以id作为键读取多个标签为一个或多个元组，即{id:(cat1，cat2，.)}

Question

我正在建模一个多标签文本分类算法。下面是我的labels.txt文件的一个片段，我想把这些记录转换成一个字典，字典中包含有相应类别的元组或列表，即{ id :(cat1，cat2)}。记录不是分开的新行。我在如何将这种数据转换成字典上陷入困境。

B0027DQHA0
  Movies & TV, TV
  Music, Classical
0756400120
  Books, Literature & Fiction, Anthologies & Literary Collections, General
  Books, Literature & Fiction, United States
  Books, Science Fiction & Fantasy, Science Fiction, Anthologies
  Books, Science Fiction & Fantasy, Science Fiction, Short Stories
B0000012D5
  Music, Blues
  Music, Pop
  Music, R&B

Answer 1

如果类别名称总是缩进空格，而ID不缩进，则可以使用它来区分它们，并将类别名称附加到循环中由ID索引的dict中的列表中：

r = '''B0027DQHA0
  Movies & TV, TV
  Music, Classical
0756400120
  Books, Literature & Fiction, Anthologies & Literary Collections, General
  Books, Literature & Fiction, United States
  Books, Science Fiction & Fantasy, Science Fiction, Anthologies
  Books, Science Fiction & Fantasy, Science Fiction, Short Stories
B0000012D5
  Music, Blues
  Music, Pop
  Music, R&B'''
d = {}
for l in r.splitlines():
    if l.startswith(' '):
        d.setdefault(i, []).append(l.lstrip())
    else:
        i = l
print(d)

这一产出如下：

{'B0027DQHA0': ['Movies & TV, TV', 'Music, Classical'], '0756400120': ['Books, Literature & Fiction, Anthologies & Literary Collections, General', 'Books, Literature & Fiction, United States', 'Books, Science Fiction & Fantasy, Science Fiction, Anthologies', 'Books, Science Fiction & Fantasy, Science Fiction, Short Stories'], 'B0000012D5': ['Music, Blues', 'Music, Pop', 'Music, R&B']}