I’m back for additional information, about lazy, greedy and possessive quantifiers. it’s fundamental to, correctly, understand the differences, between these 3 types of quantifiers !

So, let’s consider the simple text 12345ABCDE, in a new tab

How the regex engine interprets, for instance, the regex \w{1,10}[A-Z]{5}, with the greedy quantifier {1,10} ?. Well :

• It, first, tries to match the LONGEST range of \w => 10 Word characters. But, the part [A-Z]{5} CANNOT match anything

• Then, it backtracks and tries the first 9 Words characters. Again, the part [A-Z]{5} does NOT match the E letter

• Then, it backtracks and tries the first 8 Words characters. Again, the part [A-Z]{5} does NOT match the DE letters

• Then, it backtracks and tries the first 7 Words characters. Again, the part [A-Z]{5} does NOT match the CDE letters

• Then, it backtracks and tries the first 6 Words characters. Again, the part [A-Z]{5} does NOT match the BCDE letters

• Then, it backtracks and tries the first 5 Words characters. This time, the part [A-Z]{5} DOES match the ABCDE letters

=> After the backtracking phase, all the text is matched and selected !

Now, how the regex engine interprets the regex \w{1,10}?[A-Z]{5}, with the lazy quantifier {1,10}? ?

• It, first, tries to match the SHORTEST range of \w => 1 Word character. But, the part [A-Z]{5} CANNOT match the 2345ABCDE string

• Then, it backtracks and tries the first 2 Words characters. Again, the part [A-Z]{5} does NOT match the 345ABCDE string

• Then, it backtracks and tries the first 3 Words characters. Again, the part [A-Z]{5} does NOT match the 45ABCDE string

• Then, it backtracks and tries the first 4 Words characters. Again, the part [A-Z]{5} does NOT match the 5ABCDE string

• Then, it backtracks and tries the first 5 Words characters. This time, the part [A-Z]{5} DOES match the ABCDE letters

=> After the backtracking phase, all the text is matched and selected !

Note : Instead of the English werb backtrack, the verb fortrack would be more adapted ! Sorry, English isn’t my mother tongue !

Finally, how the regex engine interprets the regex \w{1,10}+[A-Z]{5}, with the possessive quantifier {1,10}+ ?

• It, first, tries to match the LONGEST range of \w => 10 Word characters. But, the part [A-Z]{5} CANNOT match anything

• Now, the normal process would be to backtrack. But this action is forbidden, due to the possessive quantifier ! In other words, once a match has been found, for the first part \w{1,10}+, the following parts of the regex must match the remaining of the text. But, as the first regex part have consumed all the text, the part [A-Z]{5} will NEVER match anything !
  So, the overall match fails and you get the normal message Find: Can’t find the text “\w{1,10}+[A-Z]{5}”

Using, again, the same example 12345ABCDE, in a new tab, it’s easy to verify that :

• The regex \w{1,10} matches the longest Word characters range => The whole string 12345ABCDE is matched

• The regex \w{1,10}? matches the shortest Word characters range => The 1 Word character is matched, then the 2 digit and so on…

• The regex \w{1,10}+ matches the longest Word characters range => The whole string 12345ABCDE is matched, too !

So, to sum up, here is, below, a list of all the quantifiers :

    GREEDY     quantifiers   :    *  ( = {0,} )       + ( = {1,} )      ?  ( = {0,1} )      {n}        {n,}        {m,n}

    LAZY       quantifiers   :    *? ( = {0,}? )     +? ( = {1,}? )     ?? ( = {0,1}? )     {n}?       {n,}?       {m,n}?

    POSSESSIVE quantifiers   :    *+ ( = {0,}+ )     ++ ( = {1,}+ )     ?+ ( = {0,1}+ )     {n}+       {n,}+       {m,n}+

Remark : The two {n}? and {n}+ syntaxes, although correct, are useless, as the syntax {n} could be qualified as an EXACT quantifier !

Best Regards,

guy038

• In the regex (?s)(^Block \d+.+?End block)(.*?)(^Block \d+.+?End block), of my previous post, the part (.*?) represents the shortest range of characters, even empty, stored as group 2, between the two consecutive blocks of text, that are to be swapped !

Now, if we consider the HTML example text, below :

<td height="15">
    <font size="2" color="black" face="arial, verdana"><b>Lire un message / dossier : Reçus</b></font> 
</td>

<td height="15">
    <font size="2" color="black" face="arial, verdana"><b>Lire un message / dossier : Reçus</b></font> 
</td>

Beware about the two different behaviours :

• The regex (?s)<td.+</td> looks for the largest range of characters .+, between the strings <td and </td> => It matches all the text, at once

• The regex (?s)<td.+?</td> looks for the shortest range of characters .+?, between the strings <td and </td> => it matches, successively, each block <td…</td>

Now, regarding your HTML text :

<a href=“mypage.htm” title=page">page</a>

I suppose that you do NOT get a match, using the regex (?s)(<a.+?</a>\R), because, probably, it’s the last line of your current file, which is NOT followed by any End of line character !

Indeed, in that case, the ending part \R cannot match anything. So the overall match fails :-((

Two solutions :

• Use the regex (?s)(<a.+?</a>\R?). With that syntax, the \R part is optional

• Add a line break to this last line !

For a better visualization of the End of line characters, just click on the ¶ icone

Cheers,

guy038

but if i search (?s)(<a.+?</a>\R) to get <a href=“mypage.htm” title=page">page</a> i got nothing why? what is wrong?

Here is, below, the general method in order to switch two consecutive blocks of text, separated, or not, by some stuff text

So let’s start with the example text, below :

bla
bla
bla
Block 1
of
some
text
End block
bla bla
bla bla
bla bla
bla bla
Block 2
with some
other text
End block
bla bla bla
bla bla bla
bla bla bla

To correctly determine the limits of your block of text, whatever it is, you, necessarily, need to know :

• The location of the beginning of your block. In our example, I suppose that it’s the regex ^Block \d+

• The location of the end of your block. In our case, I suppose it’s the string End block

IMPORTANT : In our example, the second block has the same limits as Block 1, but it could, perfectly, have some other limit definitions !

So, from the above definitions, we can build the regex S/R, below :

SEARCH (?s)(^Block \d+.+?End block)(.*?)(^Block \d+.+?End block)

REPLACE \3\2\1

Notes :

• The leading (?s) modifier means that the special dot character represents, absolutely, any single character. Now :

• The first part, (^Block \d+.+?End block, represents a complete individual block of text, stored as group 1

• The third part, (^Block \d+.+?End block), stands, again, for an other complete block of text, stored as group 3

• The middle part, (.*?), is the shortest range, even empty, of any character, between these two consecutive blocks of text

• In replacement, we just switch the group 1 and 3, with group 2 standing as a pivot

And, after a click on the Replace All button, you should get the expected text , below :

bla
bla
bla
Block 2
with some
other text
End block
bla bla
bla bla
bla bla
bla bla
Block 1
of
some
text
End block
bla bla bla
bla bla bla
bla bla bla

Just note that, if the two blocks are strictly consecutive, without the bla bla lines, as below :

bla
bla
bla
Block 1
of
some
text
End block
Block 2
with some
other text
End block
bla bla bla
bla bla bla
bla bla bla

We still get the correct modified text :

bla
bla
bla
Block 2
with some
other text
End block
Block 1
of
some
text
End block
bla bla bla
bla bla bla
bla bla bla

Now, Pouemes44, let’s apply this general method to your particular problem :

The general template, of your first block, is :

<noscript......
......
......
......
......
</noscript>

The general template, of your second block, is :

<img.........
.......
.......
.......
style="display:none;"></div>

Therefore, the corresponding regex S/R, in your case, is, obviously :

SEARCH (?s)(^<noscript.+?</noscript>)(.*?)(^<img.+?style="display:none;"></div>)

REPLACE \3\2\1

So, considering your example, below :

1.html
<noscript class"text" lang="example">
line1
line2
line3
line4
</noscript>
some text…
etc…
etc
<img… and …style="display:none;"></div>

2.html
<noscript class"text" lang="otherexample">
otherline1
otherline2
otherline3
otherline4
</noscript>
othersome text…
etc…
etc etc
<img… other and …style="display:none;"></div>

After performing the above S/R, you’ll get the modified text, as expected to :

1.html
<img… and …style="display:none;"></div>
some text…
etc…
etc
<noscript class"text" lang="example">
line1
line2
line3
line4
</noscript>

2.html
<img… other and …style="display:none;"></div>
othersome text…
etc…
etc etc
<noscript class"text" lang="otherexample">
otherline1
otherline2
otherline3
otherline4
</noscript>

Et voilà !!

Best Regards,

guy038

P.S. : Beware that, on our NodeBB site, text, containing starting and ending usual simple and double quotes, are changed into their Unicode equivalents, below :

• Starting simple quote ‘, of Unicode code-point \x{2018}, instead of the single quote sign ' ( x{0027 )

• Ending simple quote ’, of Unicode code-point \x{2019}, instead of the single quote sign ' ( x{0027 )

• Starting double quote “, of Unicode code-point \x{201C}, instead of the double quote sign " ( x{0022 )

• Ending double quote ”, of Unicode code-point \x{201D}, instead of the double quote sign " ( x{0022 )

Notes :

• For ANSI encoded texts, to avoid the nasty message Find: Invalid regular expression, the **correct regex syntaxes are, respectively, \x91, \x92, \x93 and \x94

BTW :

• Remember that the \x{####} regex syntax can be used :

  • For search of a true ASCII character, between \x{0000} and \x{007F}, in ANSI encoded files

  • For search of any Unicode character, between \x{0000} and \x{FFFF}, in UNICODE encoded files

• Remember that the \x{##} regex syntax can be used :

  • For search of a true ASCII character, between \x{00} and \x{7F}, in ANSI encoded files

  • For search of any Unicode character, between \x{00} and \x{FF}, in UNICODE encoded files

• To end with, remember that the \x## regex syntax can be used :

  • For search of any Unicode character, between \x00 and \xFF, either, in UNICODE or ANSI encoded files

1.html
<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>
some text…
etc…
etc
<img… and …style=“display:none;”></div>

2.html
<noscript class"text" lang=“otherexample”>
otherline1
otherline2
otherline3
otherline4
</noscript>
othersome text…
etc…
etc etc
<img… other and …style=“display:none;”></div>

without having the data we cannot say for sure what went wrong.
So, in theory, the hyperlink regex would match everyhting within the tag if
closing tags is directly followed by an eol char (no space)
and the image regex would match starting with <img class followed by
anything until ”> appears, followed directly by an eol char.

Cheers
Claudia

if i try to get an other sequence between a a tag by example… i got nothing! what is wrong?

(?s)(<a href.+?</a>\R)

with an img tag it doesnt take the first “> why? :-) thanks
(?s)(<img class.+?”>\R)

A good starting point for Regular Expressions is http://www.regular-expressions.info/

Here is my contribution, to your problem ;-)

Let’s suppose you have two files 1.html and 2.html, as below :

<!-- 1.html -->
Some text
before the
"noscript" tag
<noscript class"text" lang="example">
line1
line2
line3
line4
</noscript>
bla
bla
bla
blah
<img… and …style="display:none;"></div>
And some text,
located after
............
<noscript>
............

and

<!-- 2.html -->
bla
bla
<noscript>
bla

<noscript class"text" lang="otherexample">
Other line1
Other line2
Other line3
Other line4
Other line5
</noscript>
Other text…
etc…
etc…
etc…
<img… other and …style="display:none;"></div>
Other Text
..........
till the END of the file

• Move back, at the very beginning of current file ( Ctrl + Origin )

• Open the Replace dialog ( Ctrl + H )

• Select the Regular expression search mode

• In the Search what: box, type in the regex (?s)(<noscript class.+?</noscript>\R)(.+?style="display:none;"></div>\R)

• In the Replace with: box, type in the regex \2\1

• Click on the Replace All button

=> Your two html files should be modified, as below :

<!-- 1.html -->
Some text
before the
"noscript" tag
bla
bla
bla
blah
<img… and …style="display:none;"></div>
<noscript class"text" lang="example">
line1
line2
line3
line4
</noscript>
And some text,
located after
............
<noscript>
............

and

<!-- 2.html -->
bla
bla
<noscript>
bla

Other text…
etc…
etc…
etc…
<img… other and …style="display:none;"></div>
<noscript class"text" lang="otherexample">
Other line1
Other line2
Other line3
Other line4
Other line5
</noscript>
Other Text
..........
till the END of the file

I do hope that it’s the kind of Search/Replacement what you expect to !

Notes :

• I suppose that the ending tag </noscript> ends the current line, in your html files

• I suppose the the text <img………style="display:none;"></div> ends the current line, too

• As usual, the (?s) modifier means that a dot regex character represents any single character ( Standard or EOL chars )

• Then, the part <noscript class.+?</noscript>\R looks for any complete individual noscript block ( Note that the form .+? stands for the smallest non-null range of characters, between the two strings <noscript class and </noscript> )

• This noscript block, with the final EOL characters ( \R ), is stored as group 1, due to the enclosed parentheses

• Finally the part .+?style="display:none;"></div>\R searches for the smallest range of characters till the string style=“display:none;”></div>

• Again, this block of lines, with the final EOL characters ( \R ), is stored as group 2, due to the enclosed parentheses

• In replacement, we just rewrite these two groups, in a reverse order ! ( \2\1 )

Remarks :

• As previously said, for 400 files about, you should use the Replace in Files feature ( Ctrl + Shift + F )

• Group all your html files, which have to be modified, in a particular folder

• Do a BACKUP of this folder ( One never knows ! )

• Fill up the Filters and Directory boxes, too

• Click on the Replace in Files button, to perform a global replacement, on all your files

• IMPORTANT : Do NOT use the v7.4 or v7.4.1 version of N++. Indeed, with these last versions, there a bug, when multiple S/R are performed : a part of the files scanned are not changed -:((( Refer to :

https://notepad-plus-plus.org/community/topic/13865/replace-all-in-all-open-documents-problem-in-7-4-1/1

Cheers,

guy038

# -*- coding: utf-8 -*-

import os
FIND_WHAT = u'FIND_WHAT_REGULAR_EXPRESSION'
REPLACE_WITH = u'REPLACE_WITH_EXPRESSION'

DIRECTORY_OF_FILES = u'THE_DIRECTORY'   # like c:\\temp\\directory - doublebackslash needed

os.chdir(DIRECTORY_OF_FILES)
list_of_files = [x for x in os.listdir('.') if x.endswith('.html')]
for file in list_of_files:
    notepad.open(file)
    editor.rereplace(FIND_WHAT, REPLACE_WITH)
    notepad.save()
    notepad.close()

This script looks for html files in a defined directory.
If your files are under certain different subdirectories we need to modify the code.

Cheers
Claudia

in dos do you know how to join in one command each file of a directoty with itself
for one file its easy
copy name.htm +name.htm d:\temp\name.htm

time to sleep for me i shall search for this tomorrow
*do you know this french free texteditor http://www.gbesoft.fr/gbepad.php
its not bad, i use it in association wth notepad++

• Find = (<noscript[^>]*?class=*["“][^"“”]*?["”][^>]*?lang=["“][^"“”]*?["”]>.*?</noscript>\R*)(.*?)(<img[^>]*?style=["“]display:none;["”]></div>)
• Replace = $2$3\r\n$1
• Assumptions/Changes
  • I used [>]*? to include various unknown attributes, assuming they don’t contain a >: if they do, this won’t work
  • I used expressions that were ambivalent about smart quotes “” and normal quotes "", because you switched back and forth. Personally, I doubt there are actual smart quotes for the attributes in your HTML, but this allows my expression to work on exactly the text that you put in your examples, or on text that I think is well-formed HTML
  • Similary, I used class=* to allow class="text" or class"text", because I believe class"text" is invalid HTML, but I wanted the expression to work on your example text, just in case you are processing invalid HTML.
  • I allowed anything (or, rather, any non-quote character) inside the class="text" and lang="example", so it’s not restricted to exactly text or example: I saw you had example and otherexample, so I wanted to make sure that text and othertext would both work, too.
  • Since there is a language barrier, I am assuming by “to move = déplacer” you mean “move it, completely deleting the block from the beginning, and instead putting it after the <img display:none;">...</div>”

So that should work manually on the two example files you gave me: when I copy/pasted from your post into two files, then ran the exact FIND and REPLACE listed above once per file, it successfully found the noscript and moved it after the img.

HOWEVER (and this is a big “however”), that won’t be fun for hundreds of files. I tried having NPP open with just the two example files, then I used the Replace All in All Opened Documents option, and it replaced it in both files. Then I copied the originals so I had 100 files, and I opened them ALL in Notepad++, and the “Replace All in ALl Opened Documents” still did what I wanted. So it might work if you just open all 400 hundred files, though there’s no guarantee: I might do it in smaller groups of 50-100 files, depending on how large your files are.

If you are going to be frequently making changes like this, but the rules are changing, I highly recommend learning a scripting language: I would have done this from the command line with Perl; many others here would have used Python or Lua. The benefit of the latter two is that there are plugins for automating Notepad++ using the Python Script Plugin or the LuaScript Plugin. @Claudia-Frank could proably write the Python Script to open each of the 400 files, apply my regex, and save and close them without much difficulty. Unfortunately, I am not a Python expert, so it would take me hours more than I have available to muddle through it. (It may even be that @go2to’s ecobyte-link provides such a scripting language, or a simple interface for applying a particular regex to a multitude of files.)

my files are like this

1.html
<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>
some text…
etc…
etc
<img… and …style=“display:none;”></div>

2.html
<noscript class"text" lang=“otherexample”>
otherline1
otherline2
otherline3
otherline4
</noscript>
othersome text…
etc…
etc etc
<img… other and …style=“display:none;”></div>

i would to move (to move = déplacer in french) in each file all this <noscript… tag with the tag (generally inside there are 4 lines… can be 5)) i would to move it in each page just after style=“display:none;”></div>
and i would to do this for more than 400 files in knowing that all my noscript are differents in each page
warning; in each file i have 2 noscript tag one is just <noscript> the other is like <noscript class"text" lang=“example”>
its this second that i would to move

not easy :-)

Checkboxes: `☐ ☑ ☒ ✓ ✗`

Windows’ charmap.exe > Character Set: Unicode > Group By: Unicode Subrange > Symbols & Dingbats has the checkboxes in the first row when I’m using Font: DejaVu Sans Mono.

I also like using http://www.fileformat.info/info/unicode/char/search.htm to find Unicode characters by name or idea (like “arrow” or “box” or “check”). (I’ve actually got it as one of my custom search engines in Firefox, with the keyword “unicode”, so I just have to type “unicode ballot box” in my address bar to get http://www.fileformat.info/info/unicode/char/search.htm?q=ballot+box&preview=entity.

How did you make the little checkboxes in text? That’s pretty cool. I’ve read about some of the markdown used on this forum, but not all of it is clear to me. Not that I’m saying the checked-checkboxes are markdown…

I’m talking about this:
☑ Regular Expression

How would one do an empty checkbox as well as a checked one?

Thanks!

For example, I am assuming @kat75 wanted to take a file like

<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>

... lots of document ...

<pre>
-=-=-= REPLACE THIS SPECIFIC LINE WITH WHATEVER WAS IN THE EXAMPLE NOSCRIPT =-=-=-
</pre>

and replace it with

<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>

... lots of document ...

<pre>
line1
line2
line3
line4
</pre>

This could be done manually as I described before (using the previous FIND string, manually copying, manually moving the cursor, and manually pasting).

Or, as I added, it could be done with a Search and Replace, which I didn’t supply at the time. I was hoping @kat75 would chime in with more details, so I wouldn’t spend more time on a solution to my interpretation of the problem, without further input…

But since you seem to think NPP is not capable of this task, when it is: Using my new example file above, my interpretation of the resultant file desired would be

<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>

... lots of document ...

<pre>
line1
line2
line3
line4
</pre>

This can be accomplished in Notepad++ using a search-and-replace:

• Find What: (<noscript.*?>\s*)(.*?)(\s*</noscript>.*)(^-=-=-= REPLACE THIS SPECIFIC LINE WITH WHATEVER WAS IN THE EXAMPLE NOSCRIPT =-=-=-$)
• Replace with: $1$2$3$2
• ☑ Regular Expression and ☑ . matches newline
• REPLACE

If, instead, @kat75 wanted the entire noscript, including the opening and closing tags, a simple modification will do the trick:

• Find What: (<noscript.*?>.*?</noscript>)(.*)(^-=-=-= REPLACE THIS SPECIFIC LINE WITH WHATEVER WAS IN THE EXAMPLE NOSCRIPT =-=-=-$)
• Replace with: $1$2$1
• ☑ Regular Expression and ☑ . matches newline
• REPLACE

which would result in

<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>

... lots of document ...

<pre>
<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>
</pre>

If @kat75 wants the noscript contents or the full pair to be able to go after the indicated line rather than replacing it:

<noscript class"text" lang=“example”>
line1
line2
line3
line4
</noscript>

... lots of document ...

<pre>
-=-=-= INSERT TEXT AFTER THIS LINE =-=-=-
</pre>

Then to get the contents without the tags, use:

• Find What = (<noscript.*?>\s*)(.*?)(\s*</noscript>.*)(^-=-=-= INSERT TEXT AFTER THIS LINE =-=-=-$)
• Replace With = $1$2$3$4\r\n$2

or to include the noscript tags as well:

• Find What = (<noscript.*?>.*?</noscript>)(.*)(^-=-=-= INSERT TEXT AFTER THIS LINE =-=-=-$)
• Replace With = $1$2$3\r\n$1

If none of these accomplish what @kat75 wants, then @kat75 needs to chime in with example “before” and “after” text, so that we don’t have to guess.

(And I’m sure @guy038 would be able to improve on my regex…)
[edit: added paragraph break]

The OP kat75 let us know he had difficulties in English. deplace is not in the English dictionary. I believe he means replace as in range search-replace.

kat75 might want to follow my advice above and google ecobyte replace text. Hopefully, kat75 has a friend who is fluent in both English and their native language. kat75 should experiment with range search-replace in a temp directory first.

Copy a few html files to C:\Temp and test there until the results are as desired. That’s how I did and everything worked flawlessly.