mmar-mapi 1.0.19__tar.gz → 1.0.21__tar.gz

{mmar_mapi-1.0.19 → mmar_mapi-1.0.21}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mmar-mapi
-Version: 1.0.19
+Version: 1.0.21
 Summary: Common pure/IO utilities for multi-modal architectures team
 Keywords: 
 Author: Eugene Tagin

{mmar_mapi-1.0.19 → mmar_mapi-1.0.21}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "mmar-mapi"
 # dynamic version is not supported yet on uv_build
-version = "1.0.19"
+version = "1.0.21"
 description = "Common pure/IO utilities for multi-modal architectures team"
 authors = [{name = "Eugene Tagin", email = "tagin@airi.net"}]
 license = "MIT"

{mmar_mapi-1.0.19 → mmar_mapi-1.0.21}/src/mmar_mapi/api.py

@@ -71,13 +71,18 @@ class TextExtractorAPI:
         raise NotImplementedError
 
 
-PageRange = Annotated[tuple[int, int], AfterValidator(lambda rng: rng[0] <= rng[1])]
+def _validate_page_range(v: tuple[int, int]) -> tuple[int, int]:
+    if v[0] < 1 or v[1] < v[0]:
+        raise ValueError("Invalid page range: start must be ≥ 1 and end must be ≥ start.")
+    return v
+
+
+PageRange = Annotated[tuple[int, int], AfterValidator(_validate_page_range)]
 ForceOCR = StrEnum("ForceOCR", ["ENABLED", "DISABLED", "AUTO"])
 OutputType = StrEnum("OutputType", ["RAW", "PLAIN", "MARKDOWN"])
 
 
-class DocExtractionSpec(BaseModel):
-    page_range: PageRange | None = None
+class ExtractionEngineSpec(BaseModel):
     output_type: OutputType = OutputType.MARKDOWN
     force_ocr: ForceOCR = ForceOCR.AUTO
     do_ocr: bool = False
@@ -88,19 +93,28 @@ class DocExtractionSpec(BaseModel):
     generate_page_images: bool = False
     images_scale: float = 2.0
 
+
+class DocExtractionSpec(BaseModel):
+    page_range: PageRange | None = None
+    engine: ExtractionEngineSpec = ExtractionEngineSpec()
+
     def _update(self, **update):
         return self.model_copy(update=update)
 
+    def _update_engine(self, **engine_update):
+        return self._update(engine=self.engine.model_copy(update=engine_update))
+
     # fmt: off
-    def with_output_type_raw(self): return self._update(output_type=OutputType.RAW)
-    def with_output_type_plain(self): return self._update(output_type=OutputType.PLAIN)
-    def with_ocr(self): return self._update(do_ocr=True)
-    def with_tables(self): return self._update(do_table_structure=True, do_cell_matching=True)
-    def with_images(self): return self._update(do_image_extraction=True)
-    def with_annotations(self): return self._update(do_annotations=True)
-    def with_force_ocr_enabled(self): return self._update(force_ocr=ForceOCR.ENABLED)
-    def with_force_ocr_disabled(self): return self._update(force_ocr=ForceOCR.DISABLED)
-    def with_page_images(self): return self._update(generate_page_images=True)
+    def with_output_type_raw(self): return self._update_engine(output_type=OutputType.RAW)
+    def with_output_type_plain(self): return self._update_engine(output_type=OutputType.PLAIN)
+    def with_ocr(self): return self._update_engine(do_ocr=True)
+    def with_tables(self): return self._update_engine(do_table_structure=True, do_cell_matching=True)
+    def with_images(self): return self._update_engine(do_image_extraction=True)
+    def with_annotations(self): return self._update_engine(do_annotations=True)
+    def with_force_ocr_enabled(self): return self._update_engine(force_ocr=ForceOCR.ENABLED)
+    def with_force_ocr_disabled(self): return self._update_engine(force_ocr=ForceOCR.DISABLED)
+    def with_page_images(self): return self._update_engine(generate_page_images=True)
+
     def with_page_range(self, page_range: PageRange): return self._update(page_range=page_range)
     # fmt: on
 
@@ -130,7 +144,7 @@ class ExtractedPageImage(ExtractedImage):
 
 
 class DocExtractionOutput(BaseModel):
-    config: DocExtractionSpec
+    spec: DocExtractionSpec
     text: str = ""
     tables: list[ExtractedTable] = []
     pictures: list[ExtractedPicture] = []
@@ -138,6 +152,6 @@ class DocExtractionOutput(BaseModel):
 
 
 class DocumentExtractorAPI:
-    def extract(self, *, resource_id: ResourceId, spec: DocExtractionSpec) -> ResourceId:
+    def extract(self, *, resource_id: ResourceId, spec: DocExtractionSpec) -> ResourceId | None:
         """returns file with DocExtractionOutput"""
         raise NotImplementedError

{mmar_mapi-1.0.19 → mmar_mapi-1.0.21}/src/mmar_mapi/file_storage.py

@@ -96,6 +96,9 @@ class FileStorage:
         res = self.download_text(resource_id).split("\n")
         return res
 
+    def get_path(self, resource_id: ResourceId | None) -> Path | None:
+        return self._get_path(resource_id)
+
     def _get_path(self, resource_id: ResourceId | None) -> Path | None:
         if not resource_id:
             return None