Question

* E/CN.6/2009/1。如同工作一样(这里有各种突然尝试,要求不同的应用):

<ref name="*([^">/(?=>)]*)"* *(/*)>
<ref name="*([^">/(?=>)]*)"* *(/*)>
<ref name="*([^">/(?=>)]*)"* *(/*)>
<ref name="*([^">/(?=>)]*)"* *(/*)>

处理类似案文的案件

<ref name="Te/st12" />

缩略语

我有工作要做。

<ref name=Test1/>
<ref name=Test2 />
<ref name="Test3"/>
<ref name="Test4" />
<ref name=Test5>Foo</ref>
<ref name="Test6">Foo</ref>
<ref name=Te/st7/>
<ref name=Te/st8 />
<ref name="Te/st9"/>
<ref name="Te/st10" />
<ref name=Te/st11>Foo</ref>
<ref name="Te/st12">Foo</ref>

This regex seems like it should work (and I ve tried several variants on the theme):

<ref name="*([^">/(?=>)]*)"* *(/*)>

或者如果在某一特定应用中需要更紧急的:

<ref name="*([^">/(?=>)]*)"* *(/*)>

或更少:

<ref name="*([^">/(?=>)]*)"* *(/*)>

但是,它未能将所有包含<代码>/的编号为示例数据的案例。

更方便地研究这一问题:

<ref name="*([^">/(?=>)]*)"*

或者如果在某一特定应用中需要更紧急的:

<ref name="*([^">/(?=>)]*)"*

或更少:

<ref name="*([^">/(?=>)]*)"*

It always chokes on the / in the name data of the latter examples, despite /(?=>) (which depending on application might be escaped as /(?=>) or other variations thereof), meaning "the slash character when followed by the greater-than character". My suspicion is that (?=...) doesn t work properly inside [^...], but I ve tried various | (or) constructions, e.g. using a separate [?!>], and those didn t work either.

目标是通过下列方式将所有这类案例(包括有“姓名数据”或“姓名数据”的案件)正规化成一个前后一致的格式:

<ref name="$1"$2>

PS:我将此称为“pcre”。相关应用是Wikipedia现有编辑工具的检索/替代特征(编辑一页,在编辑室外的工具包中点击“和”“先进”,而远右面看看看成一个宽松的玻璃一孔,它带有一种reg子的搜索/替代特征,但没有确切说明什么是reg鱼。 I m 不是使用打字/...,而是在整个表述上加标记,因为申请没有使用。

Answer 1

您不能把眼光放在一个名为[<>/code>的特性中,试图将其分为两种使用OR条件的不同案例:


<ref name="?((?:[^">/]|/(?!>))*?)"? */?>

<refsname="?    # start of the `ref` tag, up to the `name` attribute

(                # group 1, what you want to capture
  (?:            # non-capturing group, for grouping the following OR condition
      [^">/]    # any character but `">/`
      |          # OR
      /(?!>)    # if it IS a `/`, then it must NOT be followed by a `>`
  )*?            # this non-capturing group repeats 0 to unlimited times, as few as possible (to avoid capturing trailing whitespaces)
)                # group 1 ends

"?s*/?>        # everything after the `name` attribute

Check the test cases here.

But be aware, matching HTML/XML tags using regex is generally considered a bad idea. It only works for some basic validation (this doesn t work for inputs like name="a>b", name= ab  etc.), for complex ones, it is recommended to use some markup language parsing library to handle it.

Answer 2

鉴于你们别无选择,只能使用reg,你可以使用reg:

<ref name=(?:"((?:[^"\]|\.)*)"|((?:(?!s*/?>).)*))s*(/?)>

这与:

<ref name= : literally <ref name=
either:
1. "((?:[^"\]|\.)*)" : a quoted string (captured in group 1) that may include escaped characters, including other quotes (from this answer); or
2. ((?:(?!s*/?>).)*) : an unquoted string (captured in group 2) which will end when the position at the regex cursor matches s*/?>. This uses a tempered greedy token to not match any further than necessary.
s*(/?)> : some spaces, an optional /, captured in group 3, and a >

您将取代以下表格:

<ref name="$1$2"$3>

Demo on regex101

友情链接