move * before

Pouemes44

$1$3

wahoo a great thanks claudia, it works
can you explain me how it works

Claudia Frank

I’m not really good in regex explanations but I’ll give it a try.

each of the () denote a matching group which can be reflected by using a dollar sign and a number.
$1 for the first matching group, $2 for the second and so on.

So what get’s matched in (<a.?>) is stored in $1,
(*\x20) in $2 and (.?</a>) in $3

(<a.*?>) = we are looking for something which starts with <a
followed by .* means anything either none or multiple time,
the question mark restricts that it should be as less as possible
chars consumed followed by >.

(*\x20) we are looking for the asteriks and \x20 is the hexcode for a space

and last but not least, (.*?</a>) which is basically similar to the first.

In the replace with field you ignore the second matching group as you want to get rid of the asteriks and space,
instead you type it in front of the $1$3 because that is what you wanted.

Better described at http://www.boost.org/doc/libs/1_55_0/libs/regex/doc/html/boost_regex/syntax/perl_syntax.html.

Cheers
Claudia

guy038

Hello, @pouemes44, @claudia-franck and All,

An other solution could be :

SEARCH (?-s)\*\x20+(?=.+?</a>)

REPLACE Leave EMPTY

Notes :

As usual, the modifier (?-s) means that the dot, ., matches a single standard character, only
Then, the regex engine is looking for a literal star symbol ( \* ) followed by at least one space character ( \x20+ )
But, ONLY, if the look-ahead is verified. That is to say, IF it exists, further on, an ending tag </a>, on the same line
As the replacement zone is empty , the combination star + space(s) is simply deleted !

Of course, if the syntax \*\x20+ occurs, ONLY, in blocks <a ......>......</a> of your file, just use :

SEARCH \*\x20+

REPLACE Leave EMPTY

Best Regards,

guy038

Pouemes44

thanks both for explanations

but guy with your solution, its only delete all *space but not move them before <a

Pouemes44

and claudia i just saw that your solution does errors

<a href=“/url/page.html”>name</a>
<a href=“/url/other.html”>name2</a>
<a href=“/url/third.html”>etc</a>
<a href=“/url/four.html”>* something</a>

give

<a href=“/url/page.html”>name</a>
<a href=“/url/other.html”>name2</a>
<a href=“/url/third.html”>etc</a>
<a href=“/url/four.html”>something</a>

i would

<a href=“/url/four.html”>something</a>

Pouemes44

like this seems to works

search (?-s)(<a.+?">)+*\x20+(?=.+?</a>)
replace * $1

Pouemes44

@guy038 said:

this two
(?-s)(<a.+?">)*\x20
replace * $1

Claudia Frank

@Pouemes44

I don’t see really the difference, I guess because the markdown used some of your chars for formatting.
To prevent this you can either indent every line with 4 spaces or escape the markdown characters with
a backslash, like

* whatever */

I had just to put 4 spaces in front of the asteriks and every single character is shown whereas here

*whatever*/

I had to write \*whatever*/ to inform markdown that the asteriks should be displayed and not used to reformat text.

Anyway, good to see that you have found your solution.

Cheers
Claudia

Pouemes44

yes claudia,
i dont know what was the poblem with your solution
anyway i succeed
a great thanks

guy038

Hi, @pouemes44, @claudia-frank and All,

I apologize, Pouemess44, because I didn’t read your post carefully :-( I should have read the line :

so to put the * before <a

and, also, the title of your post !

move * before

So, seemingly, from the initial text :

  <a href=“url/myurl.htm”>* name</a>

Your goal is to obtain the text below :

  *<a href=“url/myurl.htm”>name</a>

In that case, you’re perfectly right ! A correct regex S/R is, very similar to yours, is :

SEARCH (?-s)(<a href.+?>)\*\x20+

REPLACE *$1

Cheers,

guy038

P.S. :

I guess that Claudia did the same mistake than me !! So, the correct Claudia’s regex S/R is, rather :

SEARCH (<a.*?>)(\*\x20)(.*?<\/a>)

REPLACE *$1$3

Pouemes44

@guy038 said:

(?-s)

hello thanks for answer
i try also to learn
for claudia solution i have no error if i add (?-s)

like this
(?-s)(<a.?>)(*\x20)(.?</a>)

$1$3

without (?-s) or with (?s) i get errors in results