def children_are_recursively_converted_to_text():
element = documents.document([
documents.paragraph(
[documents.text("Hello "),
documents.text("world.")], {})
])
result = extract_raw_text_from_element(element)
assert_equal("Hello world.\n\n", result)
def main_document_is_found_using_package_relationships():
fileobj = _create_zip({
"word/document2.xml":
textwrap.dedent("""\
<?xml version="1.0" encoding="utf-8" ?>
<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
<w:body>
<w:p>
<w:r>
<w:t>Hello.</w:t>
</w:r>
</w:p>
</w:body>
</w:document>
"""),
"_rels/.rels":
textwrap.dedent("""\
<?xml version="1.0" encoding="utf-8"?>
<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
<Relationship Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="/word/document2.xml" Id="rId1"/>
</Relationships>
"""),
})
result = docx.read(fileobj=fileobj)
expected_document = documents.document(
[documents.paragraph([documents.run([documents.text("Hello.")])])])
assert_equal(expected_document, result.value)
def main_document_is_found_using_package_relationships():
fileobj = _create_zip({
"word/document2.xml": textwrap.dedent("""\
<?xml version="1.0" encoding="utf-8" ?>
<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
<w:body>
<w:p>
<w:r>
<w:t>Hello.</w:t>
</w:r>
</w:p>
</w:body>
</w:document>
"""),
"_rels/.rels": textwrap.dedent("""\
<?xml version="1.0" encoding="utf-8"?>
<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
<Relationship Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="/word/document2.xml" Id="rId1"/>
</Relationships>
"""),
})
result = docx.read(fileobj=fileobj)
expected_document = documents.document([
documents.paragraph([
documents.run([
documents.text("Hello.")
])
])
])
assert_equal(expected_document, result.value)
def subscript_runs_are_wrapped_in_sub_tags():
result = convert_document_element_to_html(
documents.run(
children=[documents.text("Hello")],
vertical_alignment=documents.VerticalAlignment.subscript,
), )
assert_equal("<sub>Hello</sub>", result.value)
def style_mapping_for_underline_runs_does_not_close_parent_elements():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")],
is_underline=True,
is_bold=True),
style_map=[_style_mapping("u => em")])
assert_equal("<strong><em>Hello</em></strong>", result.value)
def sdt_is_read_using_sdt_content(self):
element = xml_element("w:sdt", {}, [
xml_element("w:sdtContent", {}, [
xml_element("w:t", {}, [xml_text("Blackdown")]),
]),
])
result = _read_and_get_document_xml_element(element)
assert_equal(documents.text("Blackdown"), result)
def sdt_is_read_using_sdt_content():
element = xml_element("w:sdt", {}, [
xml_element("w:sdtContent", {}, [
xml_element("w:t", {}, [xml_text("Blackdown")]),
]),
])
result = _read_and_get_document_xml_element(element)
assert_equal(documents.text("Blackdown"), result)
def strikethrough_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_strikethrough=True),
style_map=[
_style_mapping("strike => del")
]
)
assert_equal("<del>Hello</del>", result.value)
def underline_runs_can_be_mapped_using_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True),
style_map=[
_style_mapping("u => em")
]
)
assert_equal("<em>Hello</em>", result.value)
def style_mapping_for_underline_runs_does_not_close_parent_elements():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True, is_bold=True),
style_map=[
_style_mapping("u => em")
]
)
assert_equal("<strong><em>Hello</em></strong>", result.value)
def strikethrough_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_strikethrough=True),
style_map=[
_style_mapping("strike => del")
]
)
assert_equal("<del>Hello</del>", result.value)
def small_caps_runs_can_be_mapped_using_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_small_caps=True),
style_map=[
_style_mapping("small-caps => span")
]
)
assert_equal("<span>Hello</span>", result.value)
def subscript_runs_are_wrapped_in_sub_tags():
result = convert_document_element_to_html(
documents.run(
children=[documents.text("Hello")],
vertical_alignment=documents.VerticalAlignment.subscript,
),
)
assert_equal("<sub>Hello</sub>", result.value)
def small_caps_runs_can_be_mapped_using_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_small_caps=True),
style_map=[
_style_mapping("small-caps => span")
]
)
assert_equal("<span>Hello</span>", result.value)
def underline_runs_can_be_mapped_using_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True),
style_map=[
_style_mapping("u => em")
]
)
assert_equal("<em>Hello</em>", result.value)
def can_read_document_with_single_paragraph_with_single_run_of_text(self):
with open(test_path("single-paragraph.docx"), "rb") as fileobj:
result = docx.read(fileobj=fileobj)
expected_document = documents.document([
documents.paragraph([
documents.run([documents.text("Walking on imported air")])
])
])
assert_equal(expected_document, result.value)
def can_read_document_with_single_paragraph_with_single_run_of_text(self):
with open(test_path("single-paragraph.docx"), "rb") as fileobj:
result = docx.read(fileobj=fileobj)
expected_document = documents.document([
documents.paragraph([
documents.run([
documents.text("Walking on imported air")
])
])
])
assert_equal(expected_document, result.value)
def italic_runs_are_wrapped_in_emphasis_tags_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_italic=True),
)
assert_equal("<em>Hello</em>", result.value)
def bold_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_bold=True),
style_map=[_style_mapping("b => em")]
)
assert_equal("<em>Hello</em>", result.value)
def italic_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_italic=True),
style_map=[_style_mapping("i => strong")])
assert_equal("<strong>Hello</strong>", result.value)
def _run_with_text(text):
return documents.run(children=[documents.text(text)])
def soft_hyphen_element_is_read_as_soft_hyphen_character():
element = xml_element("w:softHyphen")
tab = _read_and_get_document_xml_element(element)
assert_equal(documents.text(u"\u00ad"), tab)
def includes_children(self):
children = [documents.text("child 1"), documents.text("child 2")]
element = documents.paragraph(children=children)
assert_equal(children, get_descendants(element))
def includes_indirect_descendants(self):
grandchild = documents.text("grandchild")
child = documents.run(children=[grandchild])
element = documents.paragraph(children=[child])
assert_equal([grandchild, child], get_descendants(element))
def underline_runs_can_be_wrapped_in_tags():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True),
convert_underline=mammoth.underline.element("u"))
assert_equal("<u>Hello</u>", result.value)
def underline_runs_are_ignored_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True), )
assert_equal("Hello", result.value)
def italic_runs_are_wrapped_in_emphasis_tags():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_italic=True), )
assert_equal("<em>Hello</em>", result.value)
def underline_runs_can_be_wrapped_in_tags_using_convert_underline_argument():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_underline=True),
convert_underline=mammoth.underline.element("u")
)
assert_equal("<u>Hello</u>", result.value)
def no_break_hyphen_element_is_read_as_non_breaking_hyphen_character():
element = xml_element("w:noBreakHyphen")
tab = _read_and_get_document_xml_element(element)
assert_equal(documents.text(unichr(0x2011)), tab)
def bold_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_bold=True),
style_map=[_style_mapping("b => em")])
assert_equal("<em>Hello</em>", result.value)
def italic_runs_can_be_configured_with_style_mapping():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_italic=True),
style_map=[_style_mapping("i => strong")]
)
assert_equal("<strong>Hello</strong>", result.value)
def underline_runs_are_ignored_by_default():
result = convert_document_element_to_html(documents.run(children=[documents.text("Hello")], is_underline=True))
assert_equal("Hello", result.value)
def _run_with_text(text):
return documents.run(children=[documents.text(text)])
def strikethrough_runs_are_wrapped_in_s_elements_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_strikethrough=True),
)
assert_equal("<s>Hello</s>", result.value)
def includes_children(self):
children = [documents.text("child 1"), documents.text("child 2")]
element = documents.paragraph(children=children)
assert_equal(children, get_descendants(element))
def small_caps_runs_are_ignored_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_small_caps=True),
)
assert_equal("Hello", result.value)
def includes_indirect_descendants(self):
grandchild = documents.text("grandchild")
child = documents.run(children=[grandchild])
element = documents.paragraph(children=[child])
assert_equal([grandchild, child], get_descendants(element))
def no_break_hyphen_element_is_read_as_non_breaking_hyphen_character():
element = xml_element("w:noBreakHyphen")
tab = _read_and_get_document_xml_element(element)
assert_equal(documents.text(unichr(0x2011)), tab)
def bold_runs_are_wrapped_in_strong_tags_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_bold=True),
)
assert_equal("<strong>Hello</strong>", result.value)
def bold_runs_are_wrapped_in_strong_tags():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")], is_bold=True), )
assert_equal("<strong>Hello</strong>", result.value)
def strikethrough_runs_are_wrapped_in_s_elements_by_default():
result = convert_document_element_to_html(
documents.run(children=[documents.text("Hello")],
is_strikethrough=True), )
assert_equal("<s>Hello</s>", result.value)