feat: make page_range field public (#11751)

BEGIN_COMMIT_OVERRIDE
feat: make page_range field public
feat: Added SummaryOptions to ProcessOptions for the Summarizer
processor
feat: Added `ListDocuments()` method for Document AI Workbench training
documents
END_COMMIT_OVERRIDE

- [ ] Regenerate this pull request now.

fix: `OcrConfig.compute_style_info` is deprecated. Use
`PremiumFeatures.compute_style_info` instead.

PiperOrigin-RevId: 568225060

Source-Link:
googleapis/googleapis@3cd21b1

Source-Link:
googleapis/googleapis-gen@087cf52
Copy-Tag:
eyJwIjoicGFja2FnZXMvZ29vZ2xlLWNsb3VkLWRvY3VtZW50YWkvLk93bEJvdC55YW1sIiwiaCI6IjA4N2NmNTIzZDFhMGQ5ZDUzZmU3ODMzYmM0ZDdiZjM2OWY4MmQzMzMifQ==

BEGIN_NESTED_COMMIT
feat: Added `SummaryOptions` to `ProcessOptions` for the Summarizer
processor
feat: Added `ListDocuments()` method for Document AI Workbench training
documents

PiperOrigin-RevId: 567684912

Source-Link:
googleapis/googleapis@cac736f

Source-Link:
googleapis/googleapis-gen@cc556f3
Copy-Tag:
eyJwIjoicGFja2FnZXMvZ29vZ2xlLWNsb3VkLWRvY3VtZW50YWkvLk93bEJvdC55YW1sIiwiaCI6ImNjNTU2ZjM5ZmJkZWQzMjI1MTNlNWJlMTZhNzRmYWJlYmMzMTQ2MGEifQ==
END_NESTED_COMMIT

---------

Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>
Co-authored-by: omair <omairn@google.com>

Author: gcf-owl-bot[bot]

packages/google-cloud-documentai/CONTRIBUTING.rst

@@ -143,12 +143,12 @@ Running System Tests
    $ nox -s system
 
    # Run a single system test
-   $ nox -s system- -- -k <name of test>
+   $ nox -s system-3.11 -- -k <name of test>
 
 
   .. note::
 
-      System tests are only configured to run under Python.
+      System tests are only configured to run under Python 3.8, 3.9, 3.10 and 3.11.
       For expediency, we do not run them in older versions of Python 3.
 
   This alone will not run the tests. You'll need to change some local

packages/google-cloud-documentai/docs/documentai_v1beta3/document_service.rst

@@ -4,3 +4,7 @@ DocumentService
 .. automodule:: google.cloud.documentai_v1beta3.services.document_service
     :members:
     :inherited-members:
+
+.. automodule:: google.cloud.documentai_v1beta3.services.document_service.pagers
+    :members:
+    :inherited-members:

packages/google-cloud-documentai/google/cloud/documentai/gapic_version.py

@@ -13,4 +13,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-__version__ = "2.19.0"  # {x-release-please-version}
+__version__ = "0.0.0"  # {x-release-please-version}

packages/google-cloud-documentai/google/cloud/documentai_v1/gapic_version.py

@@ -13,4 +13,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-__version__ = "2.19.0"  # {x-release-please-version}
+__version__ = "0.0.0"  # {x-release-please-version}

packages/google-cloud-documentai/google/cloud/documentai_v1/types/document_io.py

@@ -48,7 +48,7 @@ class RawDocument(proto.Message):
             The display name of the document, it supports all Unicode
             characters except the following: ``*``, ``?``, ``[``, ``]``,
             ``%``, ``{``, ``}``,\ ``'``, ``\"``, ``,`` ``~``, ``=`` and
-            ``:`` are reserved. If not specified, a default ID will be
+            ``:`` are reserved. If not specified, a default ID is
             generated.
     """
 
@@ -260,8 +260,16 @@ class OcrConfig(proto.Message):
             Includes symbol level OCR information if set
             to true.
         compute_style_info (bool):
-            Turn on font id model and returns font style information.
-            Use PremiumFeatures.compute_style_info instead.
+            Turn on font identification model and return font style
+            information. Deprecated, use
+            [PremiumFeatures.compute_style_info][google.cloud.documentai.v1.OcrConfig.PremiumFeatures.compute_style_info]
+            instead.
+        disable_character_boxes_detection (bool):
+            Turn off character box detector in OCR
+            engine. Character box detection is enabled by
+            default in OCR 2.0+ processors.
+        premium_features (google.cloud.documentai_v1.types.OcrConfig.PremiumFeatures):
+            Configurations for premium OCR features.
     """
 
     class Hints(proto.Message):
@@ -285,6 +293,34 @@ class Hints(proto.Message):
             number=1,
         )
 
+    class PremiumFeatures(proto.Message):
+        r"""Configurations for premium OCR features.
+
+        Attributes:
+            enable_selection_mark_detection (bool):
+                Turn on selection mark detector in OCR
+                engine. Only available in OCR 2.0+ processors.
+            compute_style_info (bool):
+                Turn on font identification model and return
+                font style information.
+            enable_math_ocr (bool):
+                Turn on the model that can extract LaTeX math
+                formulas.
+        """
+
+        enable_selection_mark_detection: bool = proto.Field(
+            proto.BOOL,
+            number=3,
+        )
+        compute_style_info: bool = proto.Field(
+            proto.BOOL,
+            number=4,
+        )
+        enable_math_ocr: bool = proto.Field(
+            proto.BOOL,
+            number=5,
+        )
+
     hints: Hints = proto.Field(
         proto.MESSAGE,
         number=2,
@@ -310,6 +346,15 @@ class Hints(proto.Message):
         proto.BOOL,
         number=8,
     )
+    disable_character_boxes_detection: bool = proto.Field(
+        proto.BOOL,
+        number=10,
+    )
+    premium_features: PremiumFeatures = proto.Field(
+        proto.MESSAGE,
+        number=11,
+        message=PremiumFeatures,
+    )
 
 
 __all__ = tuple(sorted(__protobuf__.manifest))

packages/google-cloud-documentai/google/cloud/documentai_v1/types/document_processor_service.py

@@ -90,12 +90,63 @@
 class ProcessOptions(proto.Message):
     r"""Options for Process API
 
+    This message has `oneof`_ fields (mutually exclusive fields).
+    For each oneof, at most one member field can be set at the same time.
+    Setting any member of the oneof automatically clears all other
+    members.
+
+    .. _oneof: https://proto-plus-python.readthedocs.io/en/stable/fields.html#oneofs-mutually-exclusive-fields
+
     Attributes:
+        individual_page_selector (google.cloud.documentai_v1.types.ProcessOptions.IndividualPageSelector):
+            Which pages to process (1-indexed).
+
+            This field is a member of `oneof`_ ``page_range``.
+        from_start (int):
+            Only process certain pages from the start.
+            Process all if the document has fewer pages.
+
+            This field is a member of `oneof`_ ``page_range``.
+        from_end (int):
+            Only process certain pages from the end, same
+            as above.
+
+            This field is a member of `oneof`_ ``page_range``.
         ocr_config (google.cloud.documentai_v1.types.OcrConfig):
             Only applicable to ``OCR_PROCESSOR``. Returns error if set
             on other processor types.
     """
 
+    class IndividualPageSelector(proto.Message):
+        r"""A list of individual page numbers.
+
+        Attributes:
+            pages (MutableSequence[int]):
+                Optional. Indices of the pages (starting from
+                1).
+        """
+
+        pages: MutableSequence[int] = proto.RepeatedField(
+            proto.INT32,
+            number=1,
+        )
+
+    individual_page_selector: IndividualPageSelector = proto.Field(
+        proto.MESSAGE,
+        number=5,
+        oneof="page_range",
+        message=IndividualPageSelector,
+    )
+    from_start: int = proto.Field(
+        proto.INT32,
+        number=6,
+        oneof="page_range",
+    )
+    from_end: int = proto.Field(
+        proto.INT32,
+        number=7,
+        oneof="page_range",
+    )
     ocr_config: document_io.OcrConfig = proto.Field(
         proto.MESSAGE,
         number=1,

packages/google-cloud-documentai/google/cloud/documentai_v1beta2/gapic_version.py

@@ -13,4 +13,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-__version__ = "2.19.0"  # {x-release-please-version}
+__version__ = "0.0.0"  # {x-release-please-version}

packages/google-cloud-documentai/google/cloud/documentai_v1beta3/__init__.py

@@ -90,27 +90,37 @@
     UndeployProcessorVersionRequest,
     UndeployProcessorVersionResponse,
 )
-from .types.document_schema import DocumentSchema, EntityTypeMetadata, PropertyMetadata
+from .types.document_schema import (
+    DocumentSchema,
+    EntityTypeMetadata,
+    FieldExtractionMetadata,
+    PropertyMetadata,
+    SummaryOptions,
+)
 from .types.document_service import (
     BatchDeleteDocumentsMetadata,
     BatchDeleteDocumentsRequest,
     BatchDeleteDocumentsResponse,
     DatasetSplitType,
+    DocumentLabelingState,
+    DocumentMetadata,
     DocumentPageRange,
     GetDatasetSchemaRequest,
     GetDocumentRequest,
     GetDocumentResponse,
     ImportDocumentsMetadata,
     ImportDocumentsRequest,
     ImportDocumentsResponse,
+    ListDocumentsRequest,
+    ListDocumentsResponse,
     UpdateDatasetOperationMetadata,
     UpdateDatasetRequest,
     UpdateDatasetSchemaRequest,
 )
 from .types.evaluation import Evaluation, EvaluationReference
 from .types.geometry import BoundingPoly, NormalizedVertex, Vertex
 from .types.operation_metadata import CommonOperationMetadata
-from .types.processor import Processor, ProcessorVersion
+from .types.processor import Processor, ProcessorVersion, ProcessorVersionAlias
 from .types.processor_type import ProcessorType
 
 __all__ = (
@@ -143,6 +153,8 @@
     "DisableProcessorResponse",
     "Document",
     "DocumentId",
+    "DocumentLabelingState",
+    "DocumentMetadata",
     "DocumentOutputConfig",
     "DocumentPageRange",
     "DocumentProcessorServiceClient",
@@ -159,6 +171,7 @@
     "EvaluationReference",
     "FetchProcessorTypesRequest",
     "FetchProcessorTypesResponse",
+    "FieldExtractionMetadata",
     "GcsDocument",
     "GcsDocuments",
     "GcsPrefix",
@@ -176,6 +189,8 @@
     "ImportProcessorVersionMetadata",
     "ImportProcessorVersionRequest",
     "ImportProcessorVersionResponse",
+    "ListDocumentsRequest",
+    "ListDocumentsResponse",
     "ListEvaluationsRequest",
     "ListEvaluationsResponse",
     "ListProcessorTypesRequest",
@@ -192,6 +207,7 @@
     "Processor",
     "ProcessorType",
     "ProcessorVersion",
+    "ProcessorVersionAlias",
     "PropertyMetadata",
     "RawDocument",
     "ReviewDocumentOperationMetadata",
@@ -201,6 +217,7 @@
     "SetDefaultProcessorVersionMetadata",
     "SetDefaultProcessorVersionRequest",
     "SetDefaultProcessorVersionResponse",
+    "SummaryOptions",
     "TrainProcessorVersionMetadata",
     "TrainProcessorVersionRequest",
     "TrainProcessorVersionResponse",

packages/google-cloud-documentai/google/cloud/documentai_v1beta3/gapic_metadata.json

@@ -394,6 +394,11 @@
                 "import_documents"
               ]
             },
+            "ListDocuments": {
+              "methods": [
+                "list_documents"
+              ]
+            },
             "UpdateDataset": {
               "methods": [
                 "update_dataset"
@@ -429,6 +434,11 @@
                 "import_documents"
               ]
             },
+            "ListDocuments": {
+              "methods": [
+                "list_documents"
+              ]
+            },
             "UpdateDataset": {
               "methods": [
                 "update_dataset"
@@ -464,6 +474,11 @@
                 "import_documents"
               ]
             },
+            "ListDocuments": {
+              "methods": [
+                "list_documents"
+              ]
+            },
             "UpdateDataset": {
               "methods": [
                 "update_dataset"

packages/google-cloud-documentai/google/cloud/documentai_v1beta3/gapic_version.py

@@ -13,4 +13,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-__version__ = "2.19.0"  # {x-release-please-version}
+__version__ = "0.0.0"  # {x-release-please-version}