如何合并 python 中的多个字幕文件？

Question

我有一批文件“srt”要合并

sub1.srt

1
00:00:21,601 --> 00:00:24,130
- What happened? - It's a mess, I heard.

2
00:00:24,131 --> 00:00:25,900
- What's that? - Dead bodies?

3
00:00:25,901 --> 00:00:28,839
- What's going on? - I wish I knew.

sub2.srt

1
00:00:28,840 --> 00:00:31,310
No one knows. They won't say.

2
00:00:31,311 --> 00:00:35,276
- My gosh. - How can so many die?

3
00:00:45,191 --> 00:00:46,556
When you starve,

合并后

1
00:00:21,601 --> 00:00:24,130
- What happened? - It's a mess, I heard.

2
00:00:24,131 --> 00:00:25,900
- What's that? - Dead bodies?

3
00:00:25,901 --> 00:00:28,839
- What's going on? - I wish I knew.

4
00:00:28,840 --> 00:00:31,310
No one knows. They won't say.

5
00:00:31,311 --> 00:00:35,276
- My gosh. - How can so many die?

6
00:00:45,191 --> 00:00:46,556
When you starve,

我发现这个脚本可以工作问题出在数字上字幕顺序不对

filenames = ['sub1.srt', 'sub2.srt']
with open('output_file.srt', 'w') as outfile:
    for fname in filenames:
        with open(fname) as infile:
            for line in infile:
                outfile.write(line)

显示无序

如何修复？

Answer 1

您可以尝试以下方法：

import re

re_sub_no = re.compile(r"^\s*\d+\s*$", re.MULTILINE)

def repl(match):
    global sub_no
    sub_no += 1
    return str(sub_no)

sub_no = 0
filenames = ["sub1.srt", "sub2.srt"]
with open("sub_merged.srt", "w") as fout:
    for name in filenames:
        with open(name, "r") as fin:
            fout.write(re_sub_no.sub(repl, fin.read()) + "\n\n")

正则表达式模式 re_sub_no 正在搜索子编号，并通过 re.sub repl 函数确保编号一致。 sub_no 变量被设为 global 是因为函数本身无法存储编号的当前状态。模式的 \s* 部分只是在空白 before/after 数字的情况下的预防措施，也许你不需要它们（没有尝试）。

上面的代码没有解决文件的编码问题。当您运行为您链接到的文件执行以下操作时

import chardet

for name in ["sub1.srt", "sub2.srt"]:
    with open(name, "br") as file:
        print(chardet.detect(file.read()))

结果很可能是

{'encoding': 'UTF-8-SIG', 'confidence': 1.0, 'language': ''}
{'encoding': 'UTF-8-SIG', 'confidence': 1.0, 'language': ''}

因此，以下修改应该会奏效：

...
with open("sub_merged.srt", "w", encoding="utf-8-sig") as fout:
    for name in filenames:
        with open(name, "r", encoding="utf-8-sig") as fin:
            fout.write(re_sub_no.sub(repl, fin.read()) + "\n\n")

为了将来使用，您可以将其组合成：

import chardet, re

filenames = ["sub1.srt", "sub2.srt"]
encodings = []
for name in filenames:
    with open(name, "br") as file:
        encodings.append(chardet.detect(file.read())["encoding"])

re_sub_no = re.compile(r"^\d+\s*$", re.MULTILINE)

def repl(match):
    global sub_no
    sub_no += 1
    return str(sub_no)

sub_no = 0
with open("sub_merged.srt", "w", encoding=encodings[0]) as fout:
    for name, encoding in zip(filenames, encodings):
        with open(name, "r", encoding=encoding) as fin:
            fout.write(re_sub_no.sub(repl, fin.read()) + "\n\n")

Answer 2

此脚本会将第二个 SRT 文件中的字幕附加到第一个 SRT 文件中最接近的现有时间。

需要 srt 模块，需要与 pip3 install srt 一起安装。

用法

# by default output to subs1_MERGED.srt
python3 srt_merge.py subs1.srt subs2.srt

# output to custom file name final.srt
python3 srt_merge.py subs1.srt subs2.srt -o final.srt

#!/usr/bin/env python3

import argparse
import sys
from datetime import timedelta
from pathlib import Path

# REQUIRED MODULE: pip3 install srt
import srt


def nearest(items, pivot):
    return min(items, key=lambda x: abs(x.start - pivot))


if __name__ == '__main__':

    parser = argparse.ArgumentParser(description='merge SRT subtitles',
                                     usage="""
    Merge SRT subtitles:                                     
    \t{0} first.srt second.srt -o merged.srt
 """.format(Path(sys.argv[0]).name))

    parser.add_argument('srt1',
                        metavar='srt1',
                        help='SRT-file-1')
    parser.add_argument('srt2',
                        metavar='srt2',
                        help='SRT-file-2')
    parser.add_argument('--output-file', '-o',
                        default=None,
                        help='Output filename')
    parser.add_argument('--encoding', '-e',
                        default=None,
                        help='Input file encoding')
    args = parser.parse_args(sys.argv[1:])

    srt1_path = Path(args.srt1)
    srt2_path = Path(args.srt2)

    with srt1_path.open(encoding=args.encoding or 'utf-8') as fi1:
        subs1 = {s.index: s for s in srt.parse(fi1)}

    with srt2_path.open(encoding=args.encoding or 'utf-8') as fi2:
        subs2 = {s.index: s for s in srt.parse(fi2)}

    # iterate all subs in srt2 and find the closest EXISTING slot in srt1
    sub: srt.Subtitle
    start: int
    for idx, sub in subs2.items():

        start: timedelta = sub.start
        sub_nearest_slot: srt.Subtitle = nearest(subs1.values(), start)
        sub_nearest_slot.content = f'{sub_nearest_slot.content}<br>{sub.content}'
        subs1[sub_nearest_slot.index] = sub_nearest_slot

    if not args.output_file:
        generated_srt = srt1_path.parent / (f'{srt1_path.stem}_MERGED_{srt1_path.suffix}')
    else:
        generated_srt = Path(args.output_file)

    with generated_srt.open(mode='w', encoding='utf-8') as fout:
        fout.write(srt.compose(list(subs1.values())))

如何合并 python 中的多个字幕文件？

how to merge multiple subtitle files in python?

python

merge

用法