structifyai 1.172.0__tar.gz → 1.173.0__tar.gz

structifyai-1.173.0/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "1.173.0"
+}

{structifyai-1.172.0 → structifyai-1.173.0}/CHANGELOG.md

@@ -1,5 +1,14 @@
 # Changelog
 
+## 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-1.173.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: structifyai
-Version: 1.172.0
+Version: 1.173.0
 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-1.173.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "structifyai"
-version = "1.172.0"
+version = "1.173.0"
 description = "The official Python library for the structify API"
 dynamic = ["readme"]
 license = "Apache-2.0"

{structifyai-1.172.0 → structifyai-1.173.0}/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__ = "1.173.0"  # x-release-please-version

{structifyai-1.172.0 → structifyai-1.173.0}/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-1.173.0}/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-1.173.0}/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-1.173.0}/src/structify/types/job_event_body.py

@@ -24,6 +24,7 @@ __all__ = [
     "DatahubSchemasCreated",
     "DatahubTablesProcessed",
     "DatahubEmbeddingBatch",
+    "ViewedPdfPage",
 ]
 
 
@@ -164,6 +165,12 @@ class DatahubEmbeddingBatch(BaseModel):
     total_batches: int
 
 
+class ViewedPdfPage(BaseModel):
+    event_type: Literal["viewed_pdf_page"]
+
+    page_index: int
+
+
 JobEventBody: TypeAlias = Annotated[
     Union[
         AgentNavigated,
@@ -181,6 +188,7 @@ JobEventBody: TypeAlias = Annotated[
         DatahubSchemasCreated,
         DatahubTablesProcessed,
         DatahubEmbeddingBatch,
+        ViewedPdfPage,
     ],
     PropertyInfo(discriminator="event_type"),
 ]

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

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