find: (.*)

dicknskip · 03-14-2016, 03:40 PM

I want to do a search that finds: stuff and things and there is more than one occurrence in the paragraph. I want to find them one at a time. I am using: (.*) as my search. I looked up the regex and can't seem to work out the code that will find each occurrence rather than span from the first to the last in the paragraph. The contents could be upper, lower, punctuation, numbers or white space, at random, thus the .*. TIA.

jackie_w · 03-14-2016, 04:10 PM

I think you need Regex's advanced features 'lookahead' and 'lookbehind' to do what you want. I'm by no means expert in this but I think it would be something like

Code:

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

Google should find you more detailed info.

davidfor · 03-14-2016, 06:39 PM

I would have used:

Code:

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

The ? makes the match non-greedy.

dicknskip · 03-14-2016, 08:57 PM

That worked perfectly. Thanks.

eschwartz · 03-15-2016, 12:53 AM

Quote:

Originally Posted by jackie_w

I think you need Regex's advanced features 'lookahead' and 'lookbehind' to do what you want. I'm by no means expert in this but I think it would be something like

Code:

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

Google should find you more detailed info.

Wrong use of lookaround

that restricts you to just matching the text instead of the tags as well.
Useful feature, but not quite what the OP wanted.

...

davidfor's solution works, kind of... but the proper way to do it looks like this:

Code:

<i>((?:(?!</?i>).)*)</i>

This (the section in red) uses the power of negative lookaheads to ensure that the dot-match-everything can only match characters that are NOT followed by an "i" opening/closing tag.

...

The difference between mine and davidfor's solutions is that, given the sample text:

Code:

<i>sample <i>text</i></i>

Mine will match the single internal unit "text" whereas davidfor's will erroneously match from the first opening "i" tag until the first closing "i" tag.

This would matter a lot more if we were talking about a tag which is meant to be nested.

Like spans.

Which is why it is a very useful principle to know, although in this case you really can just make do with the easy solution.

credits: I originally learned this trick here: https://stackoverflow.com/questions/.../406408#406408

jackie_w · 03-15-2016, 09:37 AM

Quote:

Originally Posted by eschwartz

... that restricts you to just matching the text instead of the tags as well.

Unfortunately that was actually the question I thought was being asked. Now I've re-read the OP, I see I was mistaken.

It would appear my Regex skills exceed my skills in basic reading/comprehension in my native language

I wonder how common that is?

Sorry for any confusion, dicknskip.

eschwartz · 03-15-2016, 09:57 AM

Quote:

Originally Posted by jackie_w

It would appear my Regex skills exceed my skills in basic reading/comprehension in my native language

I wonder how common that is?

Happens to us all.

(Except me, obviously. I deny everything.

)

theducks · 03-15-2016, 10:24 AM

Quote:

Originally Posted by jackie_w

Unfortunately that was actually the question I thought was being asked. Now I've re-read the OP, I see I was mistaken.

It would appear my Regex skills exceed my skills in basic reading/comprehension in my native language

I wonder how common that is?

Sorry for any confusion, dicknskip.

I think this line currently wraps around a New York City block (they tend to be BIG).

You are behind me, somewhere, way back there... (And I can't see the front-o-the-line )

jackie_w · 03-15-2016, 10:50 AM

Quote:

Originally Posted by theducks

I think this line currently wraps around a New York City block (they tend to be BIG).

You are behind me, somewhere, way back there... (And I can't see the front-o-the-line )

That many! Can you see any of the world's great thinkers, movers and shakers in that line? If so we're all DOOMED!

Barook · 03-17-2016, 11:47 PM

I do a first pass like so:

Code:

   text: <span class="italic">something</span>
replace: <span class="italic">([^<]+)</span>
   with: <i>\1</i>
 result: <i>something</i>

It doesn't work when there are other tags inside the tags...

Code:

<span class="italic"><span class="bold">something</span></span>

... but a non-greedy replace usually cleans those up:

Code:

   text: <span class="italic"><span class="bold">something</span></span>
replace: <span class="italic">(.*?)</span>
   with: <i>\1</i>
 result: <i><span class="bold">something</span></i>
replace: <span class="bold">(.*?)</span>
   with: <b>\1</b>
 result: <i><b>something</b></i>

A similar pattern is useful to isolate stuff in quotes:

Code:

   text: <a id="return_from_note_1"></a><a href="notes.html#note_1">see note 1</a>
replace: <a id="([^"]+)"></a><a href="([^"]+)">(.*?)</a>
   with: <a id="\1" href="\2">\3</a>
 result: <a id="return_from_note_1" href="notes.html#note_1">see note 1</a>

... or for removing all attributes from uselessly-extravagant tags:

Code:

   text: <i class="italic" style="margin:auto;padding:auto;font-size=1em;">something</i>
replace: <i[^>]+>something</i>
   with: <i>\1</i>
 result: <i>something</i>

davidfor · 03-18-2016, 12:07 AM

I haven't hit the anchor one, but for the rest, I usually use Diap's Editing Toolbag. Much easier than trying to wrap my brain around the regex.

For the last one, why do you care about matching the close tags? Just fixing the opening should be enough. So:

Code:

text: <i class="italic" style="margin:auto;padding:auto;font-size=1em;">
replace: <i[^>]+>
   with: <i>
 result: <i>

should work the same. Am I missing something?

03-14-2016, 03:40 PM	#1
dicknskip Zealot Posts: 134 Karma: 10 Join Date: Nov 2009 Location: Okotoks, AB, Canada Device: iPad V-3	*find: <i>(.)</i>** I want to do a search that finds: <i>stuff and things</i> and there is more than one occurrence in the paragraph. I want to find them one at a time. I am using: <i>(.)</i> as my search. I looked up the regex and can't seem to work out the code that will find each occurrence rather than span from the first <i> to the last </i> in the paragraph. The contents could be upper, lower, punctuation, numbers or white space, at random, thus the .. TIA.

03-14-2016, 04:10 PM	#2
jackie_w Grand Sorcerer Posts: 6,251 Karma: 16539642 Join Date: Sep 2009 Location: UK Device: ClaraHD, Forma, Libra2, Clara2E, LibraCol, PBTouchHD3	I think you need Regex's advanced features 'lookahead' and 'lookbehind' to do what you want. I'm by no means expert in this but I think it would be something like Code: (?<=<i>)(.*)(?=</i>) Google should find you more detailed info.

03-14-2016, 06:39 PM	#3
davidfor Grand Sorcerer Posts: 24,905 Karma: 47303824 Join Date: Jul 2011 Location: Sydney, Australia Device: Kobo:Touch,Glo, AuraH2O, GloHD,AuraONE, ClaraHD, Libra H2O; tolinoepos	I would have used: Code: <i>(.*?)</i> The ? makes the match non-greedy.

03-18-2016, 12:07 AM	#11
davidfor Grand Sorcerer Posts: 24,905 Karma: 47303824 Join Date: Jul 2011 Location: Sydney, Australia Device: Kobo:Touch,Glo, AuraH2O, GloHD,AuraONE, ClaraHD, Libra H2O; tolinoepos	I haven't hit the anchor one, but for the rest, I usually use Diap's Editing Toolbag. Much easier than trying to wrap my brain around the regex. For the last one, why do you care about matching the close tags? Just fixing the opening should be enough. So: Code: text: <i class="italic" style="margin:auto;padding:auto;font-size=1em;"> replace: <i[^>]+> with: <i> result: <i> should work the same. Am I missing something?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Find/Replace Won't Find	Rand Brittain	Sigil	7	09-24-2011 04:35 AM
Get Find to find an exact phrase	evwool	Amazon Kindle	4	08-16-2011 08:47 AM
Find this NOT that	Danger	Sigil	5	12-27-2010 03:13 PM
Help me find this if you can...	Spanner	Reading Recommendations	3	10-12-2010 12:38 PM

03-14-2016, 08:57 PM	#4
dicknskip Zealot Posts: 134 Karma: 10 Join Date: Nov 2009 Location: Okotoks, AB, Canada Device: iPad V-3	That worked perfectly. Thanks.

Advert

Advert