Python regular expression select "Nissan" word except between ... or ... tag

Question

View on Live `regex101`

My regular expression pattern is

[Nn]issan(?=[^<>]*<)(?!(?:(?!</?(?:a|span)[ >/])(?:.|\n))*</(?:a|span)>)

I want to stop capture url inside nissan word.view following screenshot.

I use re.sub(pattern, new_word, paragraph, flags=re.U|re.M) function for replace this Nissan word to new_word.

enter image description here

[There was something about this...](http://stackoverflow.com/a/1732454/1763356) — Veedrac, Sep 29 '14 at 06:37
no i don't match href url link. i want to match only Nissan word that word not between **`Nissan or Nissan`**. — Jaykumar Patel, Sep 29 '14 at 06:42
@Veedrac is correct. Even worse is you're trying to use regular expressions on HTML in python, when python has the absolutely amazing beautiful soup library. It would be trivial to make soup of this html then remove all anchor and span tags. Then return any matches in the remains. — Stack of Pancakes, Sep 29 '14 at 06:46
But how to i solve my problem? If i'm not use regex to parse HTML then what is another way? I have paragraph. and when this paragraph is insert into database before i have another table stored word (word, replace_word, url) are replace this paragraph text if found. and after store but paragraph is modified someone again repeat same thing so i am ignore or tag between word. — Jaykumar Patel, Sep 29 '14 at 07:17
The other way is to use an XML or HTML parser to parse HTML. — Veedrac, Sep 29 '14 at 08:42

score 1 · Accepted Answer · answered Sep 29 '14 at 09:45

1

You can try this pattern:

[Nn]issan(?=[^<>]*<)(?!(?:(?!</?(?:a|span)[ >/])(?:.|\n))*</(?:a|span)>)

It has one flaw that I'm aware of, which is that nested <a> or <span> tags can trip it up, causing it to match stuff like this:

<a>nissan<span></span><a>

See demo.

Explanation:

[Nn]issan
(?= # make sure it's not inside an <a> or <span> tag, like <a href="nissan">
    # to do that, we'll assert that the next "<" occurs before ">".
    [^<>]*
    <
)
(?! # next, make sure it's not enclosed in an <a> or <span> tag like <a>nissan</a>
    # to do that, we'll match anything up to the next "a" or "span" tag, either opening or closing, and then assert the tag is opening.
    (?: # while...
        (?! #...there is no opening or closing "a" or "span" tag
            <
            /?
            (?:
                a|span
            )
            [ >]
        )
        (?: # consume the next character.
            .|\n
        )
    )*
    # then assert the tag is not closing.
    </
    (?:
        a|span
    )
    >
)

answered Sep 29 '14 at 09:45

Aran-Fey

39,665
11
104
149

NO this regexp select Nissan name in a tag. – Jaykumar Patel Sep 29 '14 at 10:05
Not sure what you mean. Can you provide an example? – Aran-Fey Sep 29 '14 at 10:07
okay, Let's I have paragraph. and when this paragraph is insert into database before i have another table stored word (word, replace_word, url) are replace this paragraph text if found. and after store but paragraph is modified someone again repeat same thing so i am ignore or tag between word. – Jaykumar Patel Sep 29 '14 at 10:14
Please give me some text to work with, and the expected result. Something like "There is no match in `nissan`, but there should be." – Aran-Fey Sep 29 '14 at 10:24
i update my answer with live view. pls you can check i hope you understand correctly. thanks – Jaykumar Patel Sep 29 '14 at 11:13
Hello, your pattern is work for me but still need one help. I do not capture www.Nissan.com or http://www.Nissan.com inside Nissan word. Thanks a lot you can help me wonderfull. god bless you :) – Jaykumar Patel Oct 01 '14 at 12:24
i add screenshot on question. – Jaykumar Patel Oct 01 '14 at 12:33

score -1 · Answer 2 · answered Sep 29 '14 at 06:41

-1

Nissan(?!((?!<\/a>).)*<\/a>|((?!<\/span>).)*<\/span>)

Try this.See demo.

http://regex101.com/r/dN8sA5/2

answered Sep 29 '14 at 06:41

vks

67,027
10
91
124

Why do you repeat twice each closing tag? Isn't it possible to do like this: `Nissan(?!.*<\/a>|.*<\/span>)` ? – WhiteAngel Sep 29 '14 at 06:49
Can you inform me the scope of this regexp is single line. – Jaykumar Patel Sep 29 '14 at 06:49
@user1153551 the regex you posted had `m` flag.that was single line too. – vks Sep 29 '14 at 06:50
Your regexp work for me but not 100% archived my goal and i'm not downvote. you are trying to help me so i'm give up to u. – Jaykumar Patel Sep 29 '14 at 07:20
Note that this pattern won't match `nissan`. – Aran-Fey Sep 29 '14 at 08:26
@Rawing that should not be match as per OP – vks Sep 29 '14 at 08:33
@vks: Huh? Shouldn't it match as long as `nissan` isn't enclosed in an `` or `` tag? – Aran-Fey Sep 29 '14 at 08:36
@Rawing `nissan` this is not enclosed in right? – vks Sep 29 '14 at 08:37
@vks: It is not. Judging from the sample input OP has provided, it'll probably be enclosed in some HTML tag - but the idea is that your pattern picks up an `` tag that isn't relevant (aka doesn't enclose `nissan` in any way). – Aran-Fey Sep 29 '14 at 08:41

Python regular expression select "Nissan" word except between ... or ... tag

View on Live regex101

2 Answers2

View on Live `regex101`