如何创建以下集合的字符类
+ - * / % < > = ! & ^ | ? :
我已经尝试过：
[+-*/%<>=!&^|?:]
但我认为其中一些需要转义。我怎么知道是哪些？
+ - * / % < > = ! & ^ | ? :
I've tried:
[+-*/%<>=!&^|?:]
But I think some of them will need to be escaped. How can I tell which ones?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（4）
您需要对
-进行转义，否则它会被解释为+和*之间的字符范围，这是无效的：据我所知，其他元字符在字符类中按字面意思理解。
You'll need to escape the
-as otherwise it'd be interpreted as a character range between+and*, which isn't valid:The other metacharacters, as far as I know, are taken literally in a character class.
如果您将减号作为第一个或最后一个字符放入组中，则不需要任何掩码，因为它不能像 [az] 中那样表示“直到”。同样，如果插入符号是该组的第一个字符，则它仅表示“不在该组中”： [^az] := not az。
量词 (+?*) 在组中没有任何意义，因此用于表示它们的字符在这里代表它们自己。
其他字符在正则表达式中从来没有特殊含义。
scala 中的快速演示：
You don't need any masking, if you put the minus as first or last character into your group, because then it can't mean 'until' as in [a-z]. Similarly, the caret only means 'not in this group' if it is the first character of the group: [^a-z] := not a-z.
Quantifiers (+?*) don't make any sense in a group, so the characters, used to represent them, stand for themselves here.
The other characters never mean something special in regexes.
Fast demo in scala:
您可以尝试使用字符类之一，例如 \p{Punct} （匹配 US-ASCII 标点符号）或 \W 匹配非单词字符 !(A-Za-z0-9)，而不是显式枚举要匹配的符号。
您可以使用 \p{Punct}* 或 \W* 之类的内容，它们可能会导致比您的其他搜索更广泛的内容，但这可能不是一件坏事......
Instead of explicitly enumerating the symbols to match, you can try using one of the character classes like \p{Punct} (matches US-ASCII punctuation) or \W matches non word characters !(A-Za-z0-9).
You'd use something like \p{Punct}* or \W* they may result in something a bit broader than your other search, but that might not be a bad thing....
通过阅读手册:-)？ + 表示至少匹配前一个字符，* 表示任意数量的字符，|是逻辑或，^ 定义字符的否定。
当然，您也可以逃脱所有这些！
编辑
我明白了，我应该先阅读手册:-)
by reading the manual :-)? + is match at least one previous character, * means any number of characters, | is logical or, ^ defines a negation of a character.
You can also escape all of them to be sure!
EDIT
I see, i should have also read the manual first :-)