Refactor: Remove OCR related code for entire page OCR

CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.6.7
+
+* Remove all OCR related code expect the table OCR code
+
 ## 0.6.6
 
 * Stop passing ocr_languages parameter into paddle to avoid invalid paddle language code error, this will be fixed until

Dockerfile

@@ -20,7 +20,6 @@ RUN python3.8 -m pip install pip==${PIP_VERSION} && \
   pip install --no-cache -r requirements/base.txt && \
   pip install --no-cache -r requirements/test.txt && \
   pip install --no-cache -r requirements/dev.txt && \
-  pip install "unstructured.PaddleOCR" && \
   dnf -y groupremove "Development Tools" && \
   dnf clean all
 

test_unstructured_inference/conftest.py

@@ -107,15 +107,6 @@ def mock_embedded_text_regions():
     ]
 
 
-@pytest.fixture()
-def mock_ocr_regions():
-    return [
-        EmbeddedTextRegion(10, 10, 90, 90, text="0", source=None),
-        EmbeddedTextRegion(200, 200, 300, 300, text="1", source=None),
-        EmbeddedTextRegion(500, 320, 600, 350, text="3", source=None),
-    ]
-
-
 # TODO(alan): Make a better test layout
 @pytest.fixture()
 def mock_layout(mock_embedded_text_regions):
@@ -130,19 +121,3 @@ def mock_layout(mock_embedded_text_regions):
         )
         for r in mock_embedded_text_regions
     ]
-
-
-@pytest.fixture()
-def mock_inferred_layout(mock_embedded_text_regions):
-    return [
-        LayoutElement(
-            r.x1,
-            r.y1,
-            r.x2,
-            r.y2,
-            text=None,
-            source=None,
-            type="Text",
-        )
-        for r in mock_embedded_text_regions
-    ]