def _convert_hn(self, dest):
caj = open(self.filename, "rb")
image_list = []
from pdfwutils import Colorspace, ImageFormat, convert_ImageList
import zlib
for i in range(self.page_num):
caj.seek(self._TOC_END_OFFSET + i * 20)
[
page_data_offset, size_of_text_section, images_per_page,
page_no, unk2, next_page_data_offset
] = struct.unpack("iihhii", caj.read(20))
caj.seek(page_data_offset)
text_header_read32 = caj.read(32)
if (text_header_read32[8:20] == b'COMPRESSTEXT'):
[expanded_text_size] = struct.unpack("i",
text_header_read32[20:24])
import zlib
caj.seek(page_data_offset + 24)
data = caj.read(size_of_text_section - 24)
output = zlib.decompress(data, bufsize=expanded_text_size)
if (len(output) != expanded_text_size):
raise SystemExit("Unexpected:", len(output),
expanded_text_size)
else:
caj.seek(page_data_offset)
output = caj.read(size_of_text_section)
from HNParsePage import HNParsePage
page_style = (next_page_data_offset > page_data_offset)
page_data = HNParsePage(output, page_style)
if (images_per_page > 1):
if (len(page_data.figures) == images_per_page):
image_list.append(None)
image_list.append(page_data.figures)
else:
print("Page %d, Image Count %d != %d" %
(i + 1, len(page_data.figures), images_per_page))
image_list.append(None)
image_list.append(page_data.figures[0:images_per_page])
current_offset = page_data_offset + size_of_text_section
for j in range(images_per_page):
caj.seek(current_offset)
read32 = caj.read(32)
[image_type_enum, offset_to_image_data,
size_of_image_data] = struct.unpack("iii", read32[0:12])
if (offset_to_image_data != current_offset + 12):
raise SystemExit("unusual image offset")
caj.seek(offset_to_image_data)
image_data = caj.read(size_of_image_data)
current_offset = offset_to_image_data + size_of_image_data
if (image_type[image_type_enum] == "JBIG"):
from jbigdec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_item = (Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0)
elif (image_type[image_type_enum] == "JBIG2"):
from jbig2dec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig2()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_item = (Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0)
elif (image_type[image_type_enum] == "JPEG"):
# stock libjpeg location
(SOFn, frame_length, bits_per_pixel, height,
width) = struct.unpack(">HHBHH", image_data[158:167])
if (SOFn != 0xFFC0):
# "Intel(R) JPEG Library" location
(SOFn, frame_length, bits_per_pixel, height,
width) = struct.unpack(">HHBHH",
image_data[0x272:0x27b])
if (SOFn != 0xFFC0):
# neither works, try brute-force
import imagesize
with open(".tmp.jpg", "wb") as f:
f.write(image_data)
(width, height) = imagesize.get(".tmp.jpg")
os.remove(".tmp.jpg")
if (image_type_enum == 1):
# non-inverted JPEG Images
height = -height
image_item = (Colorspace.RGB, (300, 300), ImageFormat.JPEG,
image_data, width, height, [], False, 8, 0)
else:
raise SystemExit("Unknown Image Type %d" %
(image_type_enum))
image_list.append(image_item)
if (len(image_list) == 0):
raise SystemExit("File is pure-text HN; cannot convert to pdf")
pdf_data = convert_ImageList(image_list)
with open('pdf_toc.pdf', 'wb') as f:
f.write(pdf_data)
# Add Outlines
add_outlines(self.get_toc(), "pdf_toc.pdf", dest)
os.remove("pdf_toc.pdf")
def _convert_hn(self, dest):
caj = open(self.filename, "rb")
image_list = []
from pdfwutils import Colorspace, ImageFormat, convert_ImageList
import zlib
for i in range(self.page_num):
caj.seek(self._TOC_END_OFFSET + i * 20)
[
page_data_offset, size_of_text_section, images_per_page,
page_no, unk2, unk3
] = struct.unpack("iihhii", caj.read(20))
current_offset = page_data_offset + size_of_text_section
for j in range(images_per_page):
caj.seek(current_offset)
read32 = caj.read(32)
[image_type_enum, offset_to_image_data,
size_of_image_data] = struct.unpack("iii", read32[0:12])
if (offset_to_image_data != current_offset + 12):
raise SystemExit("unusual image offset")
caj.seek(offset_to_image_data)
image_data = caj.read(size_of_image_data)
current_offset = offset_to_image_data + size_of_image_data
image_name = "image_dump_%04d" % (i + 1)
if (j > 0):
image_name = "image_dump_%04d_%04d" % (i + 1, j)
print("TODO: Multiple Images at Page %04d_%04d" %
(i + 1, j))
if (image_type[image_type_enum] == "JBIG"):
from jbigdec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_list.append(
(Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0))
elif (image_type[image_type_enum] == "JBIG2"):
from jbig2dec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig2()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_list.append(
(Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0))
elif (image_type[image_type_enum] == "JPEG"):
(height, width) = struct.unpack(">HH", image_data[163:167])
image_list.append(
(Colorspace.RGB, (300, 300), ImageFormat.JPEG,
image_data, width, height, [], False, 8, 0))
if (image_type_enum == 1):
print(
"TODO: non-inverted JPEG Images at Page %04d_%04d"
% (i + 1, j))
pdf_data = convert_ImageList(image_list)
with open(dest, 'wb') as f:
f.write(pdf_data)
def _convert_hn(self, dest):
caj = open(self.filename, "rb")
image_list = []
from pdfwutils import Colorspace, ImageFormat, convert_ImageList
import zlib
for i in range(self.page_num):
caj.seek(self._TOC_END_OFFSET + i * 20)
[
page_data_offset, size_of_text_section, images_per_page,
page_no, unk2, next_page_data_offset
] = struct.unpack("iihhii", caj.read(20))
caj.seek(page_data_offset)
text_header_read32 = caj.read(32)
if (text_header_read32[8:20] == b'COMPRESSTEXT'):
[expanded_text_size] = struct.unpack("i",
text_header_read32[20:24])
import zlib
caj.seek(page_data_offset + 24)
data = caj.read(size_of_text_section - 24)
output = zlib.decompress(data, bufsize=expanded_text_size)
if (len(output) != expanded_text_size):
raise SystemExit("Unexpected:", len(output),
expanded_text_size)
else:
caj.seek(page_data_offset)
output = caj.read(size_of_text_section)
from HNParsePage import HNParsePage
page_style = (next_page_data_offset > page_data_offset)
page_data = HNParsePage(output, page_style)
if (images_per_page > 1):
if (len(page_data.figures) == images_per_page):
image_list.append(None)
image_list.append(page_data.figures)
else:
raise SystemExit("Image Count %d != %d" %
(len(page_data.figures), images_per_page))
current_offset = page_data_offset + size_of_text_section
for j in range(images_per_page):
caj.seek(current_offset)
read32 = caj.read(32)
[image_type_enum, offset_to_image_data,
size_of_image_data] = struct.unpack("iii", read32[0:12])
if (offset_to_image_data != current_offset + 12):
raise SystemExit("unusual image offset")
caj.seek(offset_to_image_data)
image_data = caj.read(size_of_image_data)
current_offset = offset_to_image_data + size_of_image_data
if (image_type[image_type_enum] == "JBIG"):
from jbigdec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_item = (Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0)
elif (image_type[image_type_enum] == "JBIG2"):
from jbig2dec import CImage
cimage = CImage(image_data)
out = cimage.DecodeJbig2()
# PBM is only padded to 8 rather than 32.
# If the padding is larger, write padded file.
width = cimage.width
if (cimage.bytes_per_line > ((cimage.width + 7) >> 3)):
width = cimage.bytes_per_line << 3
image_item = (Colorspace.P, (300, 300), ImageFormat.PBM,
zlib.compress(out), width, cimage.height,
[0xffffff, 0], False, 1, 0)
elif (image_type[image_type_enum] == "JPEG"):
(height, width) = struct.unpack(">HH", image_data[163:167])
if (image_type_enum == 1):
# non-inverted JPEG Images
height = -height
image_item = (Colorspace.RGB, (300, 300), ImageFormat.JPEG,
image_data, width, height, [], False, 8, 0)
else:
raise SystemExit("Unknown Image Type %d" %
(image_type_enum))
image_list.append(image_item)
pdf_data = convert_ImageList(image_list)
with open(dest, 'wb') as f:
f.write(pdf_data)