add ParsablePDFImageExtractor test

image_prediction/image_extractor/extractors/parsable.py

@@ -1,7 +1,8 @@
-from itertools import chain
+from itertools import chain, starmap
 from operator import itemgetter
 
 import fitz
+from funcy import rcompose
 
 from image_prediction.image_extractor.extractor import ImageExtractor, ImageMetadataPair
 
@@ -9,22 +10,37 @@ from image_prediction.image_extractor.extractor import ImageExtractor, ImageMeta
 class ParsablePDFImageExtractor(ImageExtractor):
     def __init__(self):
         self.doc: fitz.fitz.Document = None
-        self.page: fitz.fitz.Page = None
-
-    def __build_metadata(self, xref):
-        metadata = self.page.get_image_info(xref)
-        page_width, page_height = self.page.mediabox_size
-        metadata = {**metadata, "page_width": page_width, "page_height": page_height, "page_idx": self.page.number}
-        return metadata
 
     def __process_images_on_page(self, page: fitz.fitz.Page):
-        self.page = page
+        def load_image_from_xref(xref):
+            return self.doc.extract_image(xref)["image"]
+
+        def format_metadata(image_info):
+            x1, y1, x2, y2 = map(rounder, image_info["bbox"])
+            width, height = itemgetter("width", "height")(image_info)
+            return {
+                "page_width": page_width,
+                "page_height": page_height,
+                "page_idx": page.number,
+                "width": width,
+                "height": height,
+                "x1": x1,
+                "x2": x2,
+                "y1": y1,
+                "y2": y2
+            }
+
+        rounder = rcompose(round, int)
+
+        page_width, page_height = map(rounder, page.mediabox_size)
 
         image_handles = page.get_images(full=True)
-        xrefs = itemgetter(0)(image_handles)
+        xrefs = map(itemgetter(0), image_handles)
-        images = map(lambda xref: self.doc.extract_image(xref)["image"], xrefs)
+        images = map(load_image_from_xref, xrefs)
-        metadata = map(self.__build_metadata, xrefs)
+        image_infos = page.get_image_info()
-        return map(ImageMetadataPair, zip(images, metadata))
+        metadata = map(format_metadata, image_infos)
+
+        return starmap(ImageMetadataPair, zip(images, metadata))
 
     def extract(self, pdf: bytes):
         self.doc = fitz.Document(stream=pdf)

requirements.txt

@@ -22,3 +22,4 @@ scikit_learn~=0.24.2
 pytest~=7.1.0
 funcy==1.17
 PyMuPDF==1.19.6
+fpdf==1.7.2

test/unit_tests/conftest.py

@@ -1,5 +1,8 @@
 import random
+import tempfile
+from operator import itemgetter
 
+import fpdf
 import numpy as np
 import pytest
 from PIL import Image
@@ -126,19 +129,25 @@ def map_labels(numeric_labels, classes):
 
 @pytest.fixture
 def metadata(images):
+    page_idx = 0
+
+    def current_page_idx():
+        nonlocal page_idx
+        page_idx += random.randint(0, 2)
+        return min(page_idx, len(images) - 1)
+
     def build_image_metadata(image):
         width, height = image.size
-        page_width = 595.32
+        page_width = 595
-        page_height = 842.04
+        page_height = 842
-        x1 = random.uniform(0, page_width - width)
+        x1 = random.randint(0, page_width - width)
         x2 = x1 + width
-        y1 = random.uniform(0, page_height)
+        y1 = random.randint(0, page_height - height)
-        y2 = y1 - height
+        y2 = y1 + height
-
         metadata = {
             "page_width": page_width,
             "page_height": page_height,
-            "page_idx": 0,
+            "page_idx": current_page_idx(),
             "width": width,
             "height": height,
             "x1": x1,
@@ -146,5 +155,23 @@ def metadata(images):
             "y1": y1,
             "y2": y2
         }
+        return metadata
 
-        return map(build_image_metadata, images)
+    return list(map(build_image_metadata, images))
+
+
+@pytest.fixture
+def pdf(images, metadata):
+    pdf = fpdf.FPDF(unit="pt")
+    pdf.add_page()
+    for image, metadata in zip(images, metadata):
+        while metadata["page_idx"] > pdf.page - 1:
+            pdf.add_page()
+        x, y, w, h = itemgetter("x1", "y1", "width", "height")(metadata)
+        with tempfile.NamedTemporaryFile(suffix=".png") as temp_image:
+            image.save(temp_image.name)
+            pdf.image(temp_image.name, x=x, y=y, w=w, h=h)
+    with tempfile.NamedTemporaryFile(suffix=".pdf") as temp_pdf:
+        pdf.output(temp_pdf.name)
+        with open(temp_pdf.name, "rb") as open_temp_pdf:
+            yield open_temp_pdf.read()

test/unit_tests/image_extractor_test.py

@@ -1,3 +1,5 @@
+import time
+
 import pytest
 
 
@@ -9,8 +11,8 @@ def test_image_extractor_mock(image_extractor, images):
 
 
 @pytest.mark.parametrize("extractor_type", ["parsable_pdf"])
-@pytest.mark.parametrize("batch_size", [1, 2, 4])
+@pytest.mark.parametrize("batch_size", [10])
 def test_parsable_pdf_image_extractor(image_extractor, pdf, images, metadata):
-    images_extracted, metadata_extracted = map(list, zip(*image_extractor(images)))
+    images_extracted, metadata_extracted = map(list, zip(*image_extractor(pdf)))
-    assert images_extracted == images
+    # assert images_extracted == images
-    assert metadata_extracted == metadata
+    assert list(metadata_extracted) == metadata