docling 2.54.0__py3-none-any.whl → 2.55.0__py3-none-any.whl

docling/utils/api_image_request.py

@@ -1,13 +1,15 @@
 import base64
+import json
 import logging
 from io import BytesIO
-from typing import Dict, Optional
+from typing import Dict, List, Optional
 
 import requests
 from PIL import Image
 from pydantic import AnyUrl
 
 from docling.datamodel.base_models import OpenAiApiResponse
+from docling.models.utils.generation_utils import GenerationStopper
 
 _log = logging.getLogger(__name__)
 
@@ -59,3 +61,107 @@ def api_image_request(
     api_resp = OpenAiApiResponse.model_validate_json(r.text)
     generated_text = api_resp.choices[0].message.content.strip()
     return generated_text
+
+
+def api_image_request_streaming(
+    image: Image.Image,
+    prompt: str,
+    url: AnyUrl,
+    *,
+    timeout: float = 20,
+    headers: Optional[Dict[str, str]] = None,
+    generation_stoppers: List[GenerationStopper] = [],
+    **params,
+) -> str:
+    """
+    Stream a chat completion from an OpenAI-compatible server (e.g., vLLM).
+    Parses SSE lines: 'data: {json}\\n\\n', terminated by 'data: [DONE]'.
+    Accumulates text and calls stopper.should_stop(window) as chunks arrive.
+    If stopper triggers, the HTTP connection is closed to abort server-side generation.
+    """
+    img_io = BytesIO()
+    image.save(img_io, "PNG")
+    image_b64 = base64.b64encode(img_io.getvalue()).decode("utf-8")
+
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {
+                    "type": "image_url",
+                    "image_url": {"url": f"data:image/png;base64,{image_b64}"},
+                },
+                {"type": "text", "text": prompt},
+            ],
+        }
+    ]
+
+    payload = {
+        "messages": messages,
+        "stream": True,  # <-- critical for SSE streaming
+        **params,
+    }
+
+    # Debug: Log the payload to verify temperature is included
+    _log.debug(f"API streaming request payload: {json.dumps(payload, indent=2)}")
+
+    # Some servers require Accept: text/event-stream for SSE.
+    # It's safe to set it; OpenAI-compatible servers tolerate it.
+    hdrs = {"Accept": "text/event-stream", **(headers or {})}
+
+    # Try to force temperature via header if server ignores payload parameter
+    if "temperature" in params:
+        hdrs["X-Temperature"] = str(params["temperature"])
+
+    # Stream the HTTP response
+    with requests.post(
+        str(url), headers=hdrs, json=payload, timeout=timeout, stream=True
+    ) as r:
+        if not r.ok:
+            _log.error(
+                f"Error calling the API {url} in streaming mode. Response was {r.text}"
+            )
+        r.raise_for_status()
+
+        full_text = []
+        for raw_line in r.iter_lines(decode_unicode=True):
+            if not raw_line:  # keep-alives / blank lines
+                continue
+            if not raw_line.startswith("data:"):
+                # Some proxies inject comments; ignore anything not starting with 'data:'
+                continue
+
+            data = raw_line[len("data:") :].strip()
+            if data == "[DONE]":
+                break
+
+            try:
+                obj = json.loads(data)
+            except json.JSONDecodeError:
+                _log.debug("Skipping non-JSON SSE chunk: %r", data[:200])
+                continue
+
+            # OpenAI-compatible delta format
+            # obj["choices"][0]["delta"]["content"] may be None or missing (e.g., tool calls)
+            try:
+                delta = obj["choices"][0].get("delta") or {}
+                piece = delta.get("content") or ""
+            except (KeyError, IndexError) as e:
+                _log.debug("Unexpected SSE chunk shape: %s", e)
+                piece = ""
+
+            if piece:
+                full_text.append(piece)
+                for stopper in generation_stoppers:
+                    # Respect stopper's lookback window. We use a simple string window which
+                    # works with the GenerationStopper interface.
+                    lookback = max(1, stopper.lookback_tokens())
+                    window = "".join(full_text)[-lookback:]
+                    if stopper.should_stop(window):
+                        # Break out of the loop cleanly. The context manager will handle
+                        # 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)

{docling-2.54.0.dist-info → docling-2.55.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling
-Version: 2.54.0
+Version: 2.55.0
 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
@@ -37,7 +37,7 @@ Requires-Dist: requests<3.0.0,>=2.32.2
 Requires-Dist: easyocr<2.0,>=1.7
 Requires-Dist: certifi>=2024.7.4
 Requires-Dist: rtree<2.0.0,>=1.3.0
-Requires-Dist: typer<0.17.0,>=0.12.5
+Requires-Dist: typer<0.20.0,>=0.12.5
 Requires-Dist: python-docx<2.0.0,>=1.1.2
 Requires-Dist: python-pptx<2.0.0,>=1.0.2
 Requires-Dist: beautifulsoup4<5.0.0,>=4.12.3

{docling-2.54.0.dist-info → docling-2.55.0.dist-info}/RECORD

@@ -5,17 +5,17 @@ docling/exceptions.py,sha256=K1WnCS1leK2JtMB5ewZWKkb0EaijFgl-tRzrO9ntgPM,134
 docling/py.typed,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
 docling/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/backend/abstract_backend.py,sha256=1lNxzwDTn303aXduPDVmTyXn-5ZIoWMLYqNxANGWmQQ,1658
-docling/backend/asciidoc_backend.py,sha256=RDNLrPJHxROiM7-NQdZn3DdvAyiPAndbSWcZo9PbCKU,14417
+docling/backend/asciidoc_backend.py,sha256=ARpMBzrNCV-x6g2I0KytDM3pGLac3z4ql3hDKi3FI04,14403
 docling/backend/csv_backend.py,sha256=2g9famYG2W-ID9jEdZPxc6O8QGv1vWQfjN8pL-QMBE0,4536
 docling/backend/docling_parse_backend.py,sha256=9rUo1vPxX6QLzGqF-2B2iEYglZg6YQ3Uea00XrLluTg,7918
 docling/backend/docling_parse_v2_backend.py,sha256=3ckTfke8IICjaImlIzc3TRhG7KDuxDDba0AuCEcjA-M,9500
 docling/backend/docling_parse_v4_backend.py,sha256=xCBbaaXjNNrOaod9tmBuCbe5mL_ipmTNG2XOxVbGG3w,7891
-docling/backend/html_backend.py,sha256=7I3BQSmC7P47jpzXHt3OuPNhtVedJiZVEjjLykyx5pY,42245
-docling/backend/md_backend.py,sha256=qCI7SD9hnWWGrkG_drpzQv2Z7DVBG4Tsq3hhTsYV790,22562
+docling/backend/html_backend.py,sha256=r2m3aIKwwr8Vv2Fxri1FaZFvd4EWvTQlmSPwXeD79zg,47796
+docling/backend/md_backend.py,sha256=zrOUYoIYudUfigwnXRQocb_M4G_ptYfblNgr6BNTYQw,22678
 docling/backend/mets_gbs_backend.py,sha256=EA8sY6tbmGiysKGYPPZiNlK-i7Adn8bLTo-7Ym15hTU,12774
 docling/backend/msexcel_backend.py,sha256=5JRbPwOjR1r45AMeIts1rj6InbOgLBf_CtAhvNPVmsQ,19157
 docling/backend/mspowerpoint_backend.py,sha256=wJgB2JStEPfD7MPpWQlpPN7bffPxaHFUnKD4wj8SLxU,15114
-docling/backend/msword_backend.py,sha256=kQI9hrx_lvHn__KdxW8MbvB78snoVzA_m4jXx6f_LJ8,54419
+docling/backend/msword_backend.py,sha256=Jfd57hzG8iFVAzqsOAHe5jG8LCHAIBXJhQCW0tESnMM,54405
 docling/backend/noop_backend.py,sha256=EOPbD86FzZPX-K_DpNrJh0_lC0bZz--4DpG-OagDNGY,1688
 docling/backend/pdf_backend.py,sha256=Wcd1NSrAMjXK8VicTki5p-j-JLofklt07eF0kIG17_0,3361
 docling/backend/pypdfium2_backend.py,sha256=AYhWs9S8W_TkAK0-OkRmUNf4HUZl26FP7-XYjwU5zDk,14209
@@ -27,11 +27,11 @@ docling/backend/docx/latex/omml.py,sha256=4vh9FCbXh-Tb6KJGqNwzlMUMYEnnJgBtBI24dw
 docling/backend/json/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/backend/json/docling_json_backend.py,sha256=LlFMVoZrrCfVwbDuRbNN4Xg96Lujh4xxrTBt9jGhY9I,1984
 docling/backend/xml/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/backend/xml/jats_backend.py,sha256=LPj33EFdi2MRCakkLWrRLlUAc-B-949f8zp5gKNvBcg,25238
-docling/backend/xml/uspto_backend.py,sha256=nyAMr5ht7dclxkVDwsKNeiOhLQrUtRLS8JdscB2AVJg,70924
+docling/backend/xml/jats_backend.py,sha256=_BWpQQg3SlsHAOOj0v2qRJoVqaQzL91GqN1tK9LxqWo,29463
+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=J_hXHclzT-uNu-cuKNdlc3vwCnyDRxXrJ5L2LJofzeo,32729
+docling/cli/main.py,sha256=UX-5fRGVP_yGxTQez0x1PNnaNKRgWdcXGoPCHy-0uFM,32887
 docling/cli/models.py,sha256=rw_2JfeJ-k_iOLpz3JfgL1QbJY__W9nE23nHdov6VfU,6252
 docling/cli/tools.py,sha256=QhtRxQG0TVrfsMqdv5i7J0_qQy1ZZyWYnHPwJl7b5oY,322
 docling/datamodel/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -43,12 +43,12 @@ docling/datamodel/extraction.py,sha256=7dgvtK5SuvgfB8LHAwS1FwrW1kcMQJuJG0ol8uAQg
 docling/datamodel/layout_model_specs.py,sha256=GSkJ-Z_0PVgwWGi7C7TsxbzRjlrWS9ZrHJjHumv-Z5U,2339
 docling/datamodel/pipeline_options.py,sha256=28opZ3woXA8IKaG2-BHM-lmmi-gyuScCMHGxhlxGOsk,11290
 docling/datamodel/pipeline_options_asr_model.py,sha256=7X068xl-qpbyPxC7-TwX7Q6tLyZXGT5h1osZ_xLNLM0,1454
-docling/datamodel/pipeline_options_vlm_model.py,sha256=AcqqThSW74hwQ6x7pazzm57LnJiUqB7gQi5wFayGlbk,2628
+docling/datamodel/pipeline_options_vlm_model.py,sha256=Szdq5_MhqQ8xBCvOUkdn_LLV29ZMQJcF4xnItYlkmXQ,3090
 docling/datamodel/settings.py,sha256=c0MTw6pO5be_BKxHKYl4SaBJAw_qL-aapxp-g5HHj1A,2084
-docling/datamodel/vlm_model_specs.py,sha256=UMXiTzWCXcx2BtF5slYfWhjRXAx0s1oiAvE-vCzrATo,9686
+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=-zisU32pgDRbychyG6-neB0qweNbPaYnLXwiGT7SEdI,2859
-docling/models/base_model.py,sha256=beMGyrpl-yYX3YnLzQkLfxMLxwmDWnbcFhkjbUlWJSU,7146
+docling/models/api_vlm_model.py,sha256=iNQ9LiT031Mch-LHn8O2CskVXYkr4weEetZPxynU_9U,4236
+docling/models/base_model.py,sha256=LSaJWkSaDyLBVB4Fv9fkw6kmJ67QnG0t32iGn_u2WjE,7256
 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
@@ -61,7 +61,7 @@ docling/models/picture_description_api_model.py,sha256=o3EkV5aHW_6WzE_fdj_VRnNCr
 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/rapid_ocr_model.py,sha256=anUVUwaj9Wubgu4FnHdYMuOVkQP_hJiLY1qRToelBoc,7700
-docling/models/readingorder_model.py,sha256=bZoXHaSwUsa8niSmJrbCuy784ixCeBXT-RQBUfgHJ4A,14925
+docling/models/readingorder_model.py,sha256=_usJdpM4GMWeGGneEwLLxa9grIGQb0XnNMugV72jGbY,14911
 docling/models/table_structure_model.py,sha256=7g_mFf1YzfF8PXQfefNu6XYZu7TzJAn86zKb6IEUdCg,12518
 docling/models/tesseract_ocr_cli_model.py,sha256=I3Gn28Y-LD8OfvyCElN9fLiNgpo2sT0uMkVt258253s,12881
 docling/models/tesseract_ocr_model.py,sha256=GdI5Cjfi87qcehVbM3wdKRvKkl_F9A4bwTUbjXZCJYA,10745
@@ -72,12 +72,13 @@ docling/models/factories/picture_description_factory.py,sha256=Ru3-TnVVEKf5O07C_
 docling/models/plugins/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/models/plugins/defaults.py,sha256=OAHWW2tCcUXSyDMFxV_lXVRjSBJ1n6z-Eb3R8cDucU4,886
 docling/models/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/models/utils/hf_model_download.py,sha256=scBEfsM4yl7xPzqe7UtPvDh9RfQZQnuOhqQKilYBHls,984
+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=wcBsUGHuPNoZn9pfQh4cyyZi1bFGCoF30T_MFVGDyyM,12128
-docling/models/vlm_models_inline/mlx_model.py,sha256=VP05v97mqzmaG4o9bOpJcxIlEqvNzAapJ15Zz3E3ACI,10169
+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=_EnK1nfpAPJky7aRlyp8SUIghiZOQO8AkDN_hHqXLZg,8615
+docling/models/vlm_models_inline/vllm_model.py,sha256=vXClayYxPGX1jzQ1Rvf3vvwtW9khgApGvcRz4Qbyu7I,10293
 docling/pipeline/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/pipeline/asr_pipeline.py,sha256=S55VHLoX3Mgauen1YP-PSUlI0LA1bgTgTkU-eC4U-dg,8481
 docling/pipeline/base_extraction_pipeline.py,sha256=GYrEz83IXv-tdIHjtNWxMBNczFwL8SZyf9vnPJ3STaI,2627
@@ -89,7 +90,7 @@ docling/pipeline/threaded_standard_pdf_pipeline.py,sha256=i67G5AOW7PIFCe5JS2sdBm
 docling/pipeline/vlm_pipeline.py,sha256=oMcdgzym_UQbVN3bajux_hENY40XGOnb6NU6Kwje2Os,15376
 docling/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/utils/accelerator_utils.py,sha256=DSajLxVx1JEVT0zt5de26llciLNlVfIDfSa2zYCFJzQ,2909
-docling/utils/api_image_request.py,sha256=_CgdzmPqdsyXmyYUFGLZcXcoH586qC6A1p5vsNbj1Q0,1416
+docling/utils/api_image_request.py,sha256=kQDmTvQT6M2IgXnGYeoNflI6sLUG6WTCcEft94CRwWg,5379
 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=sE9UR3Nv4iOk26uoIsN3bFioE7ScfAjj0orDBDneLXg,25166
@@ -100,9 +101,9 @@ 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-2.54.0.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
-docling-2.54.0.dist-info/METADATA,sha256=_GsdUYyPCv8XKeLeSO9Y0euAH8Eanr5i_y5kLvDEb1g,11252
-docling-2.54.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-docling-2.54.0.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
-docling-2.54.0.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
-docling-2.54.0.dist-info/RECORD,,
+docling-2.55.0.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
+docling-2.55.0.dist-info/METADATA,sha256=e1RK_bATZ2Q_Ie9kC6uHFCj99D7pkW678jxk_l0CHxk,11252
+docling-2.55.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling-2.55.0.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
+docling-2.55.0.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
+docling-2.55.0.dist-info/RECORD,,