nucliadb-models 6.9.2.post5276__py3-none-any.whl → 6.9.6.post5453__py3-none-any.whl

nucliadb_models/augment.py

@@ -0,0 +1,85 @@
+# Copyright 2025 Bosutech XXI S.L.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+from pydantic import BaseModel
+
+from nucliadb_models.common import FieldTypeName
+from nucliadb_models.resource import ExtractedDataTypeName, Resource
+from nucliadb_models.search import Image, ResourceProperties, SearchParamDefaults
+
+ParagraphId = str
+
+
+class AugmentedParagraph(BaseModel):
+    text: str | None = None
+
+    neighbours_before: dict[ParagraphId, str] | None = None
+    neighbours_after: dict[ParagraphId, str] | None = None
+
+    image: Image | None = None
+
+
+class AugmentedField(BaseModel):
+    page_preview_image: Image | None = None
+
+
+class AugmentedResource(Resource):
+    def updated_from(self, origin: Resource):
+        for key in origin.model_fields.keys():
+            self.__setattr__(key, getattr(origin, key))
+
+
+class AugmentResources(BaseModel):
+    given: list[str]
+
+    show: list[ResourceProperties] = SearchParamDefaults.show.to_pydantic_field()
+    extracted: list[ExtractedDataTypeName] = SearchParamDefaults.extracted.to_pydantic_field()
+    field_type_filter: list[FieldTypeName] = SearchParamDefaults.field_type_filter.to_pydantic_field()
+    # TODO: field name filter, da field prefix filter
+
+
+class AugmentParagraph(BaseModel):
+    id: ParagraphId
+
+
+class AugmentParagraphs(BaseModel):
+    given: list[AugmentParagraph]
+
+    text: bool = True
+
+    neighbours_before: int = 0
+    neighbours_after: int = 0
+
+    # paragraph extracted from an image, return an image
+    source_image: bool = False
+
+    # paragraph extracted from a table, return table image
+    table_image: bool = False
+
+    # return page_preview instead of table image if table image enabled
+    table_prefers_page_preview: bool = False
+
+    # paragraph from a page, return page preview image
+    page_preview_image: bool = False
+
+
+class AugmentRequest(BaseModel):
+    resources: AugmentResources
+    paragraphs: AugmentParagraphs
+
+
+class AugmentResponse(BaseModel):
+    resources: dict[str, AugmentedResource]
+    paragraphs: dict[str, AugmentedParagraph]

nucliadb_models/conversation.py

@@ -86,7 +86,7 @@ class FieldConversation(BaseModel):
 
 
 class InputMessageContent(BaseModel):
-    text: str = Field(max_length=10 * 1024)
+    text: str = Field()
     format: MessageFormat = MessageFormat.PLAIN
     attachments: List[FileB64] = Field(default=[], max_length=50)
     attachments_fields: List[FieldRef] = Field(default=[], max_length=50)
@@ -129,7 +129,6 @@ class InputConversationField(BaseModel):
     messages: List[InputMessage] = Field(
         default_factory=list,
         description="List of messages in the conversation field. Each message must have a unique ident. A single conversation can contain up to 51,200 messages. You can add up to 2,048 messages per request.",
-        max_length=2048,
     )
     extract_strategy: Optional[str] = Field(
         default=None,

nucliadb_models/hydration.py

@@ -246,10 +246,11 @@ class Hydration(BaseModel, extra="forbid"):
 ParagraphId = Annotated[
     str,
     StringConstraints(
-        pattern=r"^[0-9a-f]{32}/[acftu]/[a-zA-Z0-9:_-]+/[0-9]+-[0-9]+$",
-        min_length=32 + 1 + 1 + 1 + 1 + 1 + 3,
+        # resource-uuid/field-type/field-id/[split-id/]paragraph-id
+        pattern=r"^[0-9a-f]{32}/[acftu]/[a-zA-Z0-9:_-]+(/[^/]{1,128})?/[0-9]+-[0-9]+$",
+        min_length=32 + 1 + 1 + 1 + 1 + 0 + 0 + 1 + 3,
         # max field id of 250 and 10 digit paragraphs. More than enough
-        max_length=32 + 1 + 1 + 1 + 250 + 1 + 21,
+        max_length=32 + 1 + 1 + 1 + 250 + 1 + 128 + 1 + 21,
     ),
 ]
 

nucliadb_models/search.py

@@ -347,10 +347,12 @@ SortOrderMap = {
 
 class SortOptions(BaseModel):
     field: SortField
-    limit: Optional[int] = Field(None, gt=0)
     order: SortOrder = SortOrder.DESC
 
 
+MAX_RANK_FUSION_WINDOW = 500
+
+
 class RankFusionName(str, Enum):
     RECIPROCAL_RANK_FUSION = "rrf"
 
@@ -380,7 +382,7 @@ class ReciprocalRankFusion(_BaseRankFusion):
     )
     window: Optional[int] = Field(
         default=None,
-        le=500,
+        le=MAX_RANK_FUSION_WINDOW,
         title="RRF window",
         description="Number of elements for retrieval to do RRF. Window must be greater or equal to top_k. Greater values will increase probability of multi match at cost of retrieval time",  # noqa: E501
     )
@@ -503,10 +505,18 @@ class SearchParamDefaults:
     )
     top_k = ParamDefault(
         default=20,
+        gt=-1,
         le=200,
         title="Top k",
         description="The number of results search should return. The maximum number of results allowed is 200.",
     )
+    offset = ParamDefault(
+        default=0,
+        gt=-1,
+        le=1000,
+        title="Results offset",
+        description="The number of results to skip, starting from the beginning in sort order. Used for pagination. It can only be used with the keyword and fulltext indexes.",
+    )
     highlight = ParamDefault(
         default=False,
         title="Highlight",
@@ -532,12 +542,6 @@ class SearchParamDefaults:
         title="Sort order",
         description="Order to sort results with",
     )
-    sort_limit = ParamDefault(
-        default=None,
-        title="Sort limit",
-        description="",
-        gt=0,
-    )
     sort_field = ParamDefault(
         default=None,
         title="Sort field",
@@ -938,12 +942,32 @@ class SearchRequest(BaseSearchRequest):
     )
     faceted: list[str] = SearchParamDefaults.faceted.to_pydantic_field()
     sort: Optional[SortOptions] = SearchParamDefaults.sort.to_pydantic_field()
+    offset: int = SearchParamDefaults.offset.to_pydantic_field()
 
     @field_validator("faceted")
     @classmethod
     def nested_facets_not_supported(cls, facets):
         return validate_facets(facets)
 
+    @model_validator(mode="after")
+    def offset_sort_only_on_keyword_indexes(self):
+        has_non_keyword_indexes = set(self.features) & {SearchOptions.SEMANTIC, SearchOptions.RELATIONS}
+        if has_non_keyword_indexes:
+            if self.offset > 0:
+                raise ValueError("offset cannot be used with the semantic or relations index")
+            if self.sort and self.sort.field != SortField.SCORE:
+                raise ValueError("sort by date cannot be used with the semantic or relations index")
+
+        return self
+
+    @field_validator("sort", mode="after")
+    @classmethod
+    def sorting_by_title_not_supported(cls, value: Optional[SortOptions]) -> Optional[SortOptions]:
+        if value and value.field == SortField.TITLE:
+            raise ValueError("sorting by title not supported in /search")
+
+        return value
+
 
 class Author(str, Enum):
     NUCLIA = "NUCLIA"

{nucliadb_models-6.9.2.post5276.dist-info → nucliadb_models-6.9.6.post5453.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nucliadb_models
-Version: 6.9.2.post5276
+Version: 6.9.6.post5453
 Author-email: Nuclia <nucliadb@nuclia.com>
 License-Expression: Apache-2.0
 Project-URL: Homepage, https://nuclia.com

{nucliadb_models-6.9.2.post5276.dist-info → nucliadb_models-6.9.6.post5453.dist-info}/RECORD

@@ -1,15 +1,16 @@
 nucliadb_models/__init__.py,sha256=3y8-htogKuCZcbhaUZdSjTeEjUSeec9aRWyL8AlKCyM,1077
+nucliadb_models/augment.py,sha256=vAtFh4D4eC4nvfwaRTlfeuAMOL9Z9TFZnUNiRAMasss,2543
 nucliadb_models/common.py,sha256=2dtKG4ZNi9p-yoNY76Uvyu1SlMeNYpH-MnuU3Q6w9Js,8169
 nucliadb_models/configuration.py,sha256=BBrJsNjP324Cw_5J3dBrGwvpkHQYbXEo3TUaI9IqAOg,2449
 nucliadb_models/content_types.py,sha256=36Ga-iGf4ivCqgtXC7imFgegrwHB117s9eqP62JtGv0,3456
-nucliadb_models/conversation.py,sha256=Ts-h7gMf18rw9tbjF2X__vNtMshHRDqbLua_VXb1qOE,5126
+nucliadb_models/conversation.py,sha256=k9bKhkDiqhqmdrDfDPNoUfG7-2H_-KAyuOnETd8zV0E,5081
 nucliadb_models/entities.py,sha256=i-7Y8qmFRRTih5zw0ajv1U_iiXexe66M3TK8hUikQZk,2356
 nucliadb_models/export_import.py,sha256=mNm9IArOLnC6TLupkwqVFhxD5d08mpIVOVFneECv8UA,1073
 nucliadb_models/external_index_providers.py,sha256=pL3leo4MkuJOnKlU1Sg6GT_mnK_VUBxGui-RPmDYVWU,1126
 nucliadb_models/extracted.py,sha256=Owz7LC3le3Dvau3TtRiO8NY84meOf6IxN-RrOqqpMPs,5593
 nucliadb_models/file.py,sha256=tXtgB9c7i2ADsnJ7HdbXyroAmXadGvOeA49htBh7BZo,2263
 nucliadb_models/filters.py,sha256=NQI2-4AFzzJuZy8NeY3jXlTbbU5wxiwMCP-5DrD-7lE,14759
-nucliadb_models/hydration.py,sha256=7SFnAcTQRE9etVccpph6aA1AUqsHVwkzT4YF6Uzl0Gs,14262
+nucliadb_models/hydration.py,sha256=SlAzraJE6DX0uOpZWxu2k_9-ikYorsj0t8xwsWSBQZY,14363
 nucliadb_models/labels.py,sha256=9zqRgkpZuX3kUPwsTTgCH7JyOWK7dM5pwyuHJR86YdU,3949
 nucliadb_models/link.py,sha256=PF5hHLwdOed5TMBTxtokkgWtMh1bFnORZjybh0NwVCw,2526
 nucliadb_models/metadata.py,sha256=OOKGy_83NtlG1QKQZEwMuwu4wbVEe7P30Y2QvnGSDto,8933
@@ -17,7 +18,7 @@ nucliadb_models/notifications.py,sha256=mna8-AoD_29Wds0Thl0AF0zpERnJmYGLZX1w1fUo
 nucliadb_models/processing.py,sha256=nhKuHQjqCdb9zJVkYGPTLub23tK9e_lwL5OCDVymZjY,719
 nucliadb_models/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nucliadb_models/resource.py,sha256=RzCos0QRgSMkaV-p7EoceSmt7UTzt9G9be5BKF-iGrQ,9021
-nucliadb_models/search.py,sha256=gQEXJ9bXXcxswr7aOzvBeGIQlrq5TgRWIqTxKEbSoCE,96409
+nucliadb_models/search.py,sha256=_vn3pDXcK4iwiCfim3BtlD5EaQAeXoxl2IfNDsrKesA,97514
 nucliadb_models/security.py,sha256=opxaDLfvk3aU0sjesK0jGrYLx5h4YCwlKKN0moYs_ig,1150
 nucliadb_models/synonyms.py,sha256=afbaVqSQSxGLwi2PusVaLSRpkOtA5AZmWOKd1f4nl2E,690
 nucliadb_models/text.py,sha256=60bxZnOjRHnDdezR8VfR3AZsXTOwePFPs2BKB8wxBak,3414
@@ -33,7 +34,7 @@ nucliadb_models/graph/responses.py,sha256=Sdq8OgFAL1YT-1lJyLLrkqcScvj7YTEqAUwQ-k
 nucliadb_models/internal/__init__.py,sha256=zG33bUz1rHFPtvqQPWn4rDwBJt3FJodGuQYD45quiQg,583
 nucliadb_models/internal/predict.py,sha256=Pnx6MmLfK65eExe1XnVxqmSlvMwdowewwks9BOEoqMw,2029
 nucliadb_models/internal/shards.py,sha256=__y1OZtWGiNcPQEWfSFOj8yw458WGi7mM4vZe0K-L1Y,1691
-nucliadb_models-6.9.2.post5276.dist-info/METADATA,sha256=-kzs6LaR18FXQmyL87mIkJuOTwmGhctfqrU9Rn1AGuY,745
-nucliadb_models-6.9.2.post5276.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-nucliadb_models-6.9.2.post5276.dist-info/top_level.txt,sha256=UrY1I8oeovIRwkXLYplssTrxQdUjhSEFDFbnwaIV3tA,16
-nucliadb_models-6.9.2.post5276.dist-info/RECORD,,
+nucliadb_models-6.9.6.post5453.dist-info/METADATA,sha256=x-3GFIapOwe9kpigwte5HkdfaNqPk-iL7CuE_NM3PAE,745
+nucliadb_models-6.9.6.post5453.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+nucliadb_models-6.9.6.post5453.dist-info/top_level.txt,sha256=UrY1I8oeovIRwkXLYplssTrxQdUjhSEFDFbnwaIV3tA,16
+nucliadb_models-6.9.6.post5453.dist-info/RECORD,,