feat: added font_family to document.proto (#404)

* feat: added font_family to document.proto
feat: added ImageQualityScores message to document.proto
feat: added PropertyMetadata and EntityTypeMetadata to document_schema.proto

PiperOrigin-RevId: 486975621

Source-Link: googleapis/googleapis@398c9f9

Source-Link: googleapis/googleapis-gen@7cd1f5f
Copy-Tag: eyJwIjoiLmdpdGh1Yi8uT3dsQm90LnlhbWwiLCJoIjoiN2NkMWY1ZjRlNDM1Nzc3Y2I4MjRhZjI2OGRjOGQzNzEzNDYxM2U2YSJ9

* 🦉 Updates from OwlBot post-processor

See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md

* Update constraints-3.7.txt

Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>
Co-authored-by: Holt Skinner <13262395+holtskinner@users.noreply.github.com>

google/cloud/documentai_v1/services/document_processor_service/async_client.py

@@ -39,7 +39,11 @@
 from google.protobuf import timestamp_pb2  # type: ignore
 
 from google.cloud.documentai_v1.services.document_processor_service import pagers
-from google.cloud.documentai_v1.types import document, document_processor_service
+from google.cloud.documentai_v1.types import (
+    document,
+    document_processor_service,
+    document_schema,
+)
 from google.cloud.documentai_v1.types import processor
 from google.cloud.documentai_v1.types import processor as gcd_processor
 from google.cloud.documentai_v1.types import processor_type

google/cloud/documentai_v1/services/document_processor_service/client.py

@@ -42,7 +42,11 @@
 from google.protobuf import timestamp_pb2  # type: ignore
 
 from google.cloud.documentai_v1.services.document_processor_service import pagers
-from google.cloud.documentai_v1.types import document, document_processor_service
+from google.cloud.documentai_v1.types import (
+    document,
+    document_processor_service,
+    document_schema,
+)
 from google.cloud.documentai_v1.types import processor
 from google.cloud.documentai_v1.types import processor as gcd_processor
 from google.cloud.documentai_v1.types import processor_type

google/cloud/documentai_v1/types/barcode.py

@@ -28,25 +28,41 @@ class Barcode(proto.Message):
 
     Attributes:
         format_ (str):
-            Format of a barcode. The supported formats are: CODE_128:
-            Code 128 type. CODE_39: Code 39 type. CODE_93: Code 93 type.
-            CODABAR: Codabar type. DATA_MATRIX: 2D Data Matrix type.
-            ITF: ITF type. EAN_13: EAN-13 type. EAN_8: EAN-8 type.
-            QR_CODE: 2D QR code type. UPC_A: UPC-A type. UPC_E: UPC-E
-            type. PDF417: PDF417 type. AZTEC: 2D Aztec code type.
-            DATABAR: GS1 DataBar code type.
+            Format of a barcode. The supported formats are:
+
+            -  ``CODE_128``: Code 128 type.
+            -  ``CODE_39``: Code 39 type.
+            -  ``CODE_93``: Code 93 type.
+            -  ``CODABAR``: Codabar type.
+            -  ``DATA_MATRIX``: 2D Data Matrix type.
+            -  ``ITF``: ITF type.
+            -  ``EAN_13``: EAN-13 type.
+            -  ``EAN_8``: EAN-8 type.
+            -  ``QR_CODE``: 2D QR code type.
+            -  ``UPC_A``: UPC-A type.
+            -  ``UPC_E``: UPC-E type.
+            -  ``PDF417``: PDF417 type.
+            -  ``AZTEC``: 2D Aztec code type.
+            -  ``DATABAR``: GS1 DataBar code type.
         value_format (str):
             Value format describes the format of the value that a
-            barcode encodes. The supported formats are: CONTACT_INFO:
-            Contact information. EMAIL: Email address. ISBN: ISBN
-            identifier. PHONE: Phone number. PRODUCT: Product. SMS: SMS
-            message. TEXT: Text string. URL: URL address. WIFI: Wifi
-            information. GEO: Geo-localization. CALENDAR_EVENT: Calendar
-            event. DRIVER_LICENSE: Driver's license.
+            barcode encodes. The supported formats are:
+
+            -  ``CONTACT_INFO``: Contact information.
+            -  ``EMAIL``: Email address.
+            -  ``ISBN``: ISBN identifier.
+            -  ``PHONE``: Phone number.
+            -  ``PRODUCT``: Product.
+            -  ``SMS``: SMS message.
+            -  ``TEXT``: Text string.
+            -  ``URL``: URL address.
+            -  ``WIFI``: Wifi information.
+            -  ``GEO``: Geo-localization.
+            -  ``CALENDAR_EVENT``: Calendar event.
+            -  ``DRIVER_LICENSE``: Driver's license.
         raw_value (str):
-            Raw value encoded in the barcode.
-            For example,
-            'MEBKM:TITLE:Google;URL:https://www.google.com;;'.
+            Raw value encoded in the barcode. For example:
+            ``'MEBKM:TITLE:Google;URL:https://www.google.com;;'``.
     """
 
     format_ = proto.Field(

google/cloud/documentai_v1/types/document.py

@@ -85,9 +85,10 @@ class Document(proto.Message):
             [Document.entities][google.cloud.documentai.v1.Document.entities].
         text_changes (Sequence[google.cloud.documentai_v1.types.Document.TextChange]):
             Placeholder. A list of text corrections made to
-            [Document.text]. This is usually used for annotating
-            corrections to OCR mistakes. Text changes for a given
-            revision may not overlap with each other.
+            [Document.text][google.cloud.documentai.v1.Document.text].
+            This is usually used for annotating corrections to OCR
+            mistakes. Text changes for a given revision may not overlap
+            with each other.
         shard_info (google.cloud.documentai_v1.types.Document.ShardInfo):
             Information about the sharding if this
             document is sharded part of a larger document.
@@ -153,6 +154,9 @@ class Style(proto.Message):
                 https://www.w3schools.com/cssref/pr_text_text-decoration.asp
             font_size (google.cloud.documentai_v1.types.Document.Style.FontSize):
                 Font size.
+            font_family (str):
+                Font family such as ``Arial``, ``Times New Roman``.
+                https://www.w3schools.com/cssref/pr_font_font-family.asp
         """
 
         class FontSize(proto.Message):
@@ -207,6 +211,10 @@ class FontSize(proto.Message):
             number=7,
             message="Document.Style.FontSize",
         )
+        font_family = proto.Field(
+            proto.STRING,
+            number=8,
+        )
 
     class Page(proto.Message):
         r"""A page in a [Document][google.cloud.documentai.v1.Document].
@@ -266,6 +274,8 @@ class Page(proto.Message):
                 page.
             detected_barcodes (Sequence[google.cloud.documentai_v1.types.Document.Page.DetectedBarcode]):
                 A list of detected barcodes.
+            image_quality_scores (google.cloud.documentai_v1.types.Document.Page.ImageQualityScores):
+                Image Quality Scores.
             provenance (google.cloud.documentai_v1.types.Document.Provenance):
                 The history of this page.
         """
@@ -374,7 +384,7 @@ class Layout(proto.Message):
                     [Layout][google.cloud.documentai.v1.Document.Page.Layout]
                     within context of the object this layout is for. e.g.
                     confidence can be for a single token, a table, a visual
-                    element, etc. depending on context. Range [0, 1].
+                    element, etc. depending on context. Range ``[0, 1]``.
                 bounding_poly (google.cloud.documentai_v1.types.BoundingPoly):
                     The bounding polygon for the
                     [Layout][google.cloud.documentai.v1.Document.Page.Layout].
@@ -520,7 +530,7 @@ class Token(proto.Message):
                     A list of detected languages together with
                     confidence.
                 provenance (google.cloud.documentai_v1.types.Document.Provenance):
-                    The  history of this annotation.
+                    The history of this annotation.
             """
 
             class DetectedBreak(proto.Message):
@@ -636,6 +646,8 @@ class Table(proto.Message):
                 detected_languages (Sequence[google.cloud.documentai_v1.types.Document.Page.DetectedLanguage]):
                     A list of detected languages together with
                     confidence.
+                provenance (google.cloud.documentai_v1.types.Document.Provenance):
+                    The history of this table.
             """
 
             class TableRow(proto.Message):
@@ -708,6 +720,11 @@ class TableCell(proto.Message):
                 number=4,
                 message="Document.Page.DetectedLanguage",
             )
+            provenance = proto.Field(
+                proto.MESSAGE,
+                number=5,
+                message="Document.Provenance",
+            )
 
         class FormField(proto.Message):
             r"""A form field detected on the page.
@@ -818,11 +835,11 @@ class DetectedLanguage(proto.Message):
 
             Attributes:
                 language_code (str):
-                    The BCP-47 language code, such as "en-US" or "sr-Latn". For
-                    more information, see
+                    The BCP-47 language code, such as ``en-US`` or ``sr-Latn``.
+                    For more information, see
                     https://www.unicode.org/reports/tr35/#Unicode_locale_identifier.
                 confidence (float):
-                    Confidence of detected language. Range [0, 1].
+                    Confidence of detected language. Range ``[0, 1]``.
             """
 
             language_code = proto.Field(
@@ -834,6 +851,56 @@ class DetectedLanguage(proto.Message):
                 number=2,
             )
 
+        class ImageQualityScores(proto.Message):
+            r"""Image Quality Scores for the page image
+
+            Attributes:
+                quality_score (float):
+                    The overall quality score. Range ``[0, 1]`` where 1 is
+                    perfect quality.
+                detected_defects (Sequence[google.cloud.documentai_v1.types.Document.Page.ImageQualityScores.DetectedDefect]):
+                    A list of detected defects.
+            """
+
+            class DetectedDefect(proto.Message):
+                r"""Image Quality Defects
+
+                Attributes:
+                    type_ (str):
+                        Name of the defect type. Supported values are:
+
+                        -  ``quality/defect_blurry``
+                        -  ``quality/defect_noisy``
+                        -  ``quality/defect_dark``
+                        -  ``quality/defect_faint``
+                        -  ``quality/defect_text_too_small``
+                        -  ``quality/defect_document_cutoff``
+                        -  ``quality/defect_text_cutoff``
+                        -  ``quality/defect_glare``
+                    confidence (float):
+                        Confidence of detected defect. Range ``[0, 1]`` where 1
+                        indicates strong confidence of that the defect exists.
+                """
+
+                type_ = proto.Field(
+                    proto.STRING,
+                    number=1,
+                )
+                confidence = proto.Field(
+                    proto.FLOAT,
+                    number=2,
+                )
+
+            quality_score = proto.Field(
+                proto.FLOAT,
+                number=1,
+            )
+            detected_defects = proto.RepeatedField(
+                proto.MESSAGE,
+                number=2,
+                message="Document.Page.ImageQualityScores.DetectedDefect",
+            )
+
         page_number = proto.Field(
             proto.INT32,
             number=1,
@@ -908,6 +975,11 @@ class DetectedLanguage(proto.Message):
             number=15,
             message="Document.Page.DetectedBarcode",
         )
+        image_quality_scores = proto.Field(
+            proto.MESSAGE,
+            number=17,
+            message="Document.Page.ImageQualityScores",
+        )
         provenance = proto.Field(
             proto.MESSAGE,
             number=16,
@@ -927,14 +999,13 @@ class Entity(proto.Message):
             type_ (str):
                 Required. Entity type from a schema e.g. ``Address``.
             mention_text (str):
-                Optional. Text value in the document e.g.
-                ``1600 Amphitheatre Pkwy``. If the entity is not present in
-                the document, this field will be empty.
+                Optional. Text value of the entity e.g.
+                ``1600 Amphitheatre Pkwy``.
             mention_id (str):
                 Optional. Deprecated. Use ``id`` field instead.
             confidence (float):
-                Optional. Confidence of detected Schema entity. Range [0,
-                1].
+                Optional. Confidence of detected Schema entity. Range
+                ``[0, 1]``.
             page_anchor (google.cloud.documentai_v1.types.Document.PageAnchor):
                 Optional. Represents the provenance of this
                 entity wrt. the location on the page where it
@@ -1230,7 +1301,7 @@ class PageRef(proto.Message):
                     a layout element on the page.
                 confidence (float):
                     Optional. Confidence of detected page element, if
-                    applicable. Range [0, 1].
+                    applicable. Range ``[0, 1]``.
             """
 
             class LayoutType(proto.Enum):

google/cloud/documentai_v1/types/document_io.py

@@ -13,6 +13,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
+from google.protobuf import field_mask_pb2  # type: ignore
 import proto  # type: ignore
 
 __protobuf__ = proto.module(
@@ -36,7 +37,8 @@ class RawDocument(proto.Message):
             Inline document content.
         mime_type (str):
             An IANA MIME type (RFC6838) indicating the nature and format
-            of the [content].
+            of the
+            [content][google.cloud.documentai.v1.RawDocument.content].
     """
 
     content = proto.Field(
@@ -113,7 +115,7 @@ class BatchDocumentsInputConfig(proto.Message):
     Attributes:
         gcs_prefix (google.cloud.documentai_v1.types.GcsPrefix):
             The set of documents that match the specified Cloud Storage
-            [gcs_prefix].
+            ``gcs_prefix``.
 
             This field is a member of `oneof`_ ``source``.
         gcs_documents (google.cloud.documentai_v1.types.GcsDocuments):
@@ -159,12 +161,22 @@ class GcsOutputConfig(proto.Message):
             gcs_uri (str):
                 The Cloud Storage uri (a directory) of the
                 output.
+            field_mask (google.protobuf.field_mask_pb2.FieldMask):
+                Specifies which fields to include in the output documents.
+                Only supports top level document and pages field so it must
+                be in the form of ``{document_field_name}`` or
+                ``pages.{page_field_name}``.
         """
 
         gcs_uri = proto.Field(
             proto.STRING,
             number=1,
         )
+        field_mask = proto.Field(
+            proto.MESSAGE,
+            number=2,
+            message=field_mask_pb2.FieldMask,
+        )
 
     gcs_output_config = proto.Field(
         proto.MESSAGE,

google/cloud/documentai_v1/types/document_processor_service.py

@@ -106,8 +106,10 @@ class ProcessRequest(proto.Message):
             Whether Human Review feature should be
             skipped for this request. Default to false.
         field_mask (google.protobuf.field_mask_pb2.FieldMask):
-            Specifies which fields to include in
-            ProcessResponse's document.
+            Specifies which fields to include in ProcessResponse's
+            document. Only supports top level document and pages field
+            so it must be in the form of ``{document_field_name}`` or
+            ``pages.{page_field_name}``.
     """
 
     inline_document = proto.Field(

google/cloud/documentai_v1/types/document_schema.py

@@ -63,16 +63,16 @@ class EntityType(proto.Message):
                 and cannot be a 'Common Type'. Besides that we use the
                 following naming conventions:
 
-                -  *use snake_casing*
+                -  *use ``snake_casing``*
                 -  name matching is case-insensitive
                 -  Maximum 64 characters.
                 -  Must start with a letter.
                 -  Allowed characters: ASCII letters ``[a-z0-9_-]``. (For
                    backward compatibility internal infrastructure and
                    tooling can handle any ascii character)
-                -  The '/' is sometimes used to denote a property of a type.
-                   For example line_item/amount. This convention is
-                   deprecated, but will still be honored for backward
+                -  The ``/`` is sometimes used to denote a property of a
+                   type. For example ``line_item/amount``. This convention
+                   is deprecated, but will still be honored for backward
                    compatibility.
             base_types (Sequence[str]):
                 The entity type that this type is derived