structifyai 1.172.0__tar.gz → 1.174.0__tar.gz

structifyai-1.174.0/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "1.174.0"
+}

{structifyai-1.172.0 → structifyai-1.174.0}/CHANGELOG.md

@@ -1,5 +1,27 @@
 # Changelog
 
+## 1.174.0 (2026-01-17)
+
+Full Changelog: [v1.173.0...v1.174.0](https://github.com/StructifyAI/structify-python/compare/v1.173.0...v1.174.0)
+
+### Features
+
+* **api:** api update ([4898eed](https://github.com/StructifyAI/structify-python/commit/4898eed760c37a3fab86f244304ba550c9e5d483))
+* **api:** api update ([55a2fda](https://github.com/StructifyAI/structify-python/commit/55a2fdad2d363f308a97ead42cf32e7d4207287e))
+* Update from Structify backend changes ([19bbfab](https://github.com/StructifyAI/structify-python/commit/19bbfab92d40fce911d38402e922030f651be63b))
+* Update from Structify backend changes ([7ead0af](https://github.com/StructifyAI/structify-python/commit/7ead0af03debdb16cfc3139cb3f60467d9ad2383))
+* Update from Structify backend changes ([a5eec29](https://github.com/StructifyAI/structify-python/commit/a5eec29b34a5c109db22c93d78d467f86f31f732))
+* Update from Structify backend changes ([35b058e](https://github.com/StructifyAI/structify-python/commit/35b058e1d132e4467881ee9ed6dea418b95ca68b))
+
+## 1.173.0 (2026-01-15)
+
+Full Changelog: [v1.172.0...v1.173.0](https://github.com/StructifyAI/structify-python/compare/v1.172.0...v1.173.0)
+
+### Features
+
+* Update from Structify backend changes ([abb90e2](https://github.com/StructifyAI/structify-python/commit/abb90e23e4321fde150a9c5ef591045c4822d82d))
+* Update from Structify backend changes ([2436bc0](https://github.com/StructifyAI/structify-python/commit/2436bc009c3b9c4913cd186b24f339825e690bb2))
+
 ## 1.172.0 (2026-01-15)
 
 Full Changelog: [v1.171.0...v1.172.0](https://github.com/StructifyAI/structify-python/compare/v1.171.0...v1.172.0)

{structifyai-1.172.0 → structifyai-1.174.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: structifyai
-Version: 1.172.0
+Version: 1.174.0
 Summary: The official Python library for the structify API
 Project-URL: Homepage, https://github.com/StructifyAI/structify-python
 Project-URL: Repository, https://github.com/StructifyAI/structify-python

{structifyai-1.172.0 → structifyai-1.174.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "structifyai"
-version = "1.172.0"
+version = "1.174.0"
 description = "The official Python library for the structify API"
 dynamic = ["readme"]
 license = "Apache-2.0"

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "structify"
-__version__ = "1.172.0"  # x-release-please-version
+__version__ = "1.174.0"  # x-release-please-version

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/resources/polars.py

@@ -935,20 +935,24 @@ class PolarsResource(SyncAPIResource):
             # Wait for all PDF processing jobs to complete
             self._client.jobs.wait_for_jobs(job_ids=job_ids, title=f"Parsing {table_name} from PDFs", node_id=node_id)
 
-            # Collect results from all processed PDFs
+            # Collect results from all processed PDFs - each result is tagged with its source row_idx
             structured_results: list[dict[str, Any]] = []
 
             def collect_pdf_results(row_idx: int, dataset_name: str) -> List[Dict[str, Any]]:
                 pdf_path = batch_rows[row_idx][path_column]
                 entities_result = self._client.datasets.view_table(dataset=dataset_name, name=table_name)
-                return [{**entity.properties, path_column: pdf_path} for entity in entities_result]
+                return [
+                    {**entity.properties, path_column: pdf_path, "__row_idx__": row_idx} for entity in entities_result
+                ]
 
             with ThreadPoolExecutor(max_workers=MAX_PARALLEL_REQUESTS) as executor:
                 collect_futures = [
                     executor.submit(collect_pdf_results, row_idx, dataset_name)
                     for row_idx, dataset_name in idx_to_dataset.items()
                 ]
-                for future in tqdm(as_completed(collect_futures), total=len(collect_futures), desc="Collecting PDF extractions"):
+                for future in tqdm(
+                    as_completed(collect_futures), total=len(collect_futures), desc="Collecting PDF extractions"
+                ):
                     results = future.result()
                     structured_results.extend(results)
 
@@ -958,17 +962,12 @@ class PolarsResource(SyncAPIResource):
                     if col_name not in result_row:
                         result_row[col_name] = None
 
-            # Create DataFrame with structured results
             if not structured_results:
-                structured_df = pl.DataFrame(
-                    {col: pl.Series([], dtype=polars_schema[col]) for col in polars_schema.names()}
-                )
-            else:
-                structured_df = pl.DataFrame(structured_results, schema=polars_schema)
+                return pl.DataFrame(schema=polars_schema)
 
-            # Join with original batch to preserve any additional columns
-            joined_df = batch_df.join(structured_df, on=path_column, how="left")
-            return joined_df
+            # Build result dataframe directly from structured_results without joining
+            # Each entity is already tagged with path_column from its source PDF
+            return pl.DataFrame(structured_results, schema=polars_schema)
 
         return document_paths.map_batches(structure_batch, schema=polars_schema, no_optimizations=True)
 

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/types/chat_create_session_params.py

@@ -44,6 +44,7 @@ class Config(TypedDict, total=False):
             "bedrock.claude-sonnet-4-bedrock",
             "bedrock.claude-sonnet-4-5-bedrock",
             "bedrock.claude-opus-4-5-bedrock",
+            "bedrock.claude-haiku-4-5-bedrock",
             "gemini.gemini-2.5-pro",
             "gemini.gemini-2.5-flash",
             "gemini.gemini-3-pro-preview",

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/types/code_generate_code_params.py

@@ -48,6 +48,7 @@ class Config(TypedDict, total=False):
             "bedrock.claude-sonnet-4-bedrock",
             "bedrock.claude-sonnet-4-5-bedrock",
             "bedrock.claude-opus-4-5-bedrock",
+            "bedrock.claude-haiku-4-5-bedrock",
             "gemini.gemini-2.5-pro",
             "gemini.gemini-2.5-flash",
             "gemini.gemini-3-pro-preview",

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/types/connector_catalog_list_response.py

@@ -5,10 +5,18 @@ from typing import List
 from .._models import BaseModel
 from .connector_catalog_with_methods import ConnectorCatalogWithMethods
 
-__all__ = ["ConnectorCatalogListResponse"]
+__all__ = ["ConnectorCatalogListResponse", "CategoryCount"]
+
+
+class CategoryCount(BaseModel):
+    category: str
+
+    count: int
 
 
 class ConnectorCatalogListResponse(BaseModel):
+    category_counts: List[CategoryCount]
+
     items: List[ConnectorCatalogWithMethods]
 
     total_count: int

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/types/job_event_body.py

@@ -18,12 +18,14 @@ __all__ = [
     "DerivedProperty",
     "Failed",
     "Completed",
+    "CacheHit",
     "AttemptedMatch",
     "DatahubPageFetched",
     "DatahubDatabasesCreated",
     "DatahubSchemasCreated",
     "DatahubTablesProcessed",
     "DatahubEmbeddingBatch",
+    "ViewedPdfPage",
 ]
 
 
@@ -106,6 +108,14 @@ class Completed(BaseModel):
     message: Optional[str] = None
 
 
+class CacheHit(BaseModel):
+    cached_from_job_id: str
+
+    event_type: Literal["cache_hit"]
+
+    message: Optional[str] = None
+
+
 class AttemptedMatch(BaseModel):
     candidates: List[Dict[str, Union[str, bool, float]]]
 
@@ -117,6 +127,8 @@ class AttemptedMatch(BaseModel):
 
     match_idx: Optional[int] = None
 
+    raw_text: Optional[str] = None
+
 
 class DatahubPageFetched(BaseModel):
     datasets_in_page: int
@@ -164,6 +176,12 @@ class DatahubEmbeddingBatch(BaseModel):
     total_batches: int
 
 
+class ViewedPdfPage(BaseModel):
+    event_type: Literal["viewed_pdf_page"]
+
+    page_index: int
+
+
 JobEventBody: TypeAlias = Annotated[
     Union[
         AgentNavigated,
@@ -175,12 +193,14 @@ JobEventBody: TypeAlias = Annotated[
         DerivedProperty,
         Failed,
         Completed,
+        CacheHit,
         AttemptedMatch,
         DatahubPageFetched,
         DatahubDatabasesCreated,
         DatahubSchemasCreated,
         DatahubTablesProcessed,
         DatahubEmbeddingBatch,
+        ViewedPdfPage,
     ],
     PropertyInfo(discriminator="event_type"),
 ]

{structifyai-1.172.0 → structifyai-1.174.0}/src/structify/types/tool_invocation.py

@@ -1,10 +1,11 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
-from typing import List, Union, Optional
+from typing import Dict, List, Union, Optional
 from typing_extensions import Literal, Annotated, TypeAlias
 
 from .._utils import PropertyInfo
 from .._models import BaseModel
+from .knowledge_graph import KnowledgeGraph
 
 __all__ = [
     "ToolInvocation",
@@ -12,6 +13,21 @@ __all__ = [
     "WebSearchInput",
     "WebNavigate",
     "WebNavigateInput",
+    "ViewPage",
+    "ViewPageInput",
+    "Save",
+    "SaveInput",
+    "SaveEntities",
+    "SaveEntitiesInput",
+    "Exit",
+    "ExitInput",
+    "APIExecute",
+    "APIExecuteInput",
+    "Javascript",
+    "JavascriptInput",
+    "NavigateToIFrame",
+    "NavigateToIFrameInput",
+    "InfiniteScroll",
     "InspectStep",
     "InspectStepInput",
     "ReadNodeLogs",
@@ -66,6 +82,8 @@ class WebSearch(BaseModel):
 class WebNavigateInput(BaseModel):
     url: str
 
+    output_format: Optional[Literal["Text", "Visual"]] = None
+
 
 class WebNavigate(BaseModel):
     input: WebNavigateInput
@@ -73,6 +91,95 @@ class WebNavigate(BaseModel):
     name: Literal["WebNavigate"]
 
 
+class ViewPageInput(BaseModel):
+    page_number: int
+
+
+class ViewPage(BaseModel):
+    input: ViewPageInput
+
+    name: Literal["ViewPage"]
+
+
+class SaveInput(BaseModel):
+    knowledge_graph: KnowledgeGraph
+    """
+    Knowledge graph info structured to deserialize and display in the same format
+    that the LLM outputs. Also the first representation of an LLM output in the
+    pipeline from raw tool output to being merged into a DB
+    """
+
+    reason: str
+
+    sources: List[str]
+
+
+class Save(BaseModel):
+    input: SaveInput
+
+    name: Literal["Save"]
+
+
+class SaveEntitiesInput(BaseModel):
+    entities: List[Dict[str, Dict[str, object]]]
+
+    reason: str
+
+    sources: List[str]
+
+
+class SaveEntities(BaseModel):
+    input: SaveEntitiesInput
+
+    name: Literal["SaveEntities"]
+
+
+class ExitInput(BaseModel):
+    reason: str
+
+
+class Exit(BaseModel):
+    input: ExitInput
+
+    name: Literal["Exit"]
+
+
+class APIExecuteInput(BaseModel):
+    code: str
+
+
+class APIExecute(BaseModel):
+    input: APIExecuteInput
+
+    name: Literal["ApiExecute"]
+
+
+class JavascriptInput(BaseModel):
+    code: str
+
+
+class Javascript(BaseModel):
+    input: JavascriptInput
+
+    name: Literal["Javascript"]
+
+
+class NavigateToIFrameInput(BaseModel):
+    index: int
+
+
+class NavigateToIFrame(BaseModel):
+    input: NavigateToIFrameInput
+
+    name: Literal["NavigateToIFrame"]
+
+
+class InfiniteScroll(BaseModel):
+    input: object
+
+    name: Literal["InfiniteScroll"]
+
+
 class InspectStepInput(BaseModel):
     step_name: str
 
@@ -329,6 +436,14 @@ ToolInvocation: TypeAlias = Annotated[
     Union[
         WebSearch,
         WebNavigate,
+        ViewPage,
+        Save,
+        SaveEntities,
+        Exit,
+        APIExecute,
+        Javascript,
+        NavigateToIFrame,
+        InfiniteScroll,
         InspectStep,
         ReadNodeLogs,
         DeleteFile,

structifyai-1.172.0/.release-please-manifest.json

@@ -1,3 +0,0 @@
-{
-  ".": "1.172.0"
-}