pdftkコマンドの`dump_data_utf8`オプションの出力ファイルから目次（Bookmark）の情報を抽出してページ範囲を出力するスクリプト

parse_pdftk_output.py

#! /usr/local/bin/python3
 # coding: utf-8

import re
import pathlib


def parse_page_range(filename, output_path=pathlib.Path('.'), pdf_filename=None):
    '''pdftkコマンドの`dump_data_utf8`オプションの出力ファイルから目次（Bookmark）の情報を抽出してページ範囲を出力する'''


    fp = pathlib.Path(filename).open('r')  # pdftkコマンドの出力ファイルを読み込む



    entry = { 'title': "", 'page_no': None}
    last_page = 0

    bk_list = []

    for line in fp:

        if line.startswith("NumberOfPages"):
            last_page = int(line.rstrip("\n").split(': ')[1])

        if line.startswith("PageMediaBegin"):
            bk_list.append(entry)
            break

        if not line.startswith("Bookmark") :
            continue

        else:

            if line.startswith("BookmarkTitle"):
                entry['title'] = line.rstrip("\n").split(': ')[1]
            if line.startswith("BookmarkPageNumber") :
                entry['page_no'] = int(line.rstrip("\n").split(': ')[1])

            if line.startswith("BookmarkBegin") :
                if len(entry['title']) > 0:
                    bk_list.append(entry)
                    entry = { 'title': "", 'page_no': None}

            if line.startswith("PageMediaBegin"):
                bk_list.append(entry)
                break

    fp.close()


    filtered_list = [ e for e in bk_list if e['page_no'] > 0  ]

    sorted_list = filtered_list

    sorted_list.sort(key=lambda x: x['page_no'])

    page_index_list = []

    for i, e in enumerate(sorted_list):
        end_page = last_page
        try:
            # print(i)
            n = sorted_list[i+1]
            next_article_page = n['page_no']

            end_page = next_article_page - 1
        except:
            pass

        index_entry = [pdf_filename, e['title'], str(e['page_no']),str(end_page)]
        #print(index_entry)

        page_index_list.append(index_entry)
        #print(pdf_filename, e['title'], e['page_no'],end_page, sep="\t")


        #return bk_list


    tsvname = 'index_' + pdf_filename.split('.')[0] + '.tsv'

    output = output_path / tsvname

    with output.open(mode='wt') as fp :

        for n in page_index_list :
            temp = "\t".join(n) + "\n"
            fp.write(temp)

    return 0


if __name__ == '__main__' :

    pl = pathlib.Path('./data')

    output_dir = pl / 'page_index'
    output_dir.mkdir(parents=True, exist_ok=True)

    dump_files = [p for p in pl.glob('output_log/dump_*.txt')]

    for f in dump_files:

        filename = str(f)
        pdf_filename = re.search('dump_(.+).txt', filename)[1] + '.pdf'

        parse_page_range(filename, output_path=output_dir, pdf_filename=pdf_filename)