Regex examples

[John Doe]

Thanks! It seems that I just got my self some bedtime reading

Now, if I want to search for something like the above, but dont know how much text there is, how do I do that?

Example:

I search for this:

<p><a id="id1">This is the text that is in the note and I dont know how much there is. It changes for every note.</a><br /></p>

And want to replace it with this:

<aside epub:type="footnote">
    <div epub:type="footnote" id="id1"><br />
      <p>This is the text that is in the note and I dont know how much there is. It changes for every note.</p>
    </div>
</aside>

The ID numbers is quite easy after your help, but how do I search and replace the rest without changing the note? I want to replace everything around the note, but the text between "<p>" and "</p>" can be changing from note to note.

[mzmm]

regular expressions will be too brittle for this kind of thing. i think you should look into a programmatic way of parsing the html.

[John Doe]

Don't wanna go into that.

I just simply search/replace in three steps.

But thankyou for your help!

[gipsy]

Hi,
there is any way to find the word "τον" between italics and replace it with "του"

For example:

<i>Αυτός ο βασιλιάς ζει στη σκιά τον αδερφού τον</i>

must be

<i>Αυτός ο βασιλιάς ζει στη σκιά του αδερφού του</i>

Thanks

[theducks]

Originally Posted by gipsy

Hi,
there is any way to find the word "τον" between italics and replace it with "του"

For example:

must be


Thanks

<i>(.*?) tov</i>

<i>\1 tou</i>

[DiapDealer]

That will only catch the second instance in the given example, Ducks (and only those instances that immediately precede a closing 'i' tag.

Pretty hard to come up with something that will catch multiple occurences only within 'i' tags (without multiple passes anyway). I have no idea how common that word is, but it might be easier to step through each occurence one by one and manually eyeball the ones that are in italic tags.

[theducks]

Originally Posted by DiapDealer

I have no idea how common that word is, but it might be easier to step through each occurence one by one and manually eyeball the ones that are in italic tags.

That is the way I would do it
Why spend all the time creating and testing a difficult REGEX for a 1 time use?
Find or Replace Next. The buttons are a tiny mouse move away.

[Skeeve]

Originally Posted by gipsy

Hi,
there is any way to find the word "τον" between italics and replace it with "του"

In perl notation:

s/τον(?=(?:[^<]+|<[^i]|<i[^>])*?<\/i>)/του/g

[Doitsu]

Originally Posted by gipsy

Is there is any way to find the word "τον" between italics and replace it with "του"?

The following very simple regex will find one occurrence of "tov" at the beginning, at the end and in the middle of a string of text enclosed by <i></i> tags:

Find:<i>(.*?)τον(.*?)</i>
Replace:<i>\1του\2</i>

You'll, of course, have to run it multiple times to find multiple occurrences of "τον" enclosed by <i></i> tags.

[Jellby]

Originally Posted by Doitsu

The following very simple regex will find one occurrence of "tov" at the beginning, at the end and in the middle of a string of text enclosed by <i></i> tags:

Find:<i>(.*?)τον(.*?)</i>
Replace:<i>\1του\2</i>

But it would also find it between two <i></i> blocks, wouldn't it?

<i>foo</i> τον <i>bar</i>

Even with non-greedy modifier, the asterisk matches as many as needed to make the whole pattern match.

[Skeeve]

Originally Posted by Jellby

But it would also find it between two <i></i> blocks, wouldn't it?

<i>foo</i> τον <i>bar</i>

Even with non-greedy modifier, the asterisk matches as many as needed to make the whole pattern match.

You are absolutely right!

BTW: Mine expression also isn't fool-proof as it would wrongly handle invalid HTML:

<i>foo</i> τον bar</i>

[Doitsu]

Originally Posted by Jellby

But it would also find it between two <i></i> blocks, wouldn't it?

You're are of course right, but as DiapDealer has already pointed out, it's pretty much impossible to come up with a regex that covers all possible permutations. However, my simple regex covers at least the three most likely scenarios. I.e., an initial, medial or final occurrence of the search string.

[Perkin]

How about

Find:<i>([^<]*?)τον([^<]*?)</i>
Replace:<i>\1του\2</i>

[Jellby]

Originally Posted by Perkin

How about

Find:<i>([^<]*?)τον([^<]*?)</i>
Replace:<i>\1του\2</i>

It wouldn't catch:

<i><b>foo</b> τον bar</i>

[gipsy]

Thanks for the replies!

I tried the

Find: (<i>)(.*?) (τον)(.*?)(</i>)
Replace: \1\2 του\4\5

and works in many issues.

@theducks there are 3552 "τον" in the doc i use as sample to check.
with the regex. 312