chore: bump version to 1.1.0 [skip ci]

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>

CHANGELOG.md

@@ -1,3 +1,15 @@
+## [v1.1.0](https://github.com/docling-project/docling-serve/releases/tag/v1.1.0) - 2025-07-30
+
+### Feature
+
+* Add docling-mcp in the distribution ([#290](https://github.com/docling-project/docling-serve/issues/290)) ([`ecb1874`](https://github.com/docling-project/docling-serve/commit/ecb1874a507bef83d102e0e031e49fed34298637))
+* Add 3.0 openapi endpoint ([#287](https://github.com/docling-project/docling-serve/issues/287)) ([`ec594d8`](https://github.com/docling-project/docling-serve/commit/ec594d84fe36df23e7d010a2fcf769856c43600b))
+* Add new source and target ([#270](https://github.com/docling-project/docling-serve/issues/270)) ([`3771c1b`](https://github.com/docling-project/docling-serve/commit/3771c1b55403bd51966d07d8f760d5c4fbcc1760))
+
+### Fix
+
+* Referenced paths relative to zip root ([#289](https://github.com/docling-project/docling-serve/issues/289)) ([`1333f71`](https://github.com/docling-project/docling-serve/commit/1333f71c9c6495342b2169d574e921f828446f15))
+
 ## [v1.0.1](https://github.com/docling-project/docling-serve/releases/tag/v1.0.1) - 2025-07-21
 
 ### Fix

docling_serve/app.py

@@ -1,4 +1,5 @@
 import asyncio
+import copy
 import importlib.metadata
 import logging
 import shutil
@@ -24,7 +25,7 @@ from fastapi.openapi.docs import (
     get_swagger_ui_html,
     get_swagger_ui_oauth2_redirect_html,
 )
-from fastapi.responses import RedirectResponse
+from fastapi.responses import JSONResponse, RedirectResponse
 from fastapi.staticfiles import StaticFiles
 from scalar_fastapi import get_scalar_api_reference
 
@@ -34,8 +35,13 @@ from docling_jobkit.datamodel.callback import (
     ProgressCallbackResponse,
 )
 from docling_jobkit.datamodel.http_inputs import FileSource, HttpSource
+from docling_jobkit.datamodel.s3_coords import S3Coordinates
 from docling_jobkit.datamodel.task import Task, TaskSource
-from docling_jobkit.datamodel.task_targets import InBodyTarget, TaskTarget, ZipTarget
+from docling_jobkit.datamodel.task_targets import (
+    InBodyTarget,
+    TaskTarget,
+    ZipTarget,
+)
 from docling_jobkit.orchestrators.base_orchestrator import (
     BaseOrchestrator,
     ProgressInvalid,
@@ -47,6 +53,7 @@ from docling_serve.datamodel.requests import (
     ConvertDocumentsRequest,
     FileSourceRequest,
     HttpSourceRequest,
+    S3SourceRequest,
     TargetName,
 )
 from docling_serve.datamodel.responses import (
@@ -54,6 +61,7 @@ from docling_serve.datamodel.responses import (
     ConvertDocumentResponse,
     HealthCheckResponse,
     MessageKind,
+    PresignedUrlConvertDocumentResponse,
     TaskStatusResponse,
     WebsocketMessage,
 )
@@ -246,6 +254,8 @@ def create_app():  # noqa: C901
                 sources.append(FileSource.model_validate(s))
             elif isinstance(s, HttpSourceRequest):
                 sources.append(HttpSource.model_validate(s))
+            elif isinstance(s, S3SourceRequest):
+                sources.append(S3Coordinates.model_validate(s))
 
         task = await orchestrator.enqueue(
             sources=sources,
@@ -286,10 +296,79 @@ def create_app():  # noqa: C901
             if elapsed_time > docling_serve_settings.max_sync_wait:
                 return False
 
+    ##########################################
+    # Downgrade openapi 3.1 to 3.0.x helpers #
+    ##########################################
+
+    def ensure_array_items(schema):
+        """Ensure that array items are defined."""
+        if "type" in schema and schema["type"] == "array":
+            if "items" not in schema or schema["items"] is None:
+                schema["items"] = {"type": "string"}
+            elif isinstance(schema["items"], dict):
+                if "type" not in schema["items"]:
+                    schema["items"]["type"] = "string"
+
+    def handle_discriminators(schema):
+        """Ensure that discriminator properties are included in required."""
+        if "discriminator" in schema and "propertyName" in schema["discriminator"]:
+            prop = schema["discriminator"]["propertyName"]
+            if "properties" in schema and prop in schema["properties"]:
+                if "required" not in schema:
+                    schema["required"] = []
+                if prop not in schema["required"]:
+                    schema["required"].append(prop)
+
+    def handle_properties(schema):
+        """Ensure that property 'kind' is included in required."""
+        if "properties" in schema and "kind" in schema["properties"]:
+            if "required" not in schema:
+                schema["required"] = []
+            if "kind" not in schema["required"]:
+                schema["required"].append("kind")
+
+    # Downgrade openapi 3.1 to 3.0.x
+    def downgrade_openapi31_to_30(spec):
+        def strip_unsupported(obj):
+            if isinstance(obj, dict):
+                obj = {
+                    k: strip_unsupported(v)
+                    for k, v in obj.items()
+                    if k not in ("const", "examples", "prefixItems")
+                }
+
+                handle_discriminators(obj)
+                ensure_array_items(obj)
+
+                # Check for oneOf and anyOf to handle nested schemas
+                for key in ["oneOf", "anyOf"]:
+                    if key in obj:
+                        for sub in obj[key]:
+                            handle_discriminators(sub)
+                            ensure_array_items(sub)
+
+                return obj
+            elif isinstance(obj, list):
+                return [strip_unsupported(i) for i in obj]
+            return obj
+
+        if "components" in spec and "schemas" in spec["components"]:
+            for schema_name, schema in spec["components"]["schemas"].items():
+                handle_properties(schema)
+
+        return strip_unsupported(copy.deepcopy(spec))
+
     #############################
     # API Endpoints definitions #
     #############################
 
+    @app.get("/openapi-3.0.json")
+    def openapi_30():
+        spec = app.openapi()
+        downgraded = downgrade_openapi31_to_30(spec)
+        downgraded["openapi"] = "3.0.3"
+        return JSONResponse(downgraded)
+
     # Favicon
     @app.get("/favicon.ico", include_in_schema=False)
     async def favicon():
@@ -443,7 +522,8 @@ def create_app():  # noqa: C901
         orchestrator: Annotated[BaseOrchestrator, Depends(get_async_orchestrator)],
         task_id: str,
         wait: Annotated[
-            float, Query(help="Number of seconds to wait for a completed status.")
+            float,
+            Query(description="Number of seconds to wait for a completed status."),
         ] = 0.0,
     ):
         try:
@@ -525,7 +605,7 @@ def create_app():  # noqa: C901
     # Task result
     @app.get(
         "/v1/result/{task_id}",
-        response_model=ConvertDocumentResponse,
+        response_model=ConvertDocumentResponse | PresignedUrlConvertDocumentResponse,
         responses={
             200: {
                 "content": {"application/zip": {}},

docling_serve/datamodel/requests.py

@@ -1,12 +1,21 @@
 import enum
 from typing import Annotated, Literal
 
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, model_validator
+from pydantic_core import PydanticCustomError
+from typing_extensions import Self
 
 from docling_jobkit.datamodel.http_inputs import FileSource, HttpSource
-from docling_jobkit.datamodel.task_targets import InBodyTarget, TaskTarget, ZipTarget
+from docling_jobkit.datamodel.s3_coords import S3Coordinates
+from docling_jobkit.datamodel.task_targets import (
+    InBodyTarget,
+    S3Target,
+    TaskTarget,
+    ZipTarget,
+)
 
 from docling_serve.datamodel.convert import ConvertDocumentsRequestOptions
+from docling_serve.settings import AsyncEngine, docling_serve_settings
 
 ## Sources
 
@@ -19,6 +28,10 @@ class HttpSourceRequest(HttpSource):
     kind: Literal["http"] = "http"
 
 
+class S3SourceRequest(S3Coordinates):
+    kind: Literal["s3"] = "s3"
+
+
 ## Multipart targets
 class TargetName(str, enum.Enum):
     INBODY = InBodyTarget().kind
@@ -27,7 +40,7 @@ class TargetName(str, enum.Enum):
 
 ## Aliases
 SourceRequestItem = Annotated[
-    FileSourceRequest | HttpSourceRequest, Field(discriminator="kind")
+    FileSourceRequest | HttpSourceRequest | S3SourceRequest, Field(discriminator="kind")
 ]
 
 
@@ -36,3 +49,24 @@ class ConvertDocumentsRequest(BaseModel):
     options: ConvertDocumentsRequestOptions = ConvertDocumentsRequestOptions()
     sources: list[SourceRequestItem]
     target: TaskTarget = InBodyTarget()
+
+    @model_validator(mode="after")
+    def validate_s3_source_and_target(self) -> Self:
+        for source in self.sources:
+            if isinstance(source, S3SourceRequest):
+                if docling_serve_settings.eng_kind != AsyncEngine.KFP:
+                    raise PydanticCustomError(
+                        "error source", 'source kind "s3" requires engine kind "KFP"'
+                    )
+                if self.target.kind != "s3":
+                    raise PydanticCustomError(
+                        "error source", 'source kind "s3" requires target kind "s3"'
+                    )
+        if isinstance(self.target, S3Target):
+            for source in self.sources:
+                if isinstance(source, S3SourceRequest):
+                    return self
+            raise PydanticCustomError(
+                "error target", 'target kind "s3" requires source kind "s3"'
+            )
+        return self

docling_serve/datamodel/responses.py

@@ -35,6 +35,11 @@ class ConvertDocumentResponse(BaseModel):
     timings: dict[str, ProfilingItem] = {}
 
 
+class PresignedUrlConvertDocumentResponse(BaseModel):
+    status: ConversionStatus
+    processing_time: float
+
+
 class ConvertDocumentErrorResponse(BaseModel):
     status: ConversionStatus
 

docling_serve/response_preparation.py

@@ -7,6 +7,7 @@ from collections.abc import Iterable
 from pathlib import Path
 from typing import Union
 
+import httpx
 from fastapi import BackgroundTasks, HTTPException
 from fastapi.responses import FileResponse
 
@@ -15,12 +16,16 @@ from docling.datamodel.document import ConversionResult, ConversionStatus
 from docling_core.types.doc import ImageRefMode
 from docling_jobkit.datamodel.convert import ConvertDocumentsOptions
 from docling_jobkit.datamodel.task import Task
-from docling_jobkit.datamodel.task_targets import InBodyTarget, TaskTarget
+from docling_jobkit.datamodel.task_targets import InBodyTarget, PutTarget, TaskTarget
 from docling_jobkit.orchestrators.base_orchestrator import (
     BaseOrchestrator,
 )
 
-from docling_serve.datamodel.responses import ConvertDocumentResponse, DocumentResponse
+from docling_serve.datamodel.responses import (
+    ConvertDocumentResponse,
+    DocumentResponse,
+    PresignedUrlConvertDocumentResponse,
+)
 from docling_serve.settings import docling_serve_settings
 from docling_serve.storage import get_scratch
 
@@ -79,11 +84,17 @@ def _export_documents_as_files(
     export_doctags: bool,
     image_export_mode: ImageRefMode,
     md_page_break_placeholder: str,
-):
+) -> ConversionStatus:
     success_count = 0
     failure_count = 0
 
+    # Default failure in case results is empty
+    conv_result = ConversionStatus.FAILURE
+
+    artifacts_dir = Path("artifacts/")  # will be relative to the fname
+
     for conv_res in conv_results:
+        conv_result = conv_res.status
         if conv_res.status == ConversionStatus.SUCCESS:
             success_count += 1
             doc_filename = conv_res.input.file.stem
@@ -93,7 +104,9 @@ def _export_documents_as_files(
                 fname = output_dir / f"{doc_filename}.json"
                 _log.info(f"writing JSON output to {fname}")
                 conv_res.document.save_as_json(
-                    filename=fname, image_mode=image_export_mode
+                    filename=fname,
+                    image_mode=image_export_mode,
+                    artifacts_dir=artifacts_dir,
                 )
 
             # Export HTML format:
@@ -101,7 +114,9 @@ def _export_documents_as_files(
                 fname = output_dir / f"{doc_filename}.html"
                 _log.info(f"writing HTML output to {fname}")
                 conv_res.document.save_as_html(
-                    filename=fname, image_mode=image_export_mode
+                    filename=fname,
+                    image_mode=image_export_mode,
+                    artifacts_dir=artifacts_dir,
                 )
 
             # Export Text format:
@@ -120,6 +135,7 @@ def _export_documents_as_files(
                 _log.info(f"writing Markdown output to {fname}")
                 conv_res.document.save_as_markdown(
                     filename=fname,
+                    artifacts_dir=artifacts_dir,
                     image_mode=image_export_mode,
                     page_break_placeholder=md_page_break_placeholder or None,
                 )
@@ -138,6 +154,7 @@ def _export_documents_as_files(
         f"Processed {success_count + failure_count} docs, "
         f"of which {failure_count} failed"
     )
+    return conv_result
 
 
 def process_results(
@@ -145,7 +162,7 @@ def process_results(
     target: TaskTarget,
     conv_results: Iterable[ConversionResult],
     work_dir: Path,
-) -> Union[ConvertDocumentResponse, FileResponse]:
+) -> Union[ConvertDocumentResponse, FileResponse, PresignedUrlConvertDocumentResponse]:
     # Let's start by processing the documents
     try:
         start_time = time.monotonic()
@@ -169,7 +186,9 @@ def process_results(
         )
 
     # We have some results, let's prepare the response
-    response: Union[FileResponse, ConvertDocumentResponse]
+    response: Union[
+        FileResponse, ConvertDocumentResponse, PresignedUrlConvertDocumentResponse
+    ]
 
     # Booleans to know what to export
     export_json = OutputFormat.JSON in conversion_options.to_formats
@@ -209,7 +228,7 @@ def process_results(
         os.getpid()
 
         # Export the documents
-        _export_documents_as_files(
+        conv_result = _export_documents_as_files(
             conv_results=conv_results,
             output_dir=output_dir,
             export_json=export_json,
@@ -236,9 +255,24 @@ def process_results(
         # Output directory
         # background_tasks.add_task(shutil.rmtree, work_dir, ignore_errors=True)
 
-        response = FileResponse(
-            file_path, filename=file_path.name, media_type="application/zip"
-        )
+        if isinstance(target, PutTarget):
+            try:
+                with open(file_path, "rb") as file_data:
+                    r = httpx.put(str(target.url), files={"file": file_data})
+                    r.raise_for_status()
+                response = PresignedUrlConvertDocumentResponse(
+                    status=conv_result,
+                    processing_time=processing_time,
+                )
+            except Exception as exc:
+                _log.error("An error occour while uploading zip to s3", exc_info=exc)
+                raise HTTPException(
+                    status_code=500, detail="An error occour while uploading zip to s3."
+                )
+        else:
+            response = FileResponse(
+                file_path, filename=file_path.name, media_type="application/zip"
+            )
 
     return response
 

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docling-serve"
-version = "1.0.1"  # DO NOT EDIT, updated automatically
+version = "1.1.0"  # DO NOT EDIT, updated automatically
 description = "Running Docling as a service"
 license = {text = "MIT"}
 authors = [
@@ -35,8 +35,8 @@ classifiers = [
 requires-python = ">=3.10"
 dependencies = [
     "docling~=2.38",
-    "docling-core>=2.32.0",
-    "docling-jobkit[kfp,vlm]~=1.1",
+    "docling-core>=2.44.1",
+    "docling-jobkit[kfp,vlm]~=1.2",
     "fastapi[standard]~=0.115",
     "httpx~=0.28",
     "pydantic~=2.10",
@@ -46,6 +46,7 @@ dependencies = [
     "uvicorn[standard]>=0.29.0,<1.0.0",
     "websockets~=14.0",
     "scalar-fastapi>=1.0.3",
+    "docling-mcp>=1.0.0",
 ]
 
 [project.optional-dependencies]
@@ -128,7 +129,7 @@ torchvision = [
   { index = "pytorch-cu126", group = "cu126" },
   { index = "pytorch-cu128", group = "cu128" },
 ]
-# docling-jobkit = { git = "https://github.com/docling-project/docling-jobkit/", rev = "refactor" }
+# docling-jobkit = { git = "https://github.com/docling-project/docling-jobkit/", rev = "main" }
 # docling-jobkit = { path = "../docling-jobkit", editable = true }
 
 [[tool.uv.index]]

tests/test_fastapi_endpoints.py

@@ -1,6 +1,8 @@
 import asyncio
+import io
 import json
 import os
+import zipfile
 
 import pytest
 import pytest_asyncio
@@ -8,6 +10,8 @@ from asgi_lifespan import LifespanManager
 from httpx import ASGITransport, AsyncClient
 from pytest_check import check
 
+from docling_core.types.doc import DoclingDocument, PictureItem
+
 from docling_serve.app import create_app
 
 
@@ -153,3 +157,37 @@ async def test_convert_file(client: AsyncClient):
             data["document"]["doctags_content"],
             msg=f"DocTags document should contain '<doctag><page_header>'. Received: {safe_slice(data['document']['doctags_content'])}",
         )
+
+
+@pytest.mark.asyncio
+async def test_referenced_artifacts(client: AsyncClient):
+    """Test that paths in the zip file are relative to the zip file root."""
+
+    endpoint = "/v1/convert/file"
+    options = {
+        "to_formats": ["json"],
+        "image_export_mode": "referenced",
+        "target_type": "zip",
+        "ocr": False,
+    }
+
+    current_dir = os.path.dirname(__file__)
+    file_path = os.path.join(current_dir, "2206.01062v1.pdf")
+
+    files = {
+        "files": ("2206.01062v1.pdf", open(file_path, "rb"), "application/pdf"),
+    }
+
+    response = await client.post(endpoint, files=files, data=options)
+    assert response.status_code == 200, "Response should be 200 OK"
+
+    with zipfile.ZipFile(io.BytesIO(response.content)) as zip_file:
+        namelist = zip_file.namelist()
+        for file in namelist:
+            if file.endswith(".json"):
+                doc = DoclingDocument.model_validate(json.loads(zip_file.read(file)))
+                for item, _level in doc.iterate_items():
+                    if isinstance(item, PictureItem):
+                        assert item.image is not None
+                        print(f"{item.image.uri}=")
+                        assert str(item.image.uri) in namelist