Python回复sub原始字符串前缀用H替换10，用P替换20

Question

我正在我的应用程序中构建一些模板，其中包含如下字符串：

templateString = r'{title} {- author} ({timestamp})'

并替换模板{}中相应的字符串（如果存在）：

# 10 things on my todo list - Brandon (2018-01-01 00:00:01)

我想在模板字符串中允许自定义字符，这样用户就可以连字符、方括号或其他任何内容，但前提是这些属性存在，例如，如果author为空，则不希望看到：

# 10 things on my todo list -  (2018-01-01 00:00:01)

你想看看：

# 10 things on my todo list (2018-01-01 00:00:01)

为此，我尝试使用捕获组来获取存在于{（标识词）和}之间的任何文本：

titleExp = re.compile(r'\{([^\{]*)title([^\}]*)\}', re.I)
authorExp = re.compile(r'\{([^\{]*)author([^\}]*)\}', re.I)
timestampExp = re.compile(r'\{([^\{]*)timestamp([^\}]*)\}', re.I)

最奇怪的是，当我尝试用原始字符串替换时，r'\1{}\2'，而不是“我的待办事项列表上的10件事”，我得到了“我的待办事项列表上的H件事”：

templateString = r'{title} {- author} ({timestamp})'
self.title = "10 things on my todo list"
renamed = re.sub(titleExp, (r'\1{}\2' if self.title else '').format(self.title or ''), renamed)
# H things on my todo list ...

我当然试过了，但没有使用原始字符串：

templateString = r'{title} {- author} ({timestamp})'
self.title = "10 things on my todo list"
renamed = re.sub(titleExp, ('\\1{}\\2' if self.title else '').format(self.title or ''), renamed)
# 10 things on my todo list ...

但同样的事情也发生了。你知道吗

怎么回事？为什么原始的弦会乱作一团？我可以看出这与数字有关，很可能是捕获组的行为不正常。你知道吗

---

端到端复制：

templateString = r'{title} {- author} ({timestamp})'

titleExp = re.compile(r'\{([^\{]*)title([^\}]*)\}', re.I)
authorExp = re.compile(r'\{([^\{]*)author([^\}]*)\}', re.I)
timestampExp = re.compile(r'\{([^\{]*)timestamp([^\}]*)\}', re.I)

title = "10 things on my todo list"
author = "Brandon"
timestamp = "2018-01-01 00:00:01"

templateString = re.sub(titleExp, r'\1{}\2'.format(title), templateString)
templateString = re.sub(authorExp, r'\1{}\2'.format(author), templateString)
templateString = re.sub(timestampExp, r'\1{}\2'.format(timestamp), templateString)

print(templateString)

# output:
# H things on my todo list - Brandon (P18-01-01 00:00:01)
# ^ ??                                ^ ??

# expected:
# 10 things on my todo list - Brandon (2018-01-01 00:00:01)

更多研究：

它似乎与替换字符串的第一个字符有关：

title = " 10 things on my todo list"
#.       ^ space
author = "Brandon"
timestamp = " 2018-01-01 00:00:01"
#.       ^ space

修复它。。。某种程度上。。。你知道吗

Answer 1

值得一提的是，如果我在不使用内联捕获组的情况下分解表达式，它的行为是正确的。我现在可以用这个作为解决办法，但我肯定会喜欢一个解释为什么。。。你知道吗

templateString = r'{title} {- author} ({timestamp})'

titleExp = re.compile(r'\{([^\{]*)title([^\}]*)\}', re.I)
authorExp = re.compile(r'\{([^\{]*)author([^\}]*)\}', re.I)
timestampExp = re.compile(r'\{([^\{]*)timestamp([^\}]*)\}', re.I)

title = "10 things on my todo list"
author = "Brandon"
timestamp = "2018-01-01 00:00:01"

match = re.search(titleExp, templateString)
title = '{}{}{}'.format(match.groups()[0], title, match.groups()[1])
templateString = re.sub(titleExp, title, templateString)

match = re.search(authorExp, templateString)
author = '{}{}{}'.format(match.groups()[0], author, match.groups()[1])
templateString = re.sub(authorExp, author, templateString)

match = re.search(timestampExp, templateString)
timestamp = '{}{}{}'.format(match.groups()[0], timestamp, match.groups()[1])
templateString = re.sub(timestampExp, timestamp, templateString)

print templateString

# output:
# 10 things on my todo list - Brandon (2018-01-01 00:00:01)

Answer 2

您可以将"-"放在{author}占位符之外，然后使用re.sub：

import re
templateString = r'{title} - {author} ({timestamp})'
title = "10 things on my todo list"
author = "Brandon"
timestamp = "2018-01-01 00:00:01"
new_data = re.sub('-\s(?=\{author)', '', templateString).format(title=title, author=author, timestamp = timestamp) if not author else templateString.format(title=title, author=author, timestamp = timestamp)
print(new_data)

输出：

10 things on my todo list - Brandon (2018-01-01 00:00:01)

当author为空时：

title = "10 things on my todo list"
author = ""
timestamp = "2018-01-01 00:00:01"
new_data = re.sub('-\s(?=\{author)', '', templateString).format(title=title, author=author, timestamp = timestamp) if not author else templateString.format(title=title, author=author, timestamp = timestamp)
print(new_data)

输出：

10 things on my todo list  (2018-01-01 00:00:01)