regex - Rails 正则表达式排除符号结束单词之间的空格

Question

我正在尝试标记两个 #BLOCK 中的每一行行。

此外，我想排除所有符号 [""," ",{},(),\n]

#BLOCK
#NAME {PC8}
#TYPE GHD3
#PROGRAM "FooBar" (2.0)
#DATE 20190501
#BASE 3740 "TXGH3789"
#BLOCK

现在，我有两个解决方案，但我想将它们合二为一。

我正在使用 Rubular，链接在这里:

示例 1:https://rubular.com/r/bd2AxaHB2QLGpt

示例 2:https://rubular.com/r/vmxm2kugNhnDCS

我已经尝试了这两种解决方案:

1. (?<=#BLOCK\n)(.*)(?=#BLOCK)这是有效的，它标记了两个 #BLOCK 中的所有内容行。

2. [^,{},(),""," ",\n]这可以排除这些符号，但不会标记两个 #BLOCK 之间的内容。行。

我怎样才能将两者结合起来以获得我在开头描述的预期结果？

预期结果是 #BLOCK 之间的两个标记行并排除符号，如 [{},(),""," ",\n] .

Answer 1

如果“标记”是指匹配，我猜你可以试试这个。
它使用 \G构造。

(注意 - Ruby 使用 //m 选项来表示全点)

(更新 - 不要让它在没有重新启动的情况下超过下一个 block )

/(?:(?:(?<=\#BLOCK\n)|(?!^)\G))[,{}()"\s]*\K(?!\#BLOCK\b)[^,{}()"\s](?=.*\#BLOCK\b)/m

https://rubular.com/r/TxlU9yhiUJkrok

解释
注意 - 此正则表达式一次匹配一个字符。

 (?:
      (?<= \#BLOCK  \n )            # A block behind
   |                              # or,
      (?! ^ )                       # Not the BOS
      \G                            # Start matching where last match left off
 )
 [,{}()"\s]*                   # Consume optional punctuation and whitespace
 \K                            # Disregard anything matched so far
 (?! \#BLOCK \b )              # Don't go past next block
 [^,{}()"\s]                   # Get a single non-punct nor whitespace char
 (?= .* \#BLOCK \b )           # Only if there is a block ahead

---

要匹配字符 block ，请使用这个。

/(?:(?<=\#BLOCK\n)|(?!^)\G)[,{}()"\s]*\K(?=.+\#BLOCK\b)(?:(?!\#BLOCK\b)[^,{}()"\s])+/m

https://rubular.com/r/kyhqnOtIrmrnJ7

解释

 (?:
      (?<= \#BLOCK  \n )            # A block behind
   |                              # or,
      (?! ^ )                       # Not the BOS
      \G                            # Start matching where last match left off
 )
 [,{}()"\s]*                   # Consume optional punctuation and whitespace
 \K                            # Disregard anything matched so far
 (?= .+ \#BLOCK \b )           # Check that there is a block ahead
 (?:
      (?! \#BLOCK \b )              # Don't go past next block
      [^,{}()"\s]                   # Get a single non-punct nor whitespace char
 )+