Add new parameter to map to skip_infer_table_types partition arg

CHANGELOG.md

@@ -1,4 +1,4 @@
-## 0.10.20-dev6
+## 0.10.20-dev7
 
 ### Enhancements
 
@@ -9,6 +9,8 @@
 * **Improve title detection in pptx documents** The default title textboxes on a pptx slide are now categorized as titles.
 * **Improve hierarchy detection in pptx documents** List items, and other slide text are properly nested under the slide title. This will enable better chunking of pptx documents.
 * **Refactor of the ingest cli workflow** The refactored approach uses a dynamically set pipeline with a snapshot along each step to save progress and accommodate continuation from a snapshot if an error occurs. This also allows the pipeline to dynamically assign any number of steps to modify the partitioned content before it gets written to a destination.
+* **Expose skip_infer_table_types in ingest CLI** For each connector a new `--skip-infer-table-types` parameter was added to map to the `skip_infer_table_types` partition argument.
+
 ### Features
 
 * **Adds `edit_distance` calculation metrics** In order to benchmark the cleaned, extracted text with unstructured, `edit_distance` (`Levenshtein distance`) is included.

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.10.20-dev6"  # pragma: no cover
+__version__ = "0.10.20-dev7"  # pragma: no cover

unstructured/ingest/cli/interfaces.py

@@ -139,6 +139,12 @@ class CliPartitionConfig(PartitionConfig, CliMixin):
     @staticmethod
     def add_cli_options(cmd: click.Command) -> None:
         options = [
+            click.Option(
+                ["--skip-infer-table-types"],
+                type=DelimitedString(),
+                default=None,
+                help="Option list of document types to skip table extraction on",
+            ),
             click.Option(
                 ["--pdf-infer-table-structure"],
                 default=False,

unstructured/ingest/interfaces.py

@@ -38,6 +38,7 @@ class BaseConfig(DataClassJsonMixin, ABC):
 class PartitionConfig(BaseConfig):
     # where to write structured data outputs
     pdf_infer_table_structure: bool = False
+    skip_infer_table_types: t.Optional[t.List[str]] = None
     strategy: str = "auto"
     ocr_languages: str = "eng"
     encoding: t.Optional[str] = None

unstructured/ingest/pipeline/partition.py

@@ -30,12 +30,19 @@ def run(self, ingest_doc_json) -> str:
             if self.partition_config.ocr_languages
             else []
         )
+        partition_kwargs = {
+            "strategy": self.partition_config.strategy,
+            "languages": languages,
+            "encoding": self.partition_config.encoding,
+            "pdf_infer_table_structure": self.partition_config.pdf_infer_table_structure,
+        }
+        if self.partition_config.skip_infer_table_types:
+            partition_kwargs[
+                "skip_infer_table_types"
+            ] = self.partition_config.skip_infer_table_types
         elements = doc.process_file(
             partition_config=self.partition_config,
-            strategy=self.partition_config.strategy,
-            languages=languages,
-            encoding=self.partition_config.encoding,
-            pdf_infer_table_structure=self.partition_config.pdf_infer_table_structure,
+            **partition_kwargs,
         )
         with open(json_path, "w", encoding="utf8") as output_f:
             logger.info(f"writing partitioned content to {json_path}")