如何添加以防止正则表达式解析字符串的某些部分？

Question

As per this answer , I am making use of a replaceAll() function to swap arbitrary strings in my javascript (Node) application:

this.replaceAll = function(str1, str2, ignoreCase)
{
    return this.replace(new RegExp(str1.replace(/([\/\,\!\\\^\$\{\}\[\]\(\)\.\*\+\?\|\<\>\-\&])/g,"\\$&"),(ignoreCase?"gi":"g")),(typeof(str2)=="string")?str2.replace(/\$/g,"$$$$"):str2);
}

I would like to expand this regex, such that it does not try to match anything inside a set of <span>...</span> tags. (I need to add HTML Span tags around parts of some strings, and I don't want to wrap anything in a span twice , if the patterns would duplicate [Ie, 'Foo' and 'Foobar' in the string "Foobarbaz"])

I am running multiple regex search/replaces on a single string, and I want to make sure that nothing gets processed multiple times.

My understanding is that I'll need a [<SPAN>] ... [</SPAN>] somehow, but I'm not quite sure on the specifics. Any advice?

Answer 1

I don't understand what your regexp does, but in general the technique is like this:

html = "replace this and this but not <span> this one </span> or <b>this</b> but this is fine"

// remove parts you don't want to be touched and place them in a buffer
tags = []
html = html.replace(/<(\w+).+?<\/\1>/g, function($0) {
    tags.push($0)
    return '@@' + (tags.length - 1)
})

// do the actual replacement
html = html.replace(/this/g, "that")

// put preserved parts back
html = html.replace(/@@(\d+)/g, function($0, $1) {
    return tags[$1]
})