what are list Regex available "named capture group" for find and replace in current version
-
@Jergen-Ross-Estaco said in what are list Regex available "named capture group" for find and replace in current version:
What is version for RegEx?
Notepad++ has been using Boost 1.78 since NPP v8.2, and 1.76 before that.
(I am not the expert on the details of named capture groups, so I’ll let someone else handle the more technical aspects of your issue)
-
-
Hello, @jergen-ross-estaco, @peterjones and All,
First, I advice you to read these two posts, first :
https://community.notepad-plus-plus.org/post/15930
https://community.notepad-plus-plus.org/post/52715
As you can see, in replacement, the following regexes :
-
\g{size} \g{size_type} | \g{file}\g{type}\r\n\t\g{path}\r\n -
\k{size} \k{size_type} | \k{file}\k{type}\r\n\t\k{path}\r\n -
\g<size> \g<size_type> | \g<file>\g<type>\r\n\t\g<path>\r\n -
\k<size> \k<size_type> | \k<file>\k<type>\r\n\t\k<path>\r\n
are totaly invalid !
The only BOOST valid syntax, for named capturing groups, in the replacement regex, is :
$+{size} $+{size_type} | $+{file}$+{type}\r\n\t$+{path}\r\n
Here is, below, your search regex, slighly modified, using the free-spacing mode
(?x), which can easily be modified, if necessary :In some parts, I used non-capturing groups, in order that you may use numbered groups, instead of the named groups, in the replacement regex
SEARCH
(?x-s) # FREE-SPACING mode + DOT matches STANDARD chars ONLY, not EOL ^ [\t\x20]* # OPTIONAL range of TAB/SPACE chars (?<size> # BEGINNING of the first NAMED group "size" [ or (?'size' ] (?: # NON-CAPTURING group [+-]? # OPTIONAL sign + or - (?: # NON-CAPTURING group [0-9]{1,3} # From 1 to 3 DIGITS (?: # NON-CAPTURING group ,[0-9]{3} # COMMA followed with THREE digits )* # END, REPEATED from 0 to MORE | # OR \d+ # ONE or MORE digits ) # END (?: # NON-CAPTURING group \. [0-9]+ # a DOT followed with some DIGITS )? # END of the OPTIONAL part ) # END | # OR \d* \. \d+ # OPTIONAL INTEGER part followed with a DOT and some DIGITS ) # END of the first NAMED group [\t\x20]* # OPTIONAL range of TAB/SPACE chars (?<size_type> # BEGINNING of the second NAMED group "size_type" [ or (?'size_type' ] (?i) gb|mb|m|g # String 'gb' or 'mb' or 'm' or 'g', INSENSITIVE ) # END of the second NAMED group [\t\x20]* # OPTIONAL TAB or SPACE characters (?<path> # BEGINNING of the third NAMED group "path" [ or (?'path' ] (?i:C|D): .+ \\ # EXACT string 'C' or 'D' followed by the GREATEST range of STANDARD chars till the LAST ANTI-SLASH of CURRENT line ) # END of the third NAMED group (?<file> # BEGINNING of the fourth NAMED group "file" [ or (?'file' ] .+ # The GREATEST range of STANDARD chars till... ) # END of the fourth NAMED group (?<type> # BEGINNING of the fifth NAMED group "type" [ or (?'type' ] \. .+ # The LAST '.' char of CURRENT line, followed with the GREATEST range of STANDARD chars, till the END of LINE ) # END of the fifth NAMED groupREPLACE
$+{size} $+{size_type} | $+{file}$+{type}\r\n\t$+{path}\r\nor
REPLACE
\1 \2 | \4\5\r\n\t\3\r\n
Note that, without the free-spacing mode
(?x), the search regex becomes :(?-s)^[\t\x20]*(?<size>(?:[+-]?(?:[0-9]{1,3}(?:,[0-9]{3})*|\d+)(?:\.[0-9]+)?)|\d*\.\d+)[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D):.+\\)(?<file>.+)(?<type>\..+)So, from this INPUT text :
5.7 GB D:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkvwhatever the search and replacement regex used, you should always get this OUTPUT text :
5.7 GB | Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkv D:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\If you decide to use the regex with the free-spacing mode, simply select all the text between
(?x-s)and# END of the fifth NAMED group(2,045bytes ). Note that the maximum of bytes, for the search field, is limited to2,046bytes !
In summary, the
6syntaxes\g{name},\g<name>,\g'name',\k{name},\k<name>and\k'name'are not deprecated but can only be used in the search part ( not in replacement ) !Best Regards
guy038
-
-
@guy038 said in what are list Regex available "named capture group" for find and replace in current version:
As you can see, in replacement, the following regexes … are totally invalid
I should have remembered that. We even added a very clear note in the replacement section of the user manual:
Please note: the \g… and \k… backreference syntaxes only work in the search expression, and are not designed or intended to work in the substitution/replacement expression.
There is a similar note in the searching section’s “numbered backreference”:
Numbered Backreference: These syntaxes match the ℕth capture group earlier in the same expression. (Backreferences are used to refer to the capture group contents only in the search/match expression; see the Substitution Escape Sequences for how to refer to capture groups in substitutions/replacements.)
… but apparently not in the [“named backreference”](Named Backreference)… I guess it needs to be spelled out there, too, even thought it’s just a few lines down in the docs.
-
@guy038 and @PeterJones.
First, I advice you to read these two posts, first :
I have already read since I tired searched before this post
I didn’t read non-capturing, am I missed read?,
also I didn’t know\x20instead\s.\x20is necessary aka space character? but why\s?you wrote “x” from
?x-s(?x-s) # FREE-SPACING mode + DOT matches STANDARD chars ONLY, not EOLbut you didn’t write “x” from
(?-s)(?-s)^[\t\x20]*(?<size>(?:[+-]?(?:[0-9]{1,3}(?:,[0-9]{3})*|\d+)(?:\.[0-9]+)?)|\d*\.\d+)[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D):.+\\)(?<file>.+)(?<type>\..+)I like that name group with start group subexp
example like my regex previous:
(?<name>subexp)and your regex
(?<name>(subexp))btw, now I can’t figure out when search can’t match part line
4.0 GB C:\pagefile.sys 1.5 GB C:\hiberfil.sysalso can’t match file type will skipped line because they don’t have extension like
100.0 MB C:\Users\Username\AppData\Local\Google\Chrome Beta\User Data\Default\Cache\Cache_Data\data_3you can help me?
If you can’t? please disregard this, fine.If you decide to use the regex with the free-spacing mode, simply select all the text between (?x-s) and # END of the fifth NAMED group ( 2,045 bytes ). Note that the maximum of bytes, for the search field, is limited to 2,046 bytes !
How I do check get bytes in search field?
anyway, thank you for help and answer.
-
Hi, @jergen-ross-estaco,
Before answering your questions, in a next post, could you tell me which of the four cases, below, will never happen ?
A 5.7 GB D:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkv B 5.7GB D:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkv C 5.7 GBD:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkv D 5.7GBD:\Movies by Jen\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP\Memory.2022.1080p.WEBRip.DD5.1.x264-NOGRP.mkv
I also revisited my post and change the part of the regex regarding the size, as below :
(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})*|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))For instance, this regex matches all the following cases, from
AtoZand@A 1 E +1 I -1 B 12345 F +12345 J -12345 C 1234.0123 G +1234.0123 K -1234.0123 D 123456.789 H +123456.789 L -123456.789 M 1,234 P +1,234 S -1,234 N 12,345.01 Q +12,345.01 T -12,345.01 O 1,234,567.0123 R +1,234,567.0123 U -1,234,567.0123 V .0 X +.0 Z -.0 W .01234 Y +.01234 @ -.01234In the same way, could you tell me which cases are sure to not happen ?
I’ll provide you a correct and final regex solution, very soon !
See you later,
Best Regards,
guy038
-
@guy038, @PeterJones and all.
I trying what I said my explain. you know I’m deaf because slow learn, mistake grammar and sentence or bad? but I hope u understand me.
Before answering your questions, in a next post, could you tell me which of the four cases, below, will never happen ?
I tested use regex to four cases are look fine working because ‘
[\t\x20]’ if no space will no failed and next step. but look similar ‘\s’.I also revisited my post and change the part of the regex regarding the size, as below :
(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})*|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))
For instance, this regex matches all the following cases, from A to Z and @
…
In the same way, could you tell me which cases are sure to not happen ?‘
A,E,I,M-Zand@’ are correctnot all, ‘
B-C,F-H, andJ-L’ are match two group that “123” and any digit. if when you use find and replace “foo” will result is “foofoo” meanwhile “123” and “4.0123” to “foo” “foo”input and output:
A 1 E +1 I -1 B 12345 F +12345 J -12345 C 1234.0123 G +1234.0123 K -1234.0123 D 123456.789 H +123456.789 L -123456.789 M 1,234 P +1,234 S -1,234 N 12,345.01 Q +12,345.01 T -12,345.01 O 1,234,567.0123 R +1,234,567.0123 U -1,234,567.0123 V .0 X +.0 Z -.0 W .01234 Y +.01234 @ -.01234 A foo E foo I foo B foofoo F foofoo J foofoo C foofoo G foofoo K foofoo D foofoo H foofoo L foofoo M foo P foo S foo N foo Q foo T foo O foo R foo U foo V foo X foo Z foo W foo Y foo @ fooI fixed replaced ‘
*’ to ‘+’ from where is end of comma and three digit group ‘(?:,[0-9]{3})+’ will matched “1234.0123”.(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})+|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))if
*is zero or more if no comma or more will match passed.
+is one or more which required at least one comma or more if no comma will next step to ‘|\d+’ is alternative match any number and decimal if don’t have comma separated number instead.input and output result: all matched without double “foo” also Replace “number” to “foo” are corrected
A 1 E +1 I -1 B 12345 F +12345 J -12345 C 1234.0123 G +1234.0123 K -1234.0123 D 123456.789 H +123456.789 L -123456.789 M 1,234 P +1,234 S -1,234 N 12,345.01 Q +12,345.01 T -12,345.01 O 1,234,567.0123 R +1,234,567.0123 U -1,234,567.0123 V .0 X +.0 Z -.0 W .01234 Y +.01234 @ -.01234 A foo E foo I foo B foo F foo J foo C foo G foo K foo D foo H foo L foo M foo P foo S foo N foo Q foo T foo O foo R foo U foo V foo X foo Z foo W foo Y foo @ foobtw, I changed that size put my regex.
I realized mentioned that
btw, now I can’t figure out when search can’t match part line
4.0 GB C:\pagefile.sys 1.5 GB C:\hiberfil.sysI mean There is no folder or it’s drive letter but I need figure out match ‘
c:/’ for ‘(?<path>)’ and now,
I have solved my myself that removed ‘:’ before dot character from ‘(?i:C|D):.+\\)’ because already “+” is one or more, this is at least match one character which is matched ‘\’ will start current position after ‘\’ when ‘\\’ will searching ‘\’ character until start current position but not found match ‘\’ error , so removed ‘:’ become when ‘+’ is matched to ‘:’ will start current position after ‘:’, found match ‘\’ from ‘C:\’ It will worked match zero folder or more level. (c:\ or c:\folder1\subfolder\subfolder2\...)Regex:
(?<path>(?i:C|D).+\\)and again I mentioned that
also can’t match file type will skipped line because they don’t have extension like
100.0 MB C:\Users\Username\AppData\Local\Google\Chrome Beta\User Data\Default\Cache\Cache_Data\data_3I added lookaround and alternative ‘
(?<file>.+(?=\.)|.+)(?<type>(?:\..+)?$)’ will match filename if have extension or not.(?<file>...)
.+(?=\.)is one or more any character before positive lookahead with match ‘.’ dot character. It’s match filename before dot character as file extension if no dot character will next step to ‘|.+’ is alternative match filename if don’t have extension instead.(?<type>...)
\..+)?$is optional dot character and one or more any character. as optional file extension.
I finally made my regex :
(?-xs)^[\t\x20]*(?<size>[+-]?(?:(?:\d{1,3}(?:,\d{3})+|\d+)(?:\.\d+)?|\.\d+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D).+\\)(?<file>.+(?=\.)|.+)(?<type>(?:\..+)?$) or (?-xs)^[\t\x20]*(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})+|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D).+\\)(?<file>.+(?=\.)|.+)(?<type>(?:\..+)?$)completed found all matched lines 100/100 in my input text.
4.0 GB C:\pagefile.sys 1.8 GB C:\Users\Jhecrose\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG.mp4 1.5 GB C:\hiberfil.sys 900.1 MB C:\Games\MBAACC\0002.p 851.8 MB C:\$MFT 569.6 MB C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\1504837401.vpkafter replaced, result output:
4.0 GB | pagefile.sys C:\ 1.8 GB | Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG.mp4 C:\Users\Username\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\ 1.5 GB | hiberfil.sys C:\ 900.1 MB | 0002.p C:\Games\MBAACC\ 851.8 MB | $MFT C:\ 569.6 MB | 1504837401.vpk C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\I’m sorry if you confuse when you read this.
thank you, so much! this case is close now but I’m waiting for my pervious question. -
Hello, @jergen-ross-estaco, @peterjones and All,
I ended up with this search regex :
SEARCH
(?-s)^[\t\x20]*(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})+|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D):.*\\)(?<file>.+)(?<type>\..+)?This regex is slightly shorter than yours :
-
I removed the useless
xmodifier, at beginning of the regex, as the regex is mono line ! -
I kept your regexes regarding the
<size>and<size_type>named groups -
I modified the
<path>,<file>and<type>named groups as below :-
I used a
*after the part(?i:C|D):, needed when files are located right under the root -
I added a
?, right after the named group(?<type>\..+), for the case of files without extension
-
Like you, my regex version matches all the cases of your INPUT file :
4.0 GB C:\pagefile.sys 1.8 GB C:\Users\Jhecrose\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG.mp4 1.5 GB C:\hiberfil.sys 900.1 MB C:\Games\MBAACC\0002.p 851.8 MB C:\$MFT 569.6 MB C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\1504837401.vpk
Now, regarding your question :
How I do check get bytes in search field?
It’s quite easy : Select a range of characters and, simply, look right after the indication
Sel :, in the status bar, at bottom of the Notepad++ window !Best Regards
guy038
-
-
Hi @guy038,
(?<type>\..+)?Same I tried but I have problem
(?<type>)won’t match instead both file and extension will match(?<file>)cause this is one.+will start current position to end of line will(?<type>)is not found match which is searching to current end of line or won’t search back to steps character.try test what’s going happen:
Replace:File:\x20\4\r\ntype:\x20\5\r\nexpected behavior:
I used my regex Find:
(?-s)^[\t\x20]*(?<size>[+-]?(?:(?:\d{1,3}(?:,\d{3})+|\d+)(?:\.\d+)?|\.\d+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D).+\\)(?<file>.+(?=\.)|.+)(?<type>(?:\..+)?)$File: The.Sea.Beast.2022.1080p.WEBRip.x264-RARBG type: .mp4 File: The.Unbearable.Weight.of.Massive.Talent.2022.1080p.WEBRip.x264-RARBG type: .mp4 File: Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG type: .mp4 File: hiberfil type: .sysactual behavior:
used your regex:
File: The.Sea.Beast.2022.1080p.WEBRip.x264-RARBG.mp4 type: File: The.Unbearable.Weight.of.Massive.Talent.2022.1080p.WEBRip.x264-RARBG.mp4 type: File: Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG.mp4 type: File: hiberfil.sys type:you have tried debugger in regex? better this, I find out happen regex getting error or problem. btw, I’m using two regex editor are:
- for regex101.com, quick hilghlight match and group and very easily debugger show one by one steps is useful
- for notepad++, before use test output depends their boost engine.
It’s quite easy : Select a range of characters and, simply, look right after the indication Sel :, in the status bar, at bottom of the Notepad++ window !
uh, I was just looking
sel :is just number?! or is that bytes? I didn’t seebytes -
Hi, @jergen-ross-estaco, @peterjones and All,
You are perfectly right about my regex : it did not respect the named groups :-((
I finally succeeded to build a correct search regex, (
2characters longer than yours, if I subsitute the\dby the[0-9]syntax ! )It uses a particular feature, not very-well known : the
Branch Resetmechanism, with the(?|pattern_1|pattern_2|....|pattern_N)syntaxRefer to this link for further explanations :
So, given this INPUT text :
4.0 GB C:\pagefile.sys 1.8 GB C:\Users\Jhecrose\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG.mp4 1.5 GB C:\hiberfil.sys 900.1 MB C:\Games\MBAACC\0002.p 851.8 MB C:\$MFT 569.6 MB C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\1504837401.vpk 1.8 MB C:\pqr.tuv.xyz.123 1.8 MB C:\xyz.123 1.8 MB C:\xyz 1.8 MB C:\abc\def\ghi\pqr.tuv.xyz.123 1.8 MB C:\abc\def\ghi\xyz.123 1.8 MB C:\abc\def\ghi\xyzThe following regex S/R :
SEARCH
(?-s)^[\t\x20]*(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})+|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D):.*\\)(?|(?<file>.+)(?<type>\..+)|([^.\r\n]+))REPLACE
Path : \3\r\nFile : \4\r\nExtension : \5\r\nwill give this OUTPUT text :
Path : C:\ File : pagefile Extension : .sys Path : C:\Users\Jhecrose\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\ File : Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG Extension : .mp4 Path : C:\ File : hiberfil Extension : .sys Path : C:\Games\MBAACC\ File : 0002 Extension : .p Path : C:\ File : $MFT Extension : Path : C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\ File : 1504837401 Extension : .vpk Path : C:\ File : pqr.tuv.xyz Extension : .123 Path : C:\ File : xyz Extension : .123 Path : C:\ File : xyz Extension : Path : C:\abc\def\ghi\ File : pqr.tuv.xyz Extension : .123 Path : C:\abc\def\ghi\ File : xyz Extension : .123 Path : C:\abc\def\ghi\ File : xyz Extension :If you prefer to use the named groups in replacement, the following regex S/R :
SEARCH
(?-s)^[\t\x20]*(?<size>[+-]?(?:(?:[0-9]{1,3}(?:,[0-9]{3})+|[0-9]+)(?:\.[0-9]+)?|\.[0-9]+))[\t\x20]*(?<size_type>(?i)gb|mb|m|g)[\t\x20]*(?<path>(?i:C|D):.*\\)(?|(?<file>.+)(?<type>\..+)|([^.\r\n]+))REPLACE
Path : $+{path}\r\nFile : $+{file}\r\nExtension : $+{type}\r\nwill return the same OUTPUT text :
Path : C:\ File : pagefile Extension : .sys Path : C:\Users\Jhecrose\Downloads\uTorrent\Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG\ File : Paws.of.Fury.The.Legend.of.Hank.2022.1080p.WEBRip.x264-RARBG Extension : .mp4 Path : C:\ File : hiberfil Extension : .sys Path : C:\Games\MBAACC\ File : 0002 Extension : .p Path : C:\ File : $MFT Extension : Path : C:\Program Files (x86)\Steam\steamapps\common\Left 4 Dead 2\left4dead2\addons\workshop\ File : 1504837401 Extension : .vpk Path : C:\ File : pqr.tuv.xyz Extension : .123 Path : C:\ File : xyz Extension : .123 Path : C:\ File : xyz Extension : Path : C:\abc\def\ghi\ File : pqr.tuv.xyz Extension : .123 Path : C:\abc\def\ghi\ File : xyz Extension : .123 Path : C:\abc\def\ghi\ File : xyz Extension :
To end with, in the status bar, the
Sel :field gives you the number of characters ( not bytes) of the current selection !BR
guy038
P.S. :
So, in the part
(?|(?<file>.+)(?<type>\..+)|([^.\r\n]+)):-
(?<file>.+)represents the named groupfile( group4) -
(?<type>\..+)represents the named grouptype( group5) -
([^.\r\n]+)represents the group4, too, due to the branch reset syntax
-
-
@guy038 said in what are list Regex available "named capture group" for find and replace in current version:
Refer to this link for further explanations :
https://www.boost.org/doc/libs/1_78_0/libs/regex/doc/html/boost_regex/syntax/perl_syntax.html#boost_regex.syntax.perl_syntax.branch_resetWhy not this LINK as well?:
