import re
import sys

from os.path import exists, splitext
from pypdf import PdfReader, PdfWriter


def _get_parent_bookmark(current_indent, history_indent, bookmarks):
    """The parent of A is the nearest bookmark whose indent is smaller than A's"""
    assert len(history_indent) == len(bookmarks)
    if current_indent == 0:
        return None
    for i in range(len(history_indent) - 1, -1, -1):
        # len(history_indent) - 1   ===>   0
        if history_indent[i] < current_indent:
            return bookmarks[i]
    return None


def addBookmark(pdf_path, bookmark_txt_path, page_offset):
    if not exists(pdf_path):
        return "Error: No such file: {}".format(pdf_path)
    if not exists(bookmark_txt_path):
        return "Error: No such file: {}".format(bookmark_txt_path)

    with open(bookmark_txt_path, "r", encoding="utf-8") as f:
        bookmark_lines = f.readlines()
    reader = PdfReader(pdf_path)
    writer = PdfWriter()
    writer.clone_document_from_reader(reader)

    maxPages = len(reader.pages)
    bookmarks, history_indent = [], []
    # decide the level of each bookmark according to the relative indent size in each line
    #   no indent:          level 1
    #     small indent:     level 2
    #       larger indent:  level 3
    #   ...
    for line in bookmark_lines:
        line2 = re.split(r"\s+", line.strip())
        if len(line2) == 1:
            continue

        indent_size = len(line) - len(line.lstrip())
        parent = _get_parent_bookmark(indent_size, history_indent, bookmarks)
        print(parent)
        history_indent.append(indent_size)

        title, page = " ".join(line2[:-1]), int(line2[-1]) - 1
        if page + page_offset >= maxPages:
            return "Error: page index out of range: %d >= %d" % (page + page_offset, maxPages)
        new_bookmark = writer.add_outline_item(title, page + page_offset, parent=parent)
        # AttributeError: 'DictionaryObject' object has no attribute 'insert_child'
        print("new_bookmark")
        bookmarks.append(new_bookmark)

    out_path = splitext(pdf_path)[0] + "-new.pdf"
    with open(out_path, "wb") as f:
        writer.write(f)

    return "The bookmarks have been added to %s" % out_path


if __name__ == "__main__":
    import sys

    args = sys.argv
    if len(args) != 4:
        print("Usage: %s [pdf] [bookmark_txt] [page_offset]" % args[0])
    else:
        print(addBookmark(args[1], args[2], int(args[3])))

# python bookmarks.py "/path/to/xxx.pdf" "/path/to/toc.txt" page_offset