def create_pdf_document(
elements: Union[List[LTComponent], Dict[int, List[LTComponent]]],
font_mapping: Optional[Dict[str, str]] = None,
font_mapping_is_regex: bool = False,
regex_flags: Union[int, re.RegexFlag] = 0,
font_size_precision: int = 1,
element_ordering: Union[ElementOrdering, Callable[
[List], List]] = ElementOrdering.LEFT_TO_RIGHT_TOP_TO_BOTTOM,
) -> "PDFDocument":
"""
Creates a PDF document with the given elements.
"elements" can be a list of elements (in which case a document with a single page
will be created) or a dictionary mapping page number to its list of elements.
"""
if not isinstance(elements, dict):
pages = {1: Page(elements=elements, width=100, height=100)}
else:
pages = {
page_number: Page(elements=elements_list, width=100, height=100)
for page_number, elements_list in elements.items()
}
return PDFDocument(
pages=pages,
font_mapping=font_mapping,
font_mapping_is_regex=font_mapping_is_regex,
regex_flags=regex_flags,
font_size_precision=font_size_precision,
element_ordering=element_ordering,
)
def test_filter_by_page(self):
elem1 = FakePDFMinerTextElement()
elem2 = FakePDFMinerTextElement()
elem3 = FakePDFMinerTextElement()
page1 = Page(width=100, height=100, elements=[elem1, elem2])
page2 = Page(width=100, height=100, elements=[elem3])
doc = PDFDocument({1: page1, 2: page2})
self.assertEqual(len(doc.elements.filter_by_page(1)), 2)
self.assert_original_element_in(elem1, doc.elements.filter_by_page(1))
self.assert_original_element_in(elem2, doc.elements.filter_by_page(1))
def test_eq(self):
with self.assertRaises(NotImplementedError):
self.elem_list == "foo"
second_elem_list = ElementList(self.doc, set([0, 1, 2, 3, 4, 5]))
self.assertTrue(self.elem_list == second_elem_list)
# Test with different indexes
second_elem_list = ElementList(self.doc, set([0, 1, 2, 3, 4]))
self.assertFalse(self.elem_list == second_elem_list)
# Test with different document
doc = PDFDocument(
pages={
1: Page(
elements=[
FakePDFMinerTextElement(),
FakePDFMinerTextElement(),
FakePDFMinerTextElement(),
FakePDFMinerTextElement(),
FakePDFMinerTextElement(),
FakePDFMinerTextElement(),
],
width=100,
height=100,
)
}
)
second_elem_list = ElementList(doc, set([0, 1, 2, 3, 4, 5]))
self.assertFalse(self.elem_list == second_elem_list)
def test_filter_partially_within_bounding_box(self, partially_within_mock):
partially_within_mock.side_effect = (
lambda self, bounding_box: self.text() == "within"
)
elem1 = FakePDFMinerTextElement(text="within")
elem2 = FakePDFMinerTextElement(text="within")
elem3 = FakePDFMinerTextElement()
elem4 = FakePDFMinerTextElement(text="within")
elem5 = FakePDFMinerTextElement()
elem6 = FakePDFMinerTextElement(text="within")
page1 = Page(elements=[elem1, elem2, elem3, elem4], width=100, height=100)
page2 = Page(elements=[elem5, elem6], width=100, height=100)
doc = PDFDocument(pages={1: page1, 2: page2})
elem_list = doc.elements
pdf_elem1 = self.extract_element_from_list(elem1, elem_list)
pdf_elem2 = self.extract_element_from_list(elem2, elem_list)
pdf_elem3 = self.extract_element_from_list(elem3, elem_list)
pdf_elem4 = self.extract_element_from_list(elem4, elem_list)
result = elem_list.filter_partially_within_bounding_box(
BoundingBox(0, 1, 0, 1), 1
)
# expected_bbox is from the left edge of elem1 to the left edge of the page
expected_bbox = BoundingBox(0, 1, 0, 1)
partially_within_mock.assert_has_calls(
[
call(pdf_elem1, expected_bbox),
call(pdf_elem2, expected_bbox),
call(pdf_elem3, expected_bbox),
call(pdf_elem4, expected_bbox),
],
any_order=True,
)
self.assertEqual(len(result), 3)
self.assertIn(pdf_elem1, result)
self.assertIn(pdf_elem2, result)
self.assertIn(pdf_elem4, result)
def test_extract_single_element(self):
with self.assertRaises(MultipleElementsFoundError):
self.elem_list.extract_single_element()
with self.assertRaises(NoElementFoundError):
self.elem_list.filter_by_tag("non_existent_tag").extract_single_element()
elem1 = FakePDFMinerTextElement()
page = Page(elements=[elem1], width=100, height=100)
doc = PDFDocument(pages={1: page})
pdf_elem_1 = self.extract_element_from_list(elem1, doc.elements)
result = doc.elements.extract_single_element()
self.assertEqual(result, pdf_elem_1)
def test_document_with_blank_page(self):
with self.assertRaises(NoElementsOnPageError):
PDFDocument(pages={1: Page(elements=[], width=100, height=100)})
def test_document(self):
el_page_1_top_left = FakePDFMinerTextElement(BoundingBox(0, 1, 2, 3))
el_page_1_top_right = FakePDFMinerTextElement(BoundingBox(2, 3, 2, 3))
el_page_1_bottom_left = FakePDFMinerTextElement(BoundingBox(
0, 1, 0, 1))
el_page_1_bottom_right = FakePDFMinerTextElement(
BoundingBox(2, 3, 0, 1))
page_1 = Page(
elements=[
el_page_1_top_left,
el_page_1_top_right,
el_page_1_bottom_left,
el_page_1_bottom_right,
],
width=100,
height=100,
)
el_page_2_top_left = FakePDFMinerTextElement(BoundingBox(0, 1, 2, 3))
el_page_2_top_right = FakePDFMinerTextElement(BoundingBox(2, 3, 2, 3))
el_page_2_bottom_left = FakePDFMinerTextElement(BoundingBox(
0, 1, 0, 1))
el_page_2_bottom_right = FakePDFMinerTextElement(
BoundingBox(2, 3, 0, 1))
page_2 = Page(
elements=[
el_page_2_bottom_right,
el_page_2_bottom_left,
el_page_2_top_right,
el_page_2_top_left,
],
width=100,
height=100,
)
document = PDFDocument(pages={1: page_1, 2: page_2})
# Checks elements were reordered
expected_ordered_list = [
el_page_1_top_left,
el_page_1_top_right,
el_page_1_bottom_left,
el_page_1_bottom_right,
el_page_2_top_left,
el_page_2_top_right,
el_page_2_bottom_left,
el_page_2_bottom_right,
]
self.assertEqual(
[elem.original_element for elem in document._element_list],
expected_ordered_list,
)
# Checks indexes were assigned properly
self.assertEqual([elem._index for elem in document._element_list],
[0, 1, 2, 3, 4, 5, 6, 7])
# Checks page numbers is correct
self.assertEqual(document.page_numbers, [1, 2])
# Checks number of pages is correct
self.assertEqual(document.number_of_pages, 2)
# Checks pages were assigned properly
self.assertEqual(
[elem.page_number for elem in document._element_list],
[1, 1, 1, 1, 2, 2, 2, 2],
)
# Checks pages were instantiated correctly
pdf_page_1 = document.get_page(1)
self.assertEqual(page_1.width, pdf_page_1.width)
self.assertEqual(page_1.height, pdf_page_1.height)
self.assertEqual(el_page_1_top_left,
pdf_page_1.start_element.original_element)
self.assertEqual(el_page_1_bottom_right,
pdf_page_1.end_element.original_element)
self.assertEqual(pdf_page_1.page_number, 1)
self.assertEqual(pdf_page_1.elements,
ElementList(document, set([0, 1, 2, 3])))
pdf_page_2 = document.get_page(2)
self.assertEqual(page_2.width, pdf_page_2.width)
self.assertEqual(page_2.height, pdf_page_2.height)
self.assertEqual(el_page_2_top_left,
pdf_page_2.start_element.original_element)
self.assertEqual(el_page_2_bottom_right,
pdf_page_2.end_element.original_element)
self.assertEqual(pdf_page_2.page_number, 2)
self.assertEqual(pdf_page_2.elements,
ElementList(document, set([4, 5, 6, 7])))
self.assertEqual(document.pages, [pdf_page_1, pdf_page_2])
self.assertEqual(document.elements,
ElementList(document, set([0, 1, 2, 3, 4, 5, 6, 7])))
with self.assertRaises(PageNotFoundError):
document.get_page(3)
def test_horizontally_in_line_with(self, partially_within_mock):
partially_within_mock.side_effect = (
lambda self, bounding_box: self.text() == "within"
)
elem1 = FakePDFMinerTextElement(
text="within", bounding_box=BoundingBox(50, 51, 50, 51)
)
elem2 = FakePDFMinerTextElement(text="within")
elem3 = FakePDFMinerTextElement()
elem4 = FakePDFMinerTextElement(text="within")
elem5 = FakePDFMinerTextElement()
elem6 = FakePDFMinerTextElement(text="within")
page1 = Page(elements=[elem1, elem2, elem3, elem4], width=100, height=100)
page2 = Page(elements=[elem5, elem6], width=100, height=100)
doc = PDFDocument(pages={1: page1, 2: page2})
elem_list = doc.elements
pdf_elem1 = self.extract_element_from_list(elem1, elem_list)
pdf_elem2 = self.extract_element_from_list(elem2, elem_list)
pdf_elem3 = self.extract_element_from_list(elem3, elem_list)
pdf_elem4 = self.extract_element_from_list(elem4, elem_list)
result = elem_list.horizontally_in_line_with(pdf_elem1)
# expected_bbox is from the left edge of elem1 to the left edge of the page
expected_bbox = BoundingBox(0, 100, 50, 51)
partially_within_mock.assert_has_calls(
[
call(pdf_elem1, expected_bbox),
call(pdf_elem2, expected_bbox),
call(pdf_elem3, expected_bbox),
call(pdf_elem4, expected_bbox),
],
any_order=True,
)
self.assertEqual(len(result), 2)
self.assertIn(pdf_elem2, result)
self.assertIn(pdf_elem4, result)
# Also test with inclusive=True
partially_within_mock.reset_mock()
result = elem_list.horizontally_in_line_with(pdf_elem1, inclusive=True)
partially_within_mock.assert_has_calls(
[
call(pdf_elem1, expected_bbox),
call(pdf_elem2, expected_bbox),
call(pdf_elem3, expected_bbox),
call(pdf_elem4, expected_bbox),
],
any_order=True,
)
self.assertEqual(len(result), 3)
self.assertIn(pdf_elem1, result)
self.assertIn(pdf_elem2, result)
self.assertIn(pdf_elem4, result)
# Test specifying tolerance
expected_bbox = BoundingBox(0, 100, 50.1, 50.9)
partially_within_mock.reset_mock()
result = elem_list.horizontally_in_line_with(pdf_elem1, tolerance=0.1)
partially_within_mock.assert_has_calls(
[
call(pdf_elem1, expected_bbox),
call(pdf_elem2, expected_bbox),
call(pdf_elem3, expected_bbox),
call(pdf_elem4, expected_bbox),
],
any_order=True,
)