def __init__(self, pdf_path):
self.pdf_path = pdf_path
fd = open(pdf_path, "rb")
doc = PDFDocument(fd)
self.viewer = SimplePDFViewer(fd)
self.pages = [p for p in doc.pages()]
def cmap_extraction_example():
pdf_filepath = './tutorial-example.pdf'
try:
fd = open(pdf_filepath, 'rb')
doc = PDFDocument(fd)
from itertools import islice
page = next(islice(doc.pages(), 2, 3))
print('page.Resources.Font = {}.'.format(page.Resources.Font))
print('len(page.Resources.Font) = {}.'.format(len(page.Resources.Font)))
font = page.Resources.Font['R26']
print('font.Subtype = {}, bool(font.ToUnicode) = {}.'.format(font.Subtype, bool(font.ToUnicode)))
# It is PostScript Type1 font, and texts use CMap provided by ToUnicode attribute.
# Font's ToUnicode attribute contains a reference to the CMap file data stream.
cmap = font.ToUnicode
print('type(cmap) = {}.'.format(type(cmap)))
print('cmap.Filter = {}.'.format(cmap.Filter))
data = cmap.filtered
with open('./sample-cmap.txt', 'wb') as fd2:
fd2.write(data)
finally:
fd.close()
def get_text_pypdf(DOI:str) -> str:
try:
"""gets the text from a given DOI"""
hostname = socket.gethostname()
path = pathlib.Path(__file__).parent.absolute()
name = hostname + str(DOI).replace("/", "") + ".pdf"
fp = Path(path / "pdfs" / name) # build filepath
url = "https://www.medrxiv.org/content/" + str(DOI) + "v1.full.pdf" # build url
response = requests.get(url)
fp.write_bytes(response.content) # save .pdf
fd = open(str(path) + "/pdfs/" + name, "rb") # open with pdfreader
doc = PDFDocument(fd)
all_pages = [p for p in doc.pages()] # get pages
viewer = SimplePDFViewer(fd) # use simple viwer
text = ""
for p in range(len(all_pages)): # for each page
viewer.navigate(p + 1) # nav to page
try:
viewer.render() # render -> clean and strip
text += (u"".join(viewer.canvas.strings).encode(sys.stdout.encoding, errors='replace').decode("windows-1252")) + '\n'
except OverflowError:
pass
fd.close()
return text.lower()
except Exception as e:
print(e, DOI)
return ""
def encrypted_and_password_protected_pdf_tutorial():
pdf_filepath = './encrypted-with-qwerty.pdf'
try:
fd = open(pdf_filepath, 'rb')
viewer = SimplePDFViewer(fd, password='******')
viewer.render()
text = ''.join(viewer.canvas.strings)
print('text = {}.'.format(text))
#--------------------
doc = PDFDocument(fd, password='******')
page_one = next(doc.pages())
print('page_one.Contents = {}.'.format(page_one.Contents))
#--------------------
try:
doc = PDFDocument(fd, password='******')
#viewer = SimplePDFViewer(fd, password='******')
except ValueError as ex:
print('ValueError raised: {}.'.format(ex))
finally:
fd.close()
def xobject_image_example():
pdf_filepath = './example-image-xobject.pdf'
try:
fd = open(pdf_filepath, 'rb')
doc = PDFDocument(fd)
# Extract XObject image.
page = next(doc.pages())
print('page.Resources.XObject = {}.'.format(page.Resources.XObject))
xobj = page.Resources.XObject['img0']
print('xobj.Type = {}, xobj.Subtype = {}.'.format(xobj.Type, xobj.Subtype))
pil_image = xobj.to_Pillow()
#pil_image.save('./extract-logo.png')
#--------------------
# Extract Images: a very simple way.
viewer = SimplePDFViewer(fd)
viewer.render()
all_page_images = viewer.canvas.images
if 'img0' in all_page_images:
img = all_page_images['img0']
print('img.Type = {}, img.Subtype = {}.'.format(img.Type, img.Subtype))
all_page_inline_images = viewer.canvas.inline_images
if all_page_inline_images:
img = all_page_inline_images[0]
print('img.Type = {}, img.Subtype = {}.'.format(img.Type, img.Subtype))
finally:
fd.close()
#--------------------
pdf_filepath = './tutorial-example.pdf'
try:
fd = open(pdf_filepath, 'rb')
viewer = SimplePDFViewer(fd)
# Extract image masks.
viewer.navigate(5)
viewer.render()
inline_images = viewer.canvas.inline_images
image_mask = next(img for img in inline_images if img.ImageMask)
pil_img = image_mask.to_Pillow()
#pil_img.save('./mask.png')
finally:
fd.close()
def init_cmb_from_pdf(month):
filename = FILE_PATH.format(str(month).zfill(2))
# logger.info(filename)
fd = open(filename, "rb")
doc = PDFDocument(fd)
all_pages = [p for p in doc.pages()]
# logger.info(len(all_pages))
viewer = SimplePDFViewer(fd)
records = []
for i in range(len(all_pages)):
viewer.navigate(i+1)
viewer.render()
records = np.append(records, viewer.canvas.strings[4:])
head = np.where(records == '记账日')[0][0]
tail = np.where(records == '本期还款总额')[0][-1]
records = records[head:tail]
# title_cn = records[:5]
# title_en = records[5:11]
records = records[11:]
column_cn = ['交易日' '交易摘要' '人民币金额' '卡号末四位' '记账日' '交易地金额']
column_en = ['transaction_date', 'transaction_description', 'transction_amount',
'card_number', 'bill_date', 'str_rmb']
# Data: ['' '掌上生活还款' '-3,011.49' '9978' '07/24' '-3,011.49']
df = pd.DataFrame(records.reshape(
[int(len(records)/6), 6]), columns=column_en)
df['type'] = 'cmb'
df['transaction_date'] = df['transaction_date'].apply(
lambda _: '2020/' + _)
df['transaction_date'] = pd.to_datetime(
df['transaction_date'], format="%Y/%m/%d", errors='coerce')
df['transction_amount'] = df['transction_amount'].apply(
lambda _: decimal_from_value(_))
df = df[['transaction_date', 'transction_amount',
'transaction_description', 'type']]
return df
def grade_document(document, verbose = False, point_flags = ('%','%') ) :
doc = PDFDocument( document )
viewer = SimplePDFViewer( document )
grade = 0
for page_number, page in enumerate( doc.pages() ) :
if verbose :
print('------------------')
print('Page:', page_number)
viewer.navigate( page_number + 1 )
viewer.render()
grade += grade_page( viewer.canvas, verbose = verbose, point_flags = point_flags )
return grade
def main():
# Get the PDF
r = requests.get(
"https://www.sos.arkansas.gov/uploads/elections/ARCountyClerks.pdf")
# Pass byte stream to PDFDocument parser (used for iterating through pages)
doc = PDFDocument(r.content)
# Pass byte stream to PDF viewer (used for reading strings on pages)
viewer = SimplePDFViewer(r.content)
navigate_pages(doc, viewer)
pprint(ELECTION_OFFICE_INFO)
def pdf_object_navigation_example():
pdf_filepath = './tutorial-example.pdf'
try:
fd = open(pdf_filepath, 'rb')
doc = PDFDocument(fd)
catalog = doc.root
print('catalog.Type = {}.'.format(catalog.Type))
print('catalog.Metadata.Type = {}, catalog.Metadata.Subtype = {}.'.format(catalog.Metadata.Type, catalog.Metadata.Subtype))
pages_tree_root = catalog.Pages
print('pages_tree_root.Type = {}.'.format(pages_tree_root.Type))
# Attribute names are cases sensitive.
# Missing or non-existing attributes have value of None.
print('(catalog.type is None) = {}.'.format(catalog.type is None))
print('(catalog.Metadata.subType is None) = {}.'.format(catalog.Metadata.subType is None))
print('(catalog.Metadata.UnkNown_AttriBute is None) = {}.'.format(catalog.Metadata.UnkNown_AttriBute is None))
# If object is an array, access its items by index.
first_page = pages_tree_root.Kids[0]
print('first_page.Type = {}.'.format(first_page.Type))
print('first_page.Contents.Length = {}.'.format(first_page.Contents.Length))
# If object is a stream, you can get either raw data (deflated in this example) or decoded content.
raw_data = first_page.Contents.stream
print('(first_page.Contents.Length == len(raw_data)) = {}.'.format(first_page.Contents.Length == len(raw_data)))
print('first_page.Contents.Filter = {}.'.format(first_page.Contents.Filter))
decoded_content = first_page.Contents.filtered
print('len(decoded_content) = {}.'.format(len(decoded_content)))
print('decoded_content.startswith(b"BT\n0 0 0 rg\n/GS0 gs") = {}.'.format(decoded_content.startswith(b'BT\n0 0 0 rg\n/GS0 gs')))
# On the file structure level all objects have unique number and generation to identify them.
num, gen = 2, 0
raw_obj = doc.locate_object(num, gen)
obj = doc.build(raw_obj)
print('obj.Type = {}.'.format(obj.Type))
finally:
fd.close()
def get_text(DOI: str) -> str:
txt = ""
name = "curr."
fp = Path(Path.cwd() / "pdfs" / "curr.pdf") # build filepath
url = "https://www.medrxiv.org/content/" + DOI + "v1.full.pdf" # build url
response = requests.get(url)
fp.write_bytes(response.content) # save .pdf
fd = open(r"pdfs\curr.pdf", "rb") # open with pdfreader
doc = PDFDocument(fd)
all_pages = [p for p in doc.pages()] # get pages
viewer = SimplePDFViewer(fd) # use simple viwer
for p in range(len(all_pages)): # for each page
viewer.navigate(p + 1) # nav to page
viewer.render() # render -? clean and strip
txt += (u"".join(viewer.canvas.strings).encode(
sys.stdout.encoding,
errors='replace').decode("windows-1252")) + '\n'
return txt
def font_extraction_example():
pdf_filepath = './example-font.pdf'
try:
fd = open(pdf_filepath, 'rb')
doc = PDFDocument(fd)
page = next(doc.pages())
print('sorted(page.Resources.Font.keys()) = {}.'.format(sorted(page.Resources.Font.keys())))
font = page.Resources.Font['T1_0']
print('font.Subtype = {}, font.BaseFont = {}, font.Encoding = {}.'.format(font.Subtype, font.BaseFont, font.Encoding))
font_file = font.FontDescriptor.FontFile
print('type(font_file) = {}.'.format(type(font_file)))
print('font_file.Filter = {}.'.format(font_file.Filter))
data = font_file.filtered
#with open('./sample-font.type1', 'wb') as fd2:
# fd2.write(data)
finally:
fd.close()
def GradeDocSafe(document, verbose=False, point_flags=defaut_point_flags):
'''
grade = GradeDoc(document, verbose = False, point_flags = defaut_point_flags )
'''
doc = PDFDocument(document)
viewer = SimplePDFViewer(document)
grade = 0
for page_number, page in enumerate(doc.pages()):
if verbose:
print('------------------')
print('Page:', page_number + 1)
viewer.navigate(page_number + 1)
viewer.render()
grade += GradePage(viewer.canvas.strings,
verbose=verbose,
point_flags=point_flags)
return grade
def navigate_pages(doc: PDFDocument, viewer: SimplePDFViewer):
for i, page in enumerate(doc.pages(), 1):
# navigate to page
viewer.navigate(i)
# render the page
viewer.render()
# collapse that ass
page_strings: List[str] = viewer.canvas.strings.copy()
merge_ranges = get_line_ranges(strings_list=page_strings)
page_strings = establish_uniformity(strings_list=page_strings,
line_range_list=merge_ranges)
get_county_election_office_info(strings_list=page_strings)
def uploaded_file():
if request.method == 'POST':
f = request.files['file']
filepath = os.path.join(app.config['UPLOAD_FOLDER'],
secure_filename(f.filename))
f.save(filepath)
fd = open(filepath, "rb")
doc = PDFDocument(fd)
version = doc.header.version
print(doc.metadata)
creationDate = doc.metadata.get('CreationDate')
dataType = doc.metadata.get('Subtype')
#data methods
viewer = SimplePDFViewer(fd)
textData = []
for canvas in viewer:
#print(canvas.strings)
textData += canvas.strings
tempstring = ''
textWords = []
for character in textData:
if character != ' ':
tempstring += character
else:
if tempstring:
textWords.append(tempstring)
tempstring = ''
print(secure_filename(f.filename))
print(creationDate)
print(textWords)
fileDocument = {
"name": secure_filename(f.filename),
"creationDate": creationDate,
"text": textWords
}
collection.insert_one(fileDocument)
return 'file uploaded successfully'
def document_tutorial():
pdf_filepath = './tutorial-example.pdf'
from io import BytesIO
with open(pdf_filepath, 'rb') as fd:
stream = BytesIO(fd.read())
doc = PDFDocument(stream)
try:
fd = open(pdf_filepath, 'rb')
doc = PDFDocument(fd)
print('doc.header.version = {}.'.format(doc.header.version))
print('doc.metadata = {}.'.format(doc.metadata))
print('doc.root.Type = {}.'.format(doc.root.Type))
print('doc.root.Metadata.Subtype = {}.'.format(doc.root.Metadata.Subtype))
print('doc.root.Outlines.First["Title"] = {}.'.format(doc.root.Outlines.First['Title']))
#--------------------
# Browse document pages.
page_one = next(doc.pages())
all_pages = [p for p in doc.pages()]
print('len(all_pages) = {}.'.format(len(all_pages)))
page_six = next(itertools.islice(doc.pages(), 5, 6))
page_five = next(itertools.islice(doc.pages(), 4, 5))
page_eight = all_pages[7]
print('page_six.MediaBox = {}.'.format(page_six.MediaBox))
print('page_six.Annots[0].Subj = {}.'.format(page_six.Annots[0].Subj))
print('page_six.Parent.Type = {}.'.format(page_six.Parent.Type))
print('page_six.Parent.Count = {}.'.format(page_six.Parent.Count))
print('len(page_six.Parent.Kids) = {}.'.format(len(page_six.Parent.Kids)))
finally:
fd.close()
# import tabula
# # df = read_pdf("Activity_Report.pdf")
# tabula.convert_into("Activity_Report.pdf", "output.csv", output_format="csv", pages='all')
# import pdftables_api
# c = pdftables_api.Client('r0tedshcbejj')
# c.xlsx('Acentria Activity Report.pdf', 'Acentria Activity Report.xlsx')
# c.xlsx('Integration/Tam_Weaver/RWI Policy Types Sample 0.pdf', 'Integration/Tam_Weaver/RWI_PolicyTypes_Sample0.xlsx')
import pdfreader
from pdfreader import PDFDocument, SimplePDFViewer
fd = open("Acentria Activity Report.pdf", "rb")
doc = PDFDocument(fd)
page = next(doc.pages())
print(doc.root)
# df = tabula.read_pdf('Acentria Activity Report.pdf', pages = 3, lattice = True)[1]
# import os
# import sys
# import pdftables_api
# from PyPDF2 import PdfFileWriter, PdfFileReader
# if len(sys.argv) < 3:
# command = os.path.basename(__file__)
# sys.exit('Usage: {} pdf-file page-number, ...'.format(command))
# pdf_input_file = sys.argv[1];
# pages_args = ",".join(sys.argv[2:]).replace(" ","")
# pages_required = [int(p) for p in filter(None, pages_args.split(","))]
elif centre_type:
print("TYPE 2")
centre_scraping_one_page(text_on_page, writer, county_name)
print("We are " + str((i / len(all_pages)) * 100) + "% done.")
elif chester_type:
print("TYPE 3")
ches_scraping_one_page(text_on_page, writer, county_name)
print("We are " + str((i / len(all_pages)) * 100) + "% done.")
else:
print("LAST TYPE")
beav_scraping_one_page(text_on_page, writer, county_name)
print("We are " + str((i / len(all_pages)) * 100) + "% done.")
#short function to grab the county name from the file name of the input pdf
def get_county(file_name):
for i in range(len(file_name)):
if file_name[i] == "_":
stop = i
break
return file_name[:stop]
if __name__ == "__main__":
file_name = "blair_results_2020.pdf"
fd = open(file_name, "rb")
doc = PDFDocument(fd)
viewer = SimplePDFViewer(fd)
county = get_county(file_name)
big_scrape(viewer, doc, county)
number_of_races += 1
#print(number_of_races)
if "PRESIDENT OF THE UNITED STATES" in text_on_page:
presidental_race(text_on_page, csv_writer, office, party)
elif number_of_races == 1:
single_race(text_on_page, csv_writer)
elif any('DELEGATE' in item for item in text_on_page):
pass
else:
double_race_page(text_on_page, csv_writer)
if __name__ == "__main__":
fd = open("bradford_results_2020.pdf", "rb")
doc = PDFDocument(fd)
viewer = SimplePDFViewer(fd)
parties = ["DEM", "REP", "NPA"]
offices = [
"PRESIDENT OF THE UNITED STATES", "ATTORNEY GENERAL",
"AUDITOR GENERAL", "STATE TREASURER"
]
presidential_candidates = [
"BERNIE SANDERS", "JOSEPH R. BIDEN", "TULSI GABBARD",
"DONALD TRUMP (W)", "Total", "Write-in", "DONALD J. TRUMP",
"ROQUE ROCKY DE LA FUENTE", "BILL WELD", "BERNIE SANDERS (W)"
]
all_pages = [p for p in doc.pages()]
with open('20200602__pa__primary__bradford__precinct.csv', 'w',
newline='') as csvfile:
writer = csv.writer(csvfile)
def main():
# Parse the arguments
parser = argparse.ArgumentParser()
parser.add_argument("file_name")
parser.add_argument("-o", "--output", default=None,
help="sets the output directory")
parser.add_argument("-v", "--verbose", default=False, type=str2bool,
const=True, nargs='?',
help="increase output verbosity")
parser.add_argument("-fp", "--first_page", default=0,
help="first page to extract from")
parser.add_argument("-lp", "--last_page", default=1000,
help="last page to extract from")
parser.add_argument("-mw", "--min_width", default=200,
help="minimum pixel width")
parser.add_argument("-mh", "--min_height", default=200,
help="minimum pixel height")
parser.add_argument("-xw", "--max_width", default=1210,
help="maximum pixel width")
parser.add_argument("-xh", "--max_height", default=1570,
help="maximum pixel height")
parser.add_argument("-mt", "--make_transparent", default=True,
type=str2bool, const=False, nargs='?',
help="flag to make the background transparent")
parser.add_argument("-wt", "--white_to_trans", default=True,
type=str2bool, const=False, nargs='?',
help="turn white pixels transparent")
parser.add_argument("-bt", "--black_to_trans", default=True,
type=str2bool, const=False, nargs='?',
help="turn black pixels transparent")
parser.add_argument("-wf", "--white_fuzz", default=1,
help="fuzz percent (0-100) for white transparency")
parser.add_argument("-bf", "--black_fuzz", default=1,
help="fuzz percent (0-100) for black transparency")
parser.add_argument("-ims", "--image_string", default="Im",
help="string that appears in all image names")
args = parser.parse_args()
if args.verbose:
print(f"Args:\n\t{args}")
# Obtain the base filename
file_name = args.file_name
assert os.path.exists(file_name)
assert file_name[-4:] == ".pdf", "must provide '.pdf' file"
base_file_name = file_name[:-4]
# Split on slashes
base_file_name = base_file_name.split("/")[-1]
base_file_name = base_file_name.split("\\")[-1]
assert len(base_file_name) > 0
# Make the output directory
if args.output is not None:
output = args.output
else:
output = base_file_name + "_images"
if args.verbose:
print(f"No output file given; outputing to {output}/")
os.makedirs(output, exist_ok=True)
# Import the pdfreader
fd = open(file_name, "rb")
doc = PDFDocument(fd)
# Check pages
assert args.first_page > -1
assert args.last_page > -1
assert args.last_page > args.first_page
# Loop over pages
for i, page in enumerate(doc.pages()):
if i < args.first_page:
continue
if i >= args.last_page:
exit()
if args.verbose:
nkeys = len(page.Resources.XObject.keys())
print(f"On page {i} -- {nkeys} XObjects detected")
# Loop over possible image objects
for key in page.Resources.XObject.keys():
if args.image_string in key or "im" in key:
xobj = page.Resources.XObject[key]
try:
pil_image = xobj.to_Pillow()
except IndexError:
if args.verbose:
print(
f"IndexError raised on page {i} {key} - skipping"
)
continue
width, height = pil_image.size
if width < args.max_width and height < args.max_height:
if width > args.min_width and height > args.min_height:
if args.verbose:
print(
f"Saving image {key} on page{i}: "+\
f"(w,h)={pil_image.size}"
)
pil_image.save(f"{output}/page{i}_{key}.png")
if args.make_transparent:
_do_transparent(args, i, key, pil_image, output)
return
glyph.attrib['name'] = ''
glyphs[ET.tostring(glyph).strip()] = name
shares = []
data = b''
glyphOrder = []
i = 0
prime = None
for line in f:
data += line
if line == b'%%EOF\n':
with TmpFile() as pdf, TmpFile() as fontfile, TmpFile() as ttxfile:
pdf.write(data)
# Convert font to xml and extract character map
font = next(PDFDocument(pdf).pages(
)).Resources.Font['F2+0'].FontDescriptor.FontFile2.filtered
fontfile.write(font)
ttx.ttDump(fontfile.name, ttxfile.name, ttx.Options('', 0))
scrambled = ET.parse(ttxfile.name).getroot()
cmap = scrambled.find('.//cmap/cmap_format_6')
# Match glyphs to their names based on their shape
for c in cmap.findall('map'):
code = int(c.attrib['code'], 16)
if 0x20 <= code <= 0x7e:
name = c.attrib['name']
scrambled_glyph = scrambled.find(
f'.//glyf/TTGlyph[@name="{name}"]')
scrambled_glyph.attrib['name'] = ''
glyphOrder.append(
glyphs[ET.tostring(scrambled_glyph).strip()])
def main():
print('Opening pdf and writing to decrypted copy')
p_pdf = open('AP12176A_20200701_142127.pdf', 'rb') #this will change
pdfReader = PyPDF2.PdfFileReader(p_pdf)
pdfWriter = PyPDF2.PdfFileWriter()
pdfReader.decrypt('')
#write all the pages in the unlocked file to a new pdf
print('Writing to decrypted copy.')
for pageNum in range(pdfReader.numPages):
pageObj = pdfReader.getPage(pageNum)
pdfWriter.addPage(pageObj)
u_name = input('File name for decrypted copy? (excluding ".pdf"): '
) #prompt for the name of the .txt output file
u_file = u_name + '.pdf'
#pdfOutputFile = open(u_file, 'wb')
print('Finishing writing to decrypted copy')
pdfOutputFile = open('temp_pdf.pdf', 'wb')
pdfWriter.write(pdfOutputFile) #write to the temporary unlocked pdf
pdfOutputFile.close()
p_pdf.close()
#open the unlocked pdf. We'll use PDFDocument and SimplePDFViewer to pull all the text
#_pdf = open(u_file, "rb")
u_pdf = open('temp_pdf.pdf', "rb")
doc = PDFDocument(u_pdf)
reader = SimplePDFViewer(u_pdf)
print('counting pages in pdf')
pgs = [p for p in doc.pages()] #count number of pages
page_ct = len(pgs)
print('cycling through pages')
with open(u_name + '.txt', 'w') as g:
for pg in range(page_ct): #cycle through pages
reader.navigate(pg + 1)
reader.render()
if (pg + 1) % 10 == 0:
print('processing page ' + str(pg + 1) + ' of ' + str(page_ct))
st = reader.canvas.strings #list with 1 line per element
for l in range(len(st)):
ln = st[l].encode('ascii', 'replace').decode(
'utf-8') #turn unknown chars into ?
g.write(ln + '\n')
group_exp = '^\s(.{8})\s(.{17})\s\s(.{8})\s\s(.{10})\s\s(.{21})\s(.{10})\s(.{18})\s(.{12})(.+)?$' #regex for grouping an invoice..
group_inv = re.compile(group_exp) #set as a regex
acct_exp = 'ACCOUNT.+(\d{6}).+DEPARTMENT\s+(\w{8})' #find acct and center
find_acct = re.compile(acct_exp)
vend_exp = '^([\w]{10})\s+(\S.+\S)\s+$' #find 10 'word' chars, then a space, then everything up to the line break
find_vendor = re.compile(vend_exp)
corp_exp = 'COMPANY\s(\w{4})\s+DATE' #find the 4 characters between COMPANY and DATE
find_corp = re.compile(corp_exp)
corp = 'XXXX'
center = 'XXXXXXXX'
account = 'XXXXXX'
v_short = 'XXXX'
v_long = 'XXXXXXXX'
data_tmp = []
ct = 0
with open(u_name + '.txt', 'r') as h:
for line in h:
ln = str(line) #remove leading/trailing spaces and newline chars
if find_corp.search(ln): #look for a new corp
if corp != find_corp.search(ln).group(1):
corp = find_corp.search(ln).group(1)
elif find_vendor.search(ln): #look for a new vendor
if v_short != str(find_vendor.search(ln).group(1)):
v_short = str(find_vendor.search(ln).group(1))
v_long = str(find_vendor.search(ln).group(2))
elif find_acct.search(ln): #look for a new acct/center
if center != find_acct.search(ln).group(
2) or account != find_acct.search(ln).group(1):
center = find_acct.search(ln).group(2)
account = find_acct.search(ln).group(1)
elif is_inv(ln): #look for an invoice
tmp = group_inv.search(ln) #print(is_inv(ln).groups())
gl_eff = tmp.group(1).strip()
inv_num = tmp.group(2).strip()
inv_date = tmp.group(3).strip()
po = tmp.group(4).strip()
desc = tmp.group(5).strip()
q = tmp.group(6).strip()
if q == '':
qty = 0
else: #qty
qty = q.replace(',', '')
prod_id = tmp.group(7).strip()
if len(tmp.group(8).strip()) != 0: #expense
exp = tmp.group(8).strip()
exp = float(str(exp).replace(',', ''))
else:
exp = 0
if len(tmp.group(9).strip()) != 0: #expense
cred = tmp.group(9).strip()
cred = float(str(cred).replace(',', ''))
else:
cred = 0
if int(qty) == 0:
per_unit = 0
else:
per_unit = round(float(exp) / float(qty), 2)
new_row = [
corp, center, account, v_short, v_long, gl_eff, inv_num,
inv_date, po, desc, qty, prod_id, exp, cred, per_unit
]
data_tmp.append(new_row)
ct += 1
if ct % 1000 == 0:
print('Finished adding row ' + str(ct))
data_cols = [
'Company', 'Center', 'Account', 'Vendor_Short', 'Vendor_Long',
'GL_Effective_Date', 'Inv_Number', 'Inv_Date', 'PO', 'Description',
'Qty', 'ProdID', 'Expense', 'Credit', 'Per_Unit_Cost'
]
col_widths = [14, 12, 12, 18, 35, 22, 20, 13, 12, 31, 10, 20, 15, 15, 18]
data_inv = pd.DataFrame(data=data_tmp, columns=data_cols)
data_inv['GL_Effective_Date'] = pd.to_datetime(
data_inv['GL_Effective_Date'])
data_inv['Inv_Date'] = pd.to_datetime(data_inv['Inv_Date'])
data_inv['Account'] = data_inv['Account'].astype('int64')
data_inv['Qty'] = data_inv['Qty'].astype('int64')
i_rows = data_inv['Company'].size
with pd.ExcelWriter(u_name + '.xlsx',
engine='xlsxwriter',
datetime_format='m/d/yyyy') as writer:
data_inv.to_excel(writer, sheet_name='DATA', index=False)
workbook = writer.book
worksheet = writer.sheets['DATA']
curr_format = workbook.add_format(
{'num_format': '$#,##0.00;[Red]($#,##0.00)'})
worksheet.set_column(12, 12, 13,
curr_format) #first col, last col, width, format
worksheet.set_column(13, 13, 13, curr_format)
worksheet.autofilter('A1:O' + str(i_rows + 1))
worksheet.freeze_panes(1, 0) #freeze 1st row
for a in range(len(col_widths)):
worksheet.set_column(a, a, col_widths[a])
print('DONE - pulled ' + str(i_rows) + ' lines into ' + u_name + '.xlsx')
src_dir = '~/Olds' #源文件目录地址
des_dir = '~/News' #新文件目录地址
num = 0
if not os.path.exists(des_dir): #如果没有目标文件夹,新建一个目标文件夹进行存储
os.makedirs(des_dir)
if os.path.exists(src_dir):
dirs = os.listdir(src_dir) #获取源文件的目录地址
print(dirs)
for dirc in dirs: #对于目录下的每一个文件
if "pdf" not in dirc:
continue
fd = open(os.path.join(src_dir, dirc), 'rb')
doc = PDFDocument(fd) #打开并建立一个PDF文件对象
viewer = SimplePDFViewer(fd)
reader = PyPDF2.PdfFileReader(fd)
print(reader.documentInfo)
#print(reader.getPage(0).extractText())
#paper_title = pdf_reader.getDocumentInfo() #获取PDF标题
viewer.render()
#print(doc.root)
#print(viewer.canvas.text_content)
#print("num : %s" % num , doc) #终端显示处理到第几个文件
# num += 1
# paper_title = str(paper_title) #标题字符化
# if paper_title.find('/') != -1: #对于'/'无法写入文件名的情况,将其用'_'代替
# new_paper_title = paper_title.replace('/','_')
# paper_title = new_paper_title
def read_file(path_to_file):
fd = open(path_to_file, "rb")
doc = PDFDocument(fd)
viewer = SimplePDFViewer(fd)
viewer.render()
return viewer
import pdfreader
from pdfreader import PDFDocument, SimplePDFViewer
from functions import cleanString
from ItemsClass import Item
from datetime import datetime
import json
import re
fd = open("order1.pdf", "rb")
doc = PDFDocument(fd)
numpages = len([p for p in doc.pages()])
viewer = SimplePDFViewer(fd)
strings = []
for num in range(0, numpages):
viewer.navigate(num + 1)
viewer.render()
strings += viewer.canvas.strings[4:]
print(strings)
prev = ""
Items = []
item = Item()
Order = {
"Items": [],
"Request": "",
"Total": "",
"Customer": "",
"Delivery": ""
}
requesting = False
def main():
"""Docstring will go here"""
print('*' * 40 + '\nERT PDF TO TXT CONVERTER\n')
print('Showing PDF files in ' + os.getcwd())
files = []
f_count = 0
for file in os.listdir():
if file.endswith(".pdf") or file.endswith(".PDF"):
f_name = file.rsplit('.', maxsplit=1)[0]
print('(' + str(f_count) + ') ' + file)
files.append(f_name) #take the file name w/o the .pdf
f_count += 1
prompt = 'Enter the number corresponding to the target pdf, or q to quit: '
choice = input(prompt)
while is_valid(choice, f_count) == False:
if choice == 'q':
print('Quitting.')
sys.exit()
else:
print('Invalid choice - try again.')
choice = input(prompt)
if choice != 'q':
pdf_to_open = str(files[int(choice)]) + '.pdf'
print(pdf_to_open)
#open the protected pdf and remove the password
print('Converting to unlocked PDF')
p_pdf = open(pdf_to_open, 'rb') #this will change
pdfReader = PyPDF2.PdfFileReader(p_pdf)
pdfReader.decrypt('')
pdfWriter = PyPDF2.PdfFileWriter()
#write all the pages in the unlocked file to a new pdf
for pageNum in range(pdfReader.numPages):
pageObj = pdfReader.getPage(pageNum)
pdfWriter.addPage(pageObj)
u_name = input('Output file name? (excluding ".txt"): '
) #prompt for the name of the new unlocked pdf
#print(Path(u_name + '.txt').is_file())
#sys.exit()
pdfOutputFile = open('temp_pdf.pdf', 'wb')
pdfWriter.write(pdfOutputFile)
pdfOutputFile.close()
p_pdf.close()
#open the unlocked pdf. We'll use PDFDocument and SimplePDFViewer to pull all the text
u_pdf = open('temp_pdf.pdf', "rb")
doc = PDFDocument(u_pdf)
reader = SimplePDFViewer(u_pdf)
start_time = time.time()
pgs = [p for p in doc.pages()] #count number of pages
page_ct = len(pgs)
print('Writing ' + str(page_ct) + ' pages to ' + u_name + '.txt ...')
with open(u_name + '.txt', 'w') as g:
for pg in range(page_ct): #cycle through pages
reader.navigate(pg + 1)
reader.render()
if (pg + 1) % 10 == 0:
print('Processing page ' + str(pg + 1) + ' of ' + str(page_ct))
st = reader.canvas.strings #list with 1 line per element
for l in range(len(st)):
ln = st[l].encode('ascii',
'replace') #turn unknown chars into ?
ln = ln.decode('ascii', 'strict')
g.write(ln + '\n')
u_pdf.close()
os.remove('temp_pdf.pdf')
print('Saved as ' + u_name + '.txt')
print("This took %s seconds." % round((time.time() - start_time), 2))
async def get_election_offices():
async with aiohttp.ClientSession() as session:
async with session.get(URL) as r:
text = await r.read()
# Prep helper vars
phone, office_supervisor, website, location_name, county_name = ("", ) * 5
doc = PDFDocument(text)
viewer = SimplePDFViewer(text)
physical_address, mailing_address = ({}, ) * 2
election_offices = []
for i, page in enumerate(doc.pages(), 1):
viewer.navigate(i)
viewer.render()
# This is parsed in the order at which pdf elements are read by the viewer.
for j, s in enumerate(viewer.canvas.strings):
if not county_name:
m = re.search(r"\D+(?=\s-)", s)
if m:
county_name = m.group(0).split(maxsplit=1)[0].capitalize()
location_name = f"{county_name} Election Office"
mapping = electionsaver.addressSchemaMapping
if not physical_address:
m = re.search(r"(?<=MUNICIPAL ADDRESS :).*", s)
if m:
physical_address = usaddress.tag(
f"{m.group(0)} {viewer.canvas.strings[j + 1]}".title(),
tag_mapping=mapping,
)[0]
physical_address["state"].upper()
physical_address["locationName"] = location_name
if not mailing_address:
m = re.search(r"(?<=MAILING ADDRESS :).*", s)
if m:
mailing_address = usaddress.tag(
f"{m.group(0)} {viewer.canvas.strings[j + 1]}".title(),
tag_mapping=mapping,
)[0]
mailing_address["state"].upper()
mailing_address["locationName"] = location_name
if not phone:
m = re.search(r"(?<=Phone 1: ).*", s)
if m:
phone = m.group(0)
election_offices.append({
"countyName": county_name,
"physicalAddress": physical_address,
"mailingAddress": mailing_address,
"phone": phone,
"officeSupervisor": office_supervisor,
"supervisorTitle": "County Clerk",
"website": website,
})
# reset for next round
phone, office_supervisor, website, location_name, county_name = (
"", ) * 5
if not office_supervisor:
m = re.search(r"(?<=COUNTY CLERK: ).*", s)
if m:
office_supervisor = m.group(0).title()
if not website:
m = re.search(r"http.*", s)
if m:
website = m.group(0)
with open(
os.path.join(ROOT_DIR, "scrapers", "wisconsin", "wisconsin.json"),
"w") as f:
json.dump(election_offices, f)
return election_offices
