embee/pyaccuwage
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

getting closer, added a FIXME to one of the fields. Having issues with columns in description fields

Browse source
This commit is contained in:
Binh 2012-07-17 15:44:28 -05:00
parent 5dde3be536
commit 40fcbdc8b8
3 changed files with 36 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

11
scripts/pyaccuwage-pdfparse
View file

@ -4,6 +4,7 @@ from pyaccuwage.pdfextract import PDFRecordFinder
import argparse
import sys
import os
import re
parser = argparse.ArgumentParser(description="Parse and convert contents of IRS files into pyaccuwage e-file classes.")
parser.add_argument("-i", "--input", nargs=1, required=True, metavar="file", type=argparse.FileType('r'), help="Source PDF file, ie: p1220.pdf")
@ -30,8 +31,12 @@ source_file = os.path.abspath(args.input[0].name)
doc = PDFRecordFinder(source_file)
records = doc.records()
builder = RecordBuilder()
for rec in records:
print 'name', rec[0]
for field in builder.load(map(lambda x:x.tuple, rec[1][1:])):
print field
sys.stdout.write("class %s(object):\n" % re.sub('[^\w]','',rec[0]))
for field in builder.load(map(lambda x:x.tuple, rec[1][1:])):
sys.stdout.write('\t' + field + '\n')
#print field