55 lines
1.6 KiB
Python
55 lines
1.6 KiB
Python
import debug
|
|
import cluster
|
|
import config
|
|
import pdfplumber
|
|
from multiprocessing.pool import ThreadPool
|
|
import queue
|
|
|
|
def main():
|
|
de_column_ify()
|
|
|
|
def log(*args):
|
|
print(*args, flush=True)
|
|
|
|
def de_column_ify():
|
|
q = queue.Queue(maxsize=4)
|
|
with pdfplumber.open(config.INPUT) as pdf:
|
|
with ThreadPool(4) as pool:
|
|
for i in range(len(pdf.pages)):
|
|
pool.apply_async(de_columnify_page, (q, pdf.pages[i], ))
|
|
for i in range(len(pdf.pages)):
|
|
log("getting", i, "of", len(pdf.pages))
|
|
got = q.get()
|
|
for got_i in got[1]:
|
|
if got_i.chars:
|
|
True or debug.debug_show(debug.debug_im(got_i.page))
|
|
textify(got_i.page)
|
|
|
|
def de_columnify_page(q, page):
|
|
result = cluster.Chars(page.chars, page).divide_into_columns()
|
|
log("putting", page.page_number)
|
|
q.put((page, result))
|
|
|
|
def textify(page):
|
|
lines = page.extract_text(layout=True).split("\n")
|
|
leading_spaces = []
|
|
for line in lines:
|
|
leading_spaces.append(len(line) - len(line.lstrip()))
|
|
median_leading_spaces = sorted(leading_spaces)[len(leading_spaces)//2]
|
|
|
|
paragraphs = []
|
|
for line in lines:
|
|
if line.startswith(median_leading_spaces * " "):
|
|
line = line[median_leading_spaces:]
|
|
if line.startswith(" "):
|
|
line = "\t" + line[1:]
|
|
if not paragraphs or line.startswith("\t"):
|
|
paragraphs.append([])
|
|
paragraphs.append([])
|
|
paragraphs[-1].append(line.rstrip())
|
|
for paragraph in paragraphs:
|
|
print(" ".join(paragraph))
|
|
|
|
if __name__ == "__main__":
|
|
main()
|