNLiEzByTVR9fk,995
@@ -91,9 +92,9 @@ natural_pdf/utils/text_extraction.py,sha256=z6Jhy11pakYCsEpkvh8ldw6DkUFsYF1hCL9Y
 natural_pdf/utils/visualization.py,sha256=30pRWQdsRJh2pSObh-brKVsFgC1n8tHmSrta_UDnVPw,8989
 natural_pdf/widgets/__init__.py,sha256=QTVaUmsw__FCweFYZebwPssQxxUFUMd0wpm_cUbGZJY,181
 natural_pdf/widgets/viewer.py,sha256=2VUY1TzWMDe9I-IVNOosKZ2LaqpjLB62ftMAdk-s6_8,24952
-natural_pdf-0.1.21.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
-natural_pdf-0.1.21.dist-info/METADATA,sha256=zo7RfNG44xTbSSdOgGjciqEVMOMCxRX3BY7qtH3roW8,6261
-natural_pdf-0.1.21.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-natural_pdf-0.1.21.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
-natural_pdf-0.1.21.dist-info/top_level.txt,sha256=Cyw1zmNDlUZfb5moU-WUWGprrwH7ln_8LDGdmMHF1xI,17
-natural_pdf-0.1.21.dist-info/RECORD,,
+natural_pdf-0.1.23.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
+natural_pdf-0.1.23.dist-info/METADATA,sha256=z7Mq5yr_sckn7pFR1KqBz_fG2sG-jBBSb2czsRrzC_k,6660
+natural_pdf-0.1.23.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+natural_pdf-0.1.23.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
+natural_pdf-0.1.23.dist-info/top_level.txt,sha256=Cyw1zmNDlUZfb5moU-WUWGprrwH7ln_8LDGdmMHF1xI,17
+natural_pdf-0.1.23.dist-info/RECORD,,