def main():
output = PdfFileWriter()
input1 = PdfFileReader(
open("/Users/odin/Downloads/What-is-Life-By-Erwin-Schrodinger.pdf",
"rb"))
print("document1.pdf has %d pages." % input1.getNumPages())
page1 = input1.getPage(0)
print(f'raw size of pdf pages: {page1.mediaBox}')
width_trim_1 = 75
width_trim_2 = 3
top_trim = 52
bottom_trim = 60
first_half_split = RectangleObject(
[width_trim_1, bottom_trim, 306 - width_trim_2, 792 - top_trim])
second_half_split = RectangleObject(
[306 + width_trim_2, bottom_trim, 612 - width_trim_1, 792 - top_trim])
for page_num in range(0, input1.getNumPages()):
p = input1.getPage(page_num)
p_copy = copy(p)
p.mediaBox = first_half_split
p_copy.mediaBox = second_half_split
output.addPage(p)
output.addPage(p_copy)
output_stream = open("pdf_to_text_output.pdf", "wb")
output.write(output_stream)
output_stream.close()
def test_writer_operations():
"""
This test just checks if the operation throws an exception.
This should be done way more thoroughly: It should be checked if the
output is as expected.
"""
pdf_path = os.path.join(RESOURCE_ROOT, "crazyones.pdf")
pdf_outline_path = os.path.join(RESOURCE_ROOT, "pdflatex-outline.pdf")
reader = PdfFileReader(pdf_path)
reader_outline = PdfFileReader(pdf_outline_path)
writer = PdfFileWriter()
page = reader.pages[0]
with pytest.raises(PageSizeNotDefinedError) as exc:
writer.addBlankPage()
assert exc.value.args == ()
writer.insertPage(page, 1)
writer.removeText()
writer.insertPage(reader_outline.pages[0], 0)
writer.addBookmarkDestination(page)
writer.addBookmark("A bookmark", 0)
# output.addNamedDestination("A named destination", 1)
writer.removeLinks()
# assert output.getNamedDestRoot() == ['A named destination', IndirectObject(9, 0, output)]
writer.addBlankPage()
writer.addURI(2, "https://example.com", RectangleObject([0, 0, 100, 100]))
writer.addLink(2, 1, RectangleObject([0, 0, 100, 100]))
assert writer.getPageLayout() is None
writer.setPageLayout("SinglePage")
assert writer.getPageLayout() == "SinglePage"
assert writer.getPageMode() is None
writer.setPageMode("UseNone")
assert writer.getPageMode() == "UseNone"
writer.insertBlankPage(width=100, height=100)
writer.insertBlankPage() # without parameters
# This gives "KeyError: '/Contents'" - is that a bug?
# output.removeImages()
writer.addMetadata({"author": "Martin Thoma"})
writer.addAttachment("foobar.gif", b"foobarcontent")
# finally, write "output" to PyPDF2-output.pdf
tmp_path = "dont_commit_writer.pdf"
with open(tmp_path, "wb") as output_stream:
writer.write(output_stream)
# cleanup
os.remove(tmp_path)
def process_page(self, page_idx, bbox_func, *args, **kwargs):
"""Process a single page and add it to the writer """
tmpfname = self.export_page(page_idx)
bbox = bbox_func(tmpfname, *args, **kwargs)
thepage = self.reader.getPage(page_idx)
thepage.cropBox = RectangleObject(bbox)
self.writer.addPage(thepage)
os.unlink(tmpfname)
return 0
def transformAnnot(p, rot, ratio, tx, ty):
"""transform the Annotations of a pdf page (takes the page and the same arguments like mergeRotatedScaledTranslatedPage)"""
if '/Annots' in p:
for a in p['/Annots']:
annot = a.getObject()
r = RectangleObject(annot['/Rect'])
(x0, y0) = r.upperLeft
(x1, y1) = r.lowerRight
if rot == 90:
x0, y0 = y0, x0
x1, y1 = y1, x1
annot.update({
NameObject('/Rect'):
RectangleObject([
x0 * ratio + tx, y0 * ratio + ty, x1 * ratio + tx,
y1 * ratio + ty
])
})
def test_add_link():
reader = PdfReader(os.path.join(RESOURCE_ROOT, "pdflatex-outline.pdf"))
writer = PdfWriter()
for page in reader.pages:
writer.add_page(page)
from PyPDF2.generic import RectangleObject
writer.add_link(
1,
2,
RectangleObject([0, 0, 100, 100]),
border=[1, 2, 3, [4]],
fit="/Fit",
)
writer.add_link(2, 3, RectangleObject([20, 30, 50, 80]), [1, 2, 3],
"/FitH", None)
writer.add_link(
3,
0,
"[ 200 300 250 350 ]",
[0, 0, 0],
"/XYZ",
0,
0,
2,
)
writer.add_link(
3,
0,
[100, 200, 150, 250],
border=[0, 0, 0],
)
# write "output" to PyPDF2-output.pdf
tmp_filename = "dont_commit_link.pdf"
with open(tmp_filename, "wb") as output_stream:
writer.write(output_stream)
# Cleanup
os.remove(tmp_filename)
def test_add_uri():
reader = PdfReader(os.path.join(RESOURCE_ROOT, "pdflatex-outline.pdf"))
writer = PdfWriter()
for page in reader.pages:
writer.add_page(page)
from PyPDF2.generic import RectangleObject
writer.add_uri(
1,
"http://www.example.com",
RectangleObject([0, 0, 100, 100]),
border=[1, 2, 3, [4]],
)
writer.add_uri(
2,
"https://pypdf2.readthedocs.io/en/latest/",
RectangleObject([20, 30, 50, 80]),
border=[1, 2, 3],
)
writer.add_uri(
3,
"https://pypdf2.readthedocs.io/en/latest/user/adding-pdf-annotations.html",
"[ 200 300 250 350 ]",
border=[0, 0, 0],
)
writer.add_uri(
3,
"https://pypdf2.readthedocs.io/en/latest/user/adding-pdf-annotations.html",
[100, 200, 150, 250],
border=[0, 0, 0],
)
# write "output" to PyPDF2-output.pdf
tmp_filename = "dont_commit_uri.pdf"
with open(tmp_filename, "wb") as output_stream:
writer.write(output_stream)
# Cleanup
os.remove(tmp_filename)
def test_page_properties():
reader = PdfReader(os.path.join(RESOURCE_ROOT, "crazyones.pdf"))
page = reader.pages[0]
assert page.mediabox == RectangleObject((0, 0, 612, 792))
assert page.cropbox == RectangleObject((0, 0, 612, 792))
assert page.bleedbox == RectangleObject((0, 0, 612, 792))
assert page.trimbox == RectangleObject((0, 0, 612, 792))
assert page.artbox == RectangleObject((0, 0, 612, 792))
page.bleedbox = RectangleObject((0, 1, 100, 101))
assert page.bleedbox == RectangleObject((0, 1, 100, 101))
def add_comment(output, page, text, rectangle):
obj = output._addObject(
DictionaryObject({
NameObject('/DA'):
TextStringObject(' /Helv 10 Tf'),
NameObject('/Subtype'):
NameObject('/FreeText'),
NameObject('/Rect'):
RectangleObject(rectangle),
NameObject('/Type'):
NameObject('/Annot'),
NameObject('/Contents'):
TextStringObject(text),
NameObject('/C'):
ArrayObject([FloatObject(1),
FloatObject(1),
FloatObject(1)]),
}))
page['/Annots'].append(obj)
def test_RectangleObject():
ro = RectangleObject((1, 2, 3, 4))
assert ro.lower_left == (1, 2)
assert ro.lower_right == (3, 2)
assert ro.upper_left == (1, 4)
assert ro.upper_right == (3, 4)
ro.lower_left = (5, 6)
assert ro.lower_left == (5, 6)
ro.lower_right = (7, 8)
assert ro.lower_right == (7, 8)
ro.upper_left = (9, 11)
assert ro.upper_left == (9, 11)
ro.upper_right = (13, 17)
assert ro.upper_right == (13, 17)
def test_writer_operations():
"""
This test just checks if the operation throws an exception.
This should be done way more thoroughly: It should be checked if the
output is as expected.
"""
pdf_path = os.path.join(RESOURCE_ROOT, "crazyones.pdf")
pdf_outline_path = os.path.join(RESOURCE_ROOT, "pdflatex-outline.pdf")
reader = PdfReader(pdf_path)
reader_outline = PdfReader(pdf_outline_path)
writer = PdfWriter()
page = reader.pages[0]
with pytest.raises(PageSizeNotDefinedError) as exc:
writer.add_blank_page()
assert exc.value.args == ()
writer.insert_page(page, 1)
writer.insert_page(reader_outline.pages[0], 0)
writer.add_bookmark_destination(page)
writer.remove_links()
writer.add_bookmark_destination(page)
bm = writer.add_bookmark("A bookmark", 0, None, (255, 0, 15), True, True,
"/FitBV", 10)
writer.add_bookmark("The XYZ fit", 0, bm, (255, 0, 15), True, True, "/XYZ",
10, 20, 3)
writer.add_bookmark("The FitH fit", 0, bm, (255, 0, 15), True, True,
"/FitH", 10)
writer.add_bookmark("The FitV fit", 0, bm, (255, 0, 15), True, True,
"/FitV", 10)
writer.add_bookmark("The FitR fit", 0, bm, (255, 0, 15), True, True,
"/FitR", 10, 20, 30, 40)
writer.add_bookmark("The FitB fit", 0, bm, (255, 0, 15), True, True,
"/FitB")
writer.add_bookmark("The FitBH fit", 0, bm, (255, 0, 15), True, True,
"/FitBH", 10)
writer.add_bookmark("The FitBV fit", 0, bm, (255, 0, 15), True, True,
"/FitBV", 10)
writer.add_blank_page()
writer.add_uri(2, "https://example.com", RectangleObject([0, 0, 100, 100]))
writer.add_link(2, 1, RectangleObject([0, 0, 100, 100]))
assert writer._get_page_layout() is None
writer._set_page_layout("/SinglePage")
assert writer._get_page_layout() == "/SinglePage"
assert writer._get_page_mode() is None
writer.set_page_mode("/UseNone")
assert writer._get_page_mode() == "/UseNone"
writer.insert_blank_page(width=100, height=100)
writer.insert_blank_page() # without parameters
# TODO: This gives "KeyError: '/Contents'" - is that a bug?
# writer.removeImages()
writer.add_metadata({"author": "Martin Thoma"})
writer.add_attachment("foobar.gif", b"foobarcontent")
# finally, write "output" to PyPDF2-output.pdf
tmp_path = "dont_commit_writer.pdf"
with open(tmp_path, "wb") as output_stream:
writer.write(output_stream)
# cleanup
os.remove(tmp_path)
def validate_mediabox(mediabox: RectangleObject, options):
width, height = (inch * options["page-size"]["width"], inch * options["page-size"]["height"])
return (mediabox.getWidth() == width and mediabox.getHeight() == height) or (mediabox.getWidth() == height and mediabox.getHeight() == width)
def annotate(fp_in, annotations):
reader = PdfFileReader(fp_in)
pdf = PdfFileWriter()
for page in reader.pages:
pdf.addPage(page)
for annotation in annotations:
page = annotation.get('page', 0)
try:
pdfpage = pdf.getPage(page)
except IndexError:
print >> sys.stderr, 'Page %d not found in pdf, not adding annotations %r' % (
page, annotation)
continue
size = pdfpage.mediaBox
angle = int(pdfpage.get('/Rotate', 0))
x = annotation['x']
y = annotation['y']
if angle == 0:
x = float(x)
y = size[3] - float(y) - 20
elif angle == 90:
x, y = float(y) - 2, float(x) - 15
else:
x = float(x)
y = float(y)
print >> sys.stderr, 'Page rotated by %d degrees not implemented yet' % (
angle)
color = annotation.get('color', None)
if isinstance(color, basestring):
if color[:1] != '#':
print >> sys.stderr, 'Unsupported color format: %s' % (color)
color = None
else:
# Assume HTML color with format "#RRGGBB".
try:
color = int(color[1:], 16)
except ValueError as e:
print >> sys.stderr, 'Unsupported color format: %s (%s)' % (
color, e)
color = None
if color is not None:
r, g, b = color >> 16, (color >> 8) & 0xff, color & 0xff
color = (r * BYTE_TO_COLOR, g * BYTE_TO_COLOR, b * BYTE_TO_COLOR)
else:
color = None
pages = pdf.getObject(pdf._pages)
pageref = pages["/Kids"][page]
anno = DictionaryObject()
anno.update({
NameObject('/Type'):
NameObject('/Annot'),
NameObject('/Subtype'):
NameObject('/Text'),
NameObject('/P'):
pageref,
NameObject('/Rect'):
RectangleObject([x, y, x + 18, y + 20]),
NameObject('/Contents'):
TextStringObject(annotation['text']),
NameObject('/C'):
ArrayObject([FloatObject(x) for x in color]),
NameObject('/Open'):
BooleanObject(True),
})
author = annotation.get('author', None)
if author:
anno[NameObject('/T')] = TextStringObject(author)
modified = annotation.get('modified', None)
if modified:
modified = time.strftime('%Y%m%d%H%M%SZ', time.gmtime(modified))
anno[NameObject('/M')] = TextStringObject(modified)
annoRef = pdf._addObject(anno)
annots = pdfpage.get('/Annots', None)
if annots is None:
annots = pdfpage[NameObject('/Annots')] = ArrayObject([annoRef])
else:
annots.append(annoRef)
fp_out = StringIO()
pdf.write(fp_out)
return fp_out.getvalue()
def applyTransform():
r = RectangleObject([0, 0])
x, y = r.lowerLeft