docling 2.58.0__py3-none-any.whl → 2.60.0__py3-none-any.whl

docling/utils/api_image_request.py

@@ -2,7 +2,7 @@ import base64
 import json
 import logging
 from io import BytesIO
-from typing import Optional
+from typing import Dict, List, Optional, Tuple
 
 import requests
 from PIL import Image
@@ -21,7 +21,7 @@ def api_image_request(
     timeout: float = 20,
     headers: Optional[dict[str, str]] = None,
     **params,
-) -> str:
+) -> Tuple[str, Optional[int]]:
     img_io = BytesIO()
     image.save(img_io, "PNG")
     image_base64 = base64.b64encode(img_io.getvalue()).decode("utf-8")
@@ -60,7 +60,8 @@ def api_image_request(
 
     api_resp = OpenAiApiResponse.model_validate_json(r.text)
     generated_text = api_resp.choices[0].message.content.strip()
-    return generated_text
+    num_tokens = api_resp.usage.total_tokens
+    return generated_text, num_tokens
 
 
 def api_image_request_streaming(
@@ -72,7 +73,7 @@ def api_image_request_streaming(
     headers: Optional[dict[str, str]] = None,
     generation_stoppers: list[GenerationStopper] = [],
     **params,
-) -> str:
+) -> Tuple[str, Optional[int]]:
     """
     Stream a chat completion from an OpenAI-compatible server (e.g., vLLM).
     Parses SSE lines: 'data: {json}\\n\\n', terminated by 'data: [DONE]'.
@@ -150,6 +151,16 @@ def api_image_request_streaming(
                 _log.debug("Unexpected SSE chunk shape: %s", e)
                 piece = ""
 
+            # Try to extract token count
+            num_tokens = None
+            try:
+                if "usage" in obj:
+                    usage = obj["usage"]
+                    num_tokens = usage.get("total_tokens")
+            except Exception as e:
+                num_tokens = None
+                _log.debug("Usage key not included in response: %s", e)
+
             if piece:
                 full_text.append(piece)
                 for stopper in generation_stoppers:
@@ -162,6 +173,6 @@ def api_image_request_streaming(
                         # closing the connection when we exit the 'with' block.
                         # vLLM/OpenAI-compatible servers will detect the client disconnect
                         # and abort the request server-side.
-                        return "".join(full_text)
+                        return "".join(full_text), num_tokens
 
-        return "".join(full_text)
+        return "".join(full_text), num_tokens

{docling-2.58.0.dist-info → docling-2.60.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling
-Version: 2.58.0
+Version: 2.60.0
 Summary: SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
 Author-email: Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>, Nikos Livathinos <nli@zurich.ibm.com>, Ahmed Nassar <ahn@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 License-Expression: MIT
@@ -22,6 +22,7 @@ Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
 Requires-Python: <4.0,>=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
@@ -45,7 +46,7 @@ Requires-Dist: beautifulsoup4<5.0.0,>=4.12.3
 Requires-Dist: pandas<3.0.0,>=2.1.4
 Requires-Dist: marko<3.0.0,>=2.1.2
 Requires-Dist: openpyxl<4.0.0,>=3.1.5
-Requires-Dist: lxml<6.0.0,>=4.0.0
+Requires-Dist: lxml<7.0.0,>=4.0.0
 Requires-Dist: pillow<12.0.0,>=10.0.0
 Requires-Dist: tqdm<5.0.0,>=4.65.0
 Requires-Dist: pluggy<2.0.0,>=1.0.0
@@ -62,15 +63,15 @@ Requires-Dist: ocrmac<2.0.0,>=1.0.0; sys_platform == "darwin" and extra == "ocrm
 Provides-Extra: vlm
 Requires-Dist: transformers<5.0.0,>=4.46.0; extra == "vlm"
 Requires-Dist: accelerate<2.0.0,>=1.2.1; extra == "vlm"
-Requires-Dist: mlx-vlm<1.0.0,>=0.3.0; (python_version >= "3.10" and sys_platform == "darwin" and platform_machine == "arm64") and extra == "vlm"
-Requires-Dist: vllm<1.0.0,>=0.10.0; (python_version >= "3.10" and sys_platform == "linux" and platform_machine == "x86_64") and extra == "vlm"
+Requires-Dist: mlx-vlm<1.0.0,>=0.3.0; (python_version >= "3.10" and python_version < "3.14" and sys_platform == "darwin" and platform_machine == "arm64") and extra == "vlm"
+Requires-Dist: vllm<1.0.0,>=0.10.0; (python_version >= "3.10" and python_version < "3.14" and sys_platform == "linux" and platform_machine == "x86_64") and extra == "vlm"
 Requires-Dist: qwen-vl-utils>=0.0.11; extra == "vlm"
 Provides-Extra: rapidocr
-Requires-Dist: rapidocr<4.0.0,>=3.3; python_version < "3.14" and extra == "rapidocr"
-Requires-Dist: onnxruntime<2.0.0,>=1.7.0; extra == "rapidocr"
+Requires-Dist: rapidocr<4.0.0,>=3.3; extra == "rapidocr"
+Requires-Dist: onnxruntime<2.0.0,>=1.7.0; python_version < "3.14" and extra == "rapidocr"
 Provides-Extra: asr
-Requires-Dist: mlx-whisper>=0.4.3; (python_version >= "3.10" and sys_platform == "darwin" and platform_machine == "arm64") and extra == "asr"
-Requires-Dist: openai-whisper>=20250625; extra == "asr"
+Requires-Dist: mlx-whisper>=0.4.3; (python_version >= "3.10" and python_version < "3.14" and sys_platform == "darwin" and platform_machine == "arm64") and extra == "asr"
+Requires-Dist: openai-whisper>=20250625; python_version < "3.14" and extra == "asr"
 Dynamic: license-file
 
 <p align="center">

{docling-2.58.0.dist-info → docling-2.60.0.dist-info}/RECORD

@@ -13,12 +13,12 @@ docling/backend/docling_parse_v4_backend.py,sha256=tBJR0BbKFOIDKSngjVDu0BrzTj7qU
 docling/backend/html_backend.py,sha256=m91kRxMhQ1w-7G6MHA9l01dgF8-YQNn8ZNx9lwG467M,52935
 docling/backend/md_backend.py,sha256=_0ToiecsGwU4H4BBso4ar9TGJi8OTwSXjgmi66vSJVQ,23513
 docling/backend/mets_gbs_backend.py,sha256=EA8sY6tbmGiysKGYPPZiNlK-i7Adn8bLTo-7Ym15hTU,12774
-docling/backend/msexcel_backend.py,sha256=-iWLdIonMZl2FCfPAXFQKIQzFOJn5InpH6KDAJ_L64o,22760
+docling/backend/msexcel_backend.py,sha256=ujU8qoevNhLDWffihMlSYFVl7B3y_Uu5g-yispWyt8Q,22868
 docling/backend/mspowerpoint_backend.py,sha256=71W_iV31Rggqn9UcMzXmsZ3QKMRpsBT8fCwdjsIIKAs,15109
 docling/backend/msword_backend.py,sha256=zNJy-KM3Ia-L8IQ4sjYxATW4owFxbg2CK0rzke8y-7w,57451
 docling/backend/noop_backend.py,sha256=EOPbD86FzZPX-K_DpNrJh0_lC0bZz--4DpG-OagDNGY,1688
 docling/backend/pdf_backend.py,sha256=UovGV3RJG6qllzMPYzhDB6GID7buGV6w1uxl5dOAEw4,3563
-docling/backend/pypdfium2_backend.py,sha256=tx0FnUW87zPsyafCvOuLcls2k5QdpPKWweyjNTfclNc,14509
+docling/backend/pypdfium2_backend.py,sha256=M1bo20H271txPogI4NgMuc7b-zHJ_BuPDJifEAfjQ88,14497
 docling/backend/webvtt_backend.py,sha256=9xPcfWVLuqhEAFrkv8aU36qHnSgjeINZAXT_C9C6XJA,19165
 docling/backend/docx/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/backend/docx/drawingml/utils.py,sha256=E9Iq8_052eEV5L1IN3ZqFX9eBidH56DKNlh6Tk7Do0I,3640
@@ -32,37 +32,37 @@ docling/backend/xml/jats_backend.py,sha256=_BWpQQg3SlsHAOOj0v2qRJoVqaQzL91GqN1tK
 docling/backend/xml/uspto_backend.py,sha256=Tv4CE7V5_QwxTNJPl90CAd_mAbwaLGy8S6s6evh1Xow,70910
 docling/chunking/__init__.py,sha256=h83TDs0AuOV6oEPLAPrn9dpGKiU-2Vg6IRNo4cv6GDA,346
 docling/cli/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/cli/main.py,sha256=x_mPS3g3Zw60_9bL_oo9OfPBmuSd-aJV7oKTPD0GjS4,36772
+docling/cli/main.py,sha256=T7MllU1e2zYoKekpEHPv7VdI4cypL6K5zzCfscHCRro,37404
 docling/cli/models.py,sha256=zZBFQJAD7C5sespnYy5M__4qC_GyqAZ-QpfWtgPRDB0,6343
 docling/cli/tools.py,sha256=QhtRxQG0TVrfsMqdv5i7J0_qQy1ZZyWYnHPwJl7b5oY,322
 docling/datamodel/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/datamodel/accelerator_options.py,sha256=wv6dOFTVAwr9onkE-0pfUqX_fDb6gX53iPPE6o8nKjI,2511
 docling/datamodel/asr_model_specs.py,sha256=gQJkW7DaSPiOuW_0QoI5OzR1_DQGRkw7yQlrVJ4hyo0,14473
 docling/datamodel/backend_options.py,sha256=2zSbJRtBmJ6Twywj8pLOKaHhklY85XaGXUmSLX_SfgQ,2473
-docling/datamodel/base_models.py,sha256=pC9CvVxMzcujKAG0TTObkYznKp8gIFdzDMDmgk5FjMQ,12697
+docling/datamodel/base_models.py,sha256=AmKIWnqjKo0WgUg6SsHJpN_et_B4rR6em0NEfJ1JKxU,12821
 docling/datamodel/document.py,sha256=T9OogC1kIm0VDSC2ZFcFgWdcOjXzw5JvGr2y2hMlx3s,18795
 docling/datamodel/extraction.py,sha256=7dgvtK5SuvgfB8LHAwS1FwrW1kcMQJuJG0ol8uAQgoQ,1323
 docling/datamodel/layout_model_specs.py,sha256=GSkJ-Z_0PVgwWGi7C7TsxbzRjlrWS9ZrHJjHumv-Z5U,2339
-docling/datamodel/pipeline_options.py,sha256=dklSaA7P6VkjbBB-Pz2OyzO2SQuV9y0I8VVr9XHJusw,11692
+docling/datamodel/pipeline_options.py,sha256=CWWpoPHnYEZBmd_lyZWvwuKrpyrRt4C_YoekjK5VlCc,11807
 docling/datamodel/pipeline_options_asr_model.py,sha256=cLqtRHBr2kbTNXRJ1ZhFGiXIK7Nl9RFmz2Wd7tJF2Jg,2172
-docling/datamodel/pipeline_options_vlm_model.py,sha256=Szdq5_MhqQ8xBCvOUkdn_LLV29ZMQJcF4xnItYlkmXQ,3090
+docling/datamodel/pipeline_options_vlm_model.py,sha256=JBdpfN3nASD5_DaAUe0tla20-Mia8fkveyNw7wVTJ4c,3131
 docling/datamodel/settings.py,sha256=c0MTw6pO5be_BKxHKYl4SaBJAw_qL-aapxp-g5HHj1A,2084
 docling/datamodel/vlm_model_specs.py,sha256=9TTmihDEFcI-TY1jJ2GTnTcrGa3bLg0e6anN4gPtFgU,10035
 docling/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/models/api_vlm_model.py,sha256=iNQ9LiT031Mch-LHn8O2CskVXYkr4weEetZPxynU_9U,4236
+docling/models/api_vlm_model.py,sha256=tZHXS_weqkhgVse1JbrpvjzAyCxW8br78eRYrlMSG3k,4321
 docling/models/auto_ocr_model.py,sha256=nn_eQfNdGUclXKrB0nodHmCqgMUNUJzG3dLq0lhlNAI,5188
 docling/models/base_model.py,sha256=QEbglxu3kT6aNq3x_5jY8T_KcD_Hhv9zr0-A4Mizhco,7252
 docling/models/base_ocr_model.py,sha256=kT8TylASOpPlY60rIG6VL6_eLVsfg5KvEVnZHzDWtR0,8193
 docling/models/code_formula_model.py,sha256=XRugm4EwifLRc-TrAk-glKlktJP-nAPneKh2EOovkJU,11308
 docling/models/document_picture_classifier.py,sha256=9JvoWeH5uQBC7levjM8zptk7UT-b8EQnD-2EnxTjTT4,6202
 docling/models/easyocr_model.py,sha256=sCWIe2xUYU1uK8g4qkPXh0OkUX-rV6BRd4Deb_na9Y4,7899
-docling/models/layout_model.py,sha256=2D7Ey2Mvtzyq9KARIFLaUZKzsR661h7Zggwn0IM9R3c,9154
+docling/models/layout_model.py,sha256=_AGu_Fw_GsLJBQPAIWAblofj1GbSsbvtU-KlhrBdbu4,9309
 docling/models/ocr_mac_model.py,sha256=y-1DSFDbACHpEwNTfQwzN9ab8r5j5rBFNPtQ48BzsrA,5396
 docling/models/page_assemble_model.py,sha256=TvN1naez7dUodLxpUUBzpuMCpqZBTf6YSpewxgjzmrg,6323
 docling/models/page_preprocessing_model.py,sha256=EmusNexws5ZmR93js_saVU0BedqZ_HIHQeY7lcf52tI,5284
 docling/models/picture_description_api_model.py,sha256=o3EkV5aHW_6WzE_fdj_VRnNCrS_btclO_ZCLAUqrfl0,2377
 docling/models/picture_description_base_model.py,sha256=kLthLhdlgwhootQ4_xhhcAk6A-vso5-qcsFJ3TcYfO0,2991
-docling/models/picture_description_vlm_model.py,sha256=Uja_BQSk7F-U1J2hm4yeLguirUzKYv1K8zRyw1IYomY,4150
+docling/models/picture_description_vlm_model.py,sha256=7-reEy5gNxKgOB-VMiysemTwoasZhO5H8VyX4NUEY-4,4272
 docling/models/rapid_ocr_model.py,sha256=JGeed1aNO64SYFgxlOifdut4fynUJyBuyyQrfuSno-4,13182
 docling/models/readingorder_model.py,sha256=gnRFfJAXH-zKtQJws5Zb1_KCVvu_dAq9pgaDYQKCt9s,17236
 docling/models/table_structure_model.py,sha256=7g_mFf1YzfF8PXQfefNu6XYZu7TzJAn86zKb6IEUdCg,12518
@@ -78,22 +78,23 @@ docling/models/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hS
 docling/models/utils/generation_utils.py,sha256=0ZfMBMbolHAWjdbMza8FbD4_jQ4VY6ReUa4gqVLwMoU,5365
 docling/models/utils/hf_model_download.py,sha256=VlKna9tLIVOGQkIRQBXfDimPIIyeRV7cFCbuOVmFQiU,1092
 docling/models/vlm_models_inline/__init__.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-docling/models/vlm_models_inline/hf_transformers_model.py,sha256=Vr6ZIKMVBGQrb0tXl0dVuqYMorDPEnNdF1axAkUpF1Q,14785
-docling/models/vlm_models_inline/mlx_model.py,sha256=ae7hDMgBsMLkqulmbKDamGSSrLJcroYsP1HApJ90IZM,13471
-docling/models/vlm_models_inline/nuextract_transformers_model.py,sha256=jLNtlkMDheUyWot7Oqq-GHQIYzJ0fZrbReq5xCnYb9E,10506
-docling/models/vlm_models_inline/vllm_model.py,sha256=vXClayYxPGX1jzQ1Rvf3vvwtW9khgApGvcRz4Qbyu7I,10293
+docling/models/vlm_models_inline/hf_transformers_model.py,sha256=ylhdnY6A2nUkLQ2Ki-o-Jn8_kjO-JbYKdhnDXmGPB7Y,15047
+docling/models/vlm_models_inline/mlx_model.py,sha256=_q1fVmVaEfnKTVp78djO4MSUA7LrF0JtCnMjTKnotT8,13749
+docling/models/vlm_models_inline/nuextract_transformers_model.py,sha256=f-Djq2G6JLT-RE2LoEP3b2Q-LI33NsGM7Qxo4f6TkeA,10768
+docling/models/vlm_models_inline/vllm_model.py,sha256=gIGZha3YCPBlJGgbjtqpRkiNrOqQszsOT3ZZZu1xbYo,11671
 docling/pipeline/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/pipeline/asr_pipeline.py,sha256=44lweVOCkFe8KikgXJjqDtfHewIotYvc242Xvgl9fV0,15744
+docling/pipeline/asr_pipeline.py,sha256=rzEMHkbZfTmCwl4mjMa2bWRlVmkajC5nKBaY0bT7qj0,16020
 docling/pipeline/base_extraction_pipeline.py,sha256=GYrEz83IXv-tdIHjtNWxMBNczFwL8SZyf9vnPJ3STaI,2627
 docling/pipeline/base_pipeline.py,sha256=NPMQDTyis-LgQ4SybY2f5AESZl5PxogF-FRQuCDckXg,12748
 docling/pipeline/extraction_vlm_pipeline.py,sha256=veUOTe8nGdnduZKaGn1RRb-NfU1H6t_EN4QAsb022Zg,8260
+docling/pipeline/legacy_standard_pdf_pipeline.py,sha256=dFTFpwwgeELNCsstHvhQe7tD4REmOXLoAIFGpGTVo3c,10491
 docling/pipeline/simple_pipeline.py,sha256=FSL_ucDd9k0D9DjNKMUkyCULIU8a057dvWfLEPmAc2A,2287
-docling/pipeline/standard_pdf_pipeline.py,sha256=xOge0zP5wli51n_6QLrFHQlwwvsivI7OMt00tht3my4,10479
-docling/pipeline/threaded_standard_pdf_pipeline.py,sha256=i67G5AOW7PIFCe5JS2sdBmPAKvAH6ScxIBhjwOGZcrI,28183
+docling/pipeline/standard_pdf_pipeline.py,sha256=pPtimWbR7Iww7u2qIWd4vy6Ruc3Dk9wWS7aXqkeP-7c,31446
+docling/pipeline/threaded_standard_pdf_pipeline.py,sha256=SO2ezB6QKpYKCJ1STctJNPoPyRQyf2jTksO1CRSukGc,200
 docling/pipeline/vlm_pipeline.py,sha256=HSbSoGZyy4eIK8eOL2g_NymrHg8r-DrB2buggJQAqHU,16189
 docling/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/utils/accelerator_utils.py,sha256=DSajLxVx1JEVT0zt5de26llciLNlVfIDfSa2zYCFJzQ,2909
-docling/utils/api_image_request.py,sha256=xrn4O8ax8wdQPkLgbRhX22qWUangCXwaOzIXy_86LCs,5367
+docling/utils/api_image_request.py,sha256=HO-FrZ8kOqMMRVJSIsH3apoNoDKM2l7xrC8NfWAEgFQ,5876
 docling/utils/export.py,sha256=VwVUnYDk3mhGmISDbVm306fwpGNnoojouStBD4UajXI,4673
 docling/utils/glm_utils.py,sha256=TKOWQqWAHsX_w4fvoAA7_2xCi_urhnp1DsmjY8_sk5w,12274
 docling/utils/layout_postprocessor.py,sha256=bwDIhgUg5rKianzccGPTotTjqjkWtIQSoZwgKio8YC4,25124
@@ -104,9 +105,9 @@ docling/utils/orientation.py,sha256=jTyLxyT31FlOodZoBMlADHNQK2lAWKYVs5z7pXd_6Cg,
 docling/utils/profiling.py,sha256=YaMGoB9MMZpagF9mb5ndoHj8Lpb9aIdb7El-Pl7IcFs,1753
 docling/utils/utils.py,sha256=kJtIYuzXeOyJHYlxmLAo7dGM5rEsDa1i84qEsUj1nio,1908
 docling/utils/visualization.py,sha256=tY2ylE2aiQKkmzlSLnFW-HTfFyqUUMguW18ldd1PLfo,2868
-docling-2.58.0.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
-docling-2.58.0.dist-info/METADATA,sha256=py9js2V38fIWft1SmMe_iD_trav0WEwojgwxlHMsNv4,11642
-docling-2.58.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-docling-2.58.0.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
-docling-2.58.0.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
-docling-2.58.0.dist-info/RECORD,,
+docling-2.60.0.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
+docling-2.60.0.dist-info/METADATA,sha256=OzgtvKwzf5G11YSPtypooStfqsna-huopIphSFZTeC8,11805
+docling-2.60.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling-2.60.0.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
+docling-2.60.0.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
+docling-2.60.0.dist-info/RECORD,,