merge.py

merge.py

import os
import re
import sys
from PyPDF2 import PdfFileMerger
import re

def extractNum(string):
    nums = re.findall("\d+", string)
    if len(nums) == 0:
        return 0
    return int(nums[0])

dirNames = [d for d in os.listdir("./") if os.path.isdir(d)]
dirNames.sort(key=extractNum)

studentPdfListDict = {}
for d in dirNames:
    files = [f for f in os.listdir(d) if f.endswith('.pdf')]
    for f in files:
        if f not in studentPdfListDict:
            studentPdfListDict[f] = []
        studentPdfListDict[f].append(os.path.join(d, f))

os.makedirs('./merged', exist_ok=True)

for (student, pdfList) in studentPdfListDict.items():
    merge = PdfFileMerger()
    print(f'{student} 학생 {len(pdfList)}개 파일 병합중...')
    for pdf in pdfList:
        merge.append(pdf)
    merge.write(os.path.join('./merged', student))
    merge.close()