JREPL.BAT v8.6 - regex text processor with support for text highlighting and alternate character sets

[dbenham]

Yes, I'm aware that the links are outdated. But I am reluctant to make changes because

• The old links continue to work, and they provide a link (button) to access the current version
• Making a change would require a new version to JREPL
• I find the old pages a bit more convenient than the new pages

I probably won't make any changes until either the old links become invalid, or if there are important functional differences described in the new pages.


Dave Benham

[colossus02]

Hey Dave,
Great Program!!!

Is there a way for this to query a list of search text and the corresponding replace text to make the change??

Example: I have a file with different countries. I need to search each country and replace with the corresponding code for each. Each country needs to be changed to a different code. But some country have the same code. I have the list in a text file of the countries and codes.

[dbenham]

There is nothing built in to do it in one step. But it is easy (and efficient) to accomplish the task with a simple batch script using two JREPL calls.

The ultimate goal is to use the /T and /V options to perform a series of find/replace operations with a single JREPL call, with the search strings specified by a FIND variable, and replace strings in REPL. The trick is to use a first JREPL call with FOR /F to parse the county/code text file and create the two variables.

Note there are two things that limit the number of find/replace pairs:

1) The number of pairs cannot exceed 99
2) The total length of each variable cannot exceed ~8190 bytes.

I'll assume your first file has one line per country/code pair, with the values delimited by a | character.

Code: Select all

@echo off
setlocal

:: Load FIND and REPL terms
for /f "delims=" %%A in (
  'jrepl "(.*?)\|(.*)" ^
         "find+=$1+'|';repl+=$2+'|';false" ^
         /jmatch ^
         /jbeg "repl='';find=''" ^
         /jend "stdout.WriteLine('find='+find.slice(0,-1)+'\nrepl='+repl.slice(0,-1))"
         /f "strings.txt"'
) do set "%%A"
set find
set repl

:: Execute the find/replace
call jrepl find repl /v /l /i /t "|" /f "input.txt" /o "output.txt"

If you have more countries than can be processed with /T and /V, then you will have to do multiple independent find/replace operations, but then you have a problem if any of your codes happen to match a country name, and it will be much slower.

Code: Select all

@echo off
copy /y "input.txt" "output.txt"
for /f "usebackq delims=| tokens=1,2" %%A in (
  "strings.txt"
) do call jrepl "%%A" "%%B" /l /i /f "output.txt" /o -

Dave Benham

[dbenham]

JREPL normally writes lines using \r\n as the line terminator (as is the Windows standard).
Version 5.0 adds the /U option to give you the ability to write lines in Unix format using \n as a line terminator.

JREPL.BAT Version 5.0

JREPL5.0.zip

(12.53 KiB) Downloaded 1232 times

Dave Benham

[dbenham]

I have augmented the behavior of the /T option:
The DelimiterChar can now be specified as the word FILE, which causes the option to treat the Search and Replace parameters as file paths.
The actual search and replace expressions are read from the respective files, one expression per line.
This feature was developed in response to a question from colossus02.

JREPL version 5.2

JREPL5.2.zip

(12.89 KiB) Downloaded 1118 times

Dave Benham

[colossus02]

Thanks so much Dave!! Sorry it took me so long to respond but this worked perfectly!! I have 7 string files to process during the find and replace and it doesn't take more then 2-3 minutes. I believe its fast because the server I run this on.

[dbenham]

I've updated my previous post from version 5.1 to 5.2.

I fixed a bug with the /T FILE option so that Search and Replace now properly ignore the /V option.

I also added an example to the documentation that shows how to use /T FILE.


Dave Benham

[dbenham]

Here is a major new release, version 6.2, with lots of new features.

JREPL6.2.zip

(16.53 KiB) Downloaded 1125 times

Many of the the new features are designed to reduce the need for user supplied JScript, and/or to facilitate use of simpler regular expressions to solve a particular task. Performance is also improved.

I'll break the new features into numbered sections, based on common themes. I'll only provide a brief summary of each new feature, plus examples. See the documentation for the complete syntax and behavior.

1) Improved Help system

New HISTORY topic

Prints out the full release history of JREPL.BAT.
Check out the lines next to 2016-10-08 v6.0: for a summary of all the new features in this release.

Code: Select all

>jrepl /?history

    2016-10-13 v6.2: Bug fix - /J, /JQ, /JMATCH, /JMATCHQ did not work with /P.
    2016-10-08 v6.1: Bug fix - v6.0 broke /JBEG and /JLIB, all fixed.
    2016-10-08 v6.0: Added /K - search and write matching lines.
                     Added /R - search and write non-matching lines.
                     Added /MATCH - search and write each match on a new line.
                     Added /P - Pre-filter regex before normal search/replace.
                     Added /PFLAG - set search flags for /P regex
                     Added /JQ and /JMATCHQ as Quick forms of /J and /JMATCH.
                     Augmented /INC and /EXC so can now specify lines by regex.
                     Changed behavior - /V now applies to /INC and /EXC.
                     Improved performance of /INC, /EXC, /T, /JBEGLN, /JENDLN.
                     Added HISTORY and UPDATE topics to the help system.
    2016-09-27 v5.2: Bug fix - Search & Replace now ignore /V if /T FILE used.
                     Added a /T FILE example to the documentation.
    2016-09-20 v5.1: Added the FILE alternative for the /T option.
    2016-09-18 v5.0: Added the /U option for Unix line terminators of /n.
    2016-08-04 v4.6: Fixed the /N documentation (reparired missing line)
    2016-08-03 v4.5: Added /D option to specify delimiter for /N and /OFF.
    2016-08-02 v4.4: Bug fix - /C count was wrong when last line did not end
                     with new line. This also affected /INC and /EXC.
    2016-07-30 v4.3: Added rpad() function and improved lpad()
    2016-06-24 v4.2: Improved the /?Options help.
    2016-06-23 v4.1: Added /T option examples to the help.
                     Added ability to request help on a single option or topic.
    2016-06-19 v4.0: Added the /INC and /EXC options.
    2016-03-27 v3.8: Bug fix - Hide leaked global variables i, lib, libs, rtn2.
                     Bug fix - Work around %~f0 bug when command is quoted.
                     Bug fix - Use /OPTIONS instead of OPTIONS as a variable
                     name within the option parser so that it is unlikely to
                     collide with a user defined variable name.
    2016-01-14 v3.7: Reworked error handling a bit.
                     Bug fix - \xnn and \unnnn could fail in a regex search
                     if result was a meta-character and /X option was used.
    2015-07-15 v3.6: Added /?? option for paged help.
    2015-06-12 v3.5: Bug fix for $n or $nn in replace string when /T is
                     used without /J or /JMATCH or /L
    2015-01-22 v3.4: Bug fix - Use /TEST instead of TEST as a variable name
                     within the option parser so that it is unlikely to
                     collide with a user defined variable name.
    2014-12-24 v3.3: Bug fix for when /JMATCH is combined with /M or /S
    2014-12-09 v3.2: Bug fix for /T without /JMATCH - fixed dynamic repl func
                     Added GOTO at top for improved startup performance
    2014-11-25 v3.1: Added /JLIB option
                     Exception handler reports when regex is bad
                     Fix /X bug with extended ASCII
    2014-11-23 v3.0: Added /JBEGLN and /JENDLN options
                     Added skip, quit, and lpad() global variables/functions
                     Exception handler reports when error in user code
    2014-11-21 v2.2: Bug fix for /T with /L option.
    2014-11-20 v2.1: Bug fix for /T option when match is an empty string
    2014-11-17 v2.0: Added /T (translate) and /C (count input lines) options
    2014-11-14 v1.0: Initial release

Easily check if there is a new version available

Opens up the first post of this thread in your browser so you can easily check if there is a newer version available.

Code: Select all

>:: The following command will open up the first post of this thread in your browser
>jrepl /?update

2) Search only, without doing any replace (more like grep, or a glorified FINDSTR)

My syntax has always required that the first two arguments are the Search and Replace strings. I didn't want to break any existing scripts that use JREPL, so I opted not to change the syntax. So the Replace argument is still required, but it is simply ignored. The easiest thing to do is pass an empty string using "".

events.txt (test file)

Code: Select all

Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
08:00 AM  Event 2 - Where did it go?
09:00 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
11:00 AM  Event 5 - The strongest man alive
12:00 PM  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look
02:00 PM  Event 8 - Now that's gross
03:00 PM  Event 9 - Hold your nose

/K Context
/K PreContext:PostContext

Keep lines that contain at least one string that matches the Search. The Context is required. It determines how many non-mathching lines to display before and/or after each matching line.

Examples:

Code: Select all

>:: Write lines that begin with a capital letter
>jrepl "^[A-Z]" "" /k 0 /f events.txt
Session 1 - The mysteries of magic
Session 2 - You won't believe your eyes
Session 3 - Don't say you weren't warned

>:: Same as above, but include 1 non-matching line after each match
>jrepl "^[A-Z]" "" /k 0:1 /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look

>:: Same as above, but include 1 non-matching line before and after each match
>jrepl "^[A-Z]" "" /k 1 /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
09:00 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
12:00 PM  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look

>:: You can easily draw your eyes to matching lines among context if you add the /OFF option.
>jrepl "^[A-Z]" "" /k 1 /off 1 /d " " /f events.txt
0 Session 1 - The mysteries of magic
  07:00 AM  Event 1 - Abracadabra
  09:00 AM  Event 3 - How could you know that?
0 Session 2 - You won't believe your eyes
  10:00 AM  Event 4 - Ouch, that's gotta hurt
  12:00 PM  Event 6 - How can she bend like that?
0 Session 3 - Don't say you weren't warned
  01:00 PM  Event 7 - I can't bear to look

/R Context
/R PreContext:PostContext

Reject lines that contain at least one string that matches the Search. In other words, print out lines that don't match. The Context feature is perhaps counter-intuitive for non-matching lines, but it works the same as for /K. A non-zero value will print out matching lines before and/or after each non-matching line.

Examples:

Code: Select all

>:: Reject lines that begin with a digit
>jrepl "^\d" "" /r 0 /f events.txt
Session 1 - The mysteries of magic
Session 2 - You won't believe your eyes
Session 3 - Don't say you weren't warned

>:: Same as above, but include 1 matching line after each non-match
>jrepl "^\d" "" /r 0:1 /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look

>:: Same as above, but include 1 matching line before and after each match
>:: Note that /OFF will not work with /R
>jrepl "^\d" "" /r 1 /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
09:00 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
12:00 PM  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look

/MATCH

Write out each string that matches on a new line. Discard all other text.

Example:

Code: Select all

>:: List out all events, without time
>jrepl "Event.*" "" /match /f events.txt
Event 1 - Abracadabra
Event 2 - Where did it go?
Event 3 - How could you know that?
Event 4 - Ouch, that's gotta hurt
Event 5 - The strongest man alive
Event 6 - How can she bend like that?
Event 7 - I can't bear to look
Event 8 - Now that's gross
Event 9 - Hold your nose!

>:: Write each number and time on a new line, and show the location of each match
>jrepl "[\d:]+( [AP]M)?" "" /match /n 2 /off 2 /d . /f events.txt
01.08.1
02.00.07:00 AM
02.16.1
03.00.08:00 AM
03.16.2
04.00.09:00 AM
04.16.3
05.08.2:
06.00.10:00 AM
06.16.4
07.00.11:00 AM
07.16.5
08.00.12:00 PM
08.16.6
09.08.3:
10.00.01:00 PM
10.16.7
11.00.02:00 PM
11.16.8
12.00.03:00 PM
12.16.9

3) Limit searches (and replaces) to selected regions within the input text

Improved /INC and /EXC options

/INC and /EXC restrict searches to blocks of lines. They introduced in version 4.0, but blocks had to be identified by line numbers. Since this filtering is line based, it cannot be used with /S or /M options.

New features for version 6.0

• Specify lines and/or rangeBegin and/or rangeEnd by regular expressions
• Apply optional offset to any line position
• Specify rangeEnd by line offset from rangeBegin

Note - Because regular expressions frequently use quoted ^, which becomes ^^ when CALL is used, a method was needed to avoid putting ^ directly in the command line.

I opted to modify /V to make it also apply to /INC and /EXC. This new behavior has the potential to break an existing script.
In the unlikely event that you already have a version 4 or 5 script that uses /V with /INC or /EXC, then you will have to modify your script to put the /INC or /EXC value in a variable.

Examples:

Code: Select all

>:: List all morning events
>jrepl "" "" /k 0 /inc "/ AM /" /f events.txt
07:00 AM  Event 1 - Abracadabra
09:00 AM  Event 3 - How could you know that?
10:00 AM  Event 4 - Ouch, that's gotta hurt

>:: List the first event of each session, without displaying the session info
>jrepl "" "" /k 0 /inc "/^session/i+1" /f events.txt
07:00 AM  Event 1 - Abracadabra
10:00 AM  Event 4 - Ouch, that's gotta hurt
01:00 PM  Event 7 - I can't bear to look

>:: List the itineary for the first session, only knowing that session lines begin with a letter
>jrepl "" "" /k 0 /inc "/^[a-z]/i/:/^[a-z]/i-1" /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
08:00 AM  Event 2 - Where did it go?
09:00 AM  Event 3 - How could you know that?

>:: List each session, plus its first event
>jrepl "" "" /k 0 /inc "/^[a-z]/i:+1" /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look

>:: Change the start time to the half hour for all events in Session 2
>jrepl "^(\d\d):\d\d" "$1:30" /inc "/^Session 2/:/^Session/" /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
08:00 AM  Event 2 - Where did it go?
09:00 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:30 AM  Event 4 - Ouch, that's gotta hurt
11:30 AM  Event 5 - The strongest man alive
12:30 PM  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 PM  Event 7 - I can't bear to look
02:00 PM  Event 8 - Now that's gross
03:00 PM  Event 9 - Hold your nose!

>:: Change the start time to the half hour for all events except for those in Session 2
>jrepl "^(\d\d):\d\d" "$1:30" /exc "/^Session 2/:/^Session/" /f events.txt
Session 1 - The mysteries of magic
07:30 AM  Event 1 - Abracadabra
08:30 AM  Event 2 - Where did it go?
09:30 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
11:00 AM  Event 5 - The strongest man alive
12:00 PM  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:30 PM  Event 7 - I can't bear to look
02:30 PM  Event 8 - Now that's gross
03:30 PM  Event 9 - Hold your nose!

>:: Combine /INC and /EXC - List the morning events for Session 2
>jrepl "" "" /k 0 /inc "/^Session 2/+1:/^Session/-1" /exc "/ PM /" /f events.txt
10:00 AM  Event 4 - Ouch, that's gotta hurt
11:00 AM  Event 5 - The strongest man alive

>:: Use the offset feature when you already have important line numbers in variables
>:: and don't want to use SET /A to define another variable. In this case I want to
>:: list lines that appear after %begin%, and before %end%.
> set /a begin=5, end=9
9
>jrepl "" "" /k 0 /inc "%begin%+1:%end%-1" /f events.txt
01:00 PM  Event 4
02:00 PM  Event 5
03:00 PM  Event 6

/P FilterRegex

Only search/replace strings that match the FilterRegex. This can be used with any type of search or search/replace.
The /PFLAG option (not shown in examples) enables control over the g and i regular expression flags.

Multiline Mode Example:
My ADVENTURE.BAT port of collossal cave adventure selectively obfuscates code by applying the ROT13 cipher to text between curly braces. A simple JREPL command can quickly encode/decode the text, without requiring any user supplied JScript!

obfuscated.txt

Code: Select all

:: Note that { boshfpngrq grkg
:: pna fcna yvarf } Pretty cool!
::{ Npgvingr qjneirf jura Unyy bs Zvfgf vf ernpurq }
if !#{qjnesFgngr}! equ 0 (
  if !#{ybp}! geq 15 set /a #{qjnesFgngr}=1
  goto :{qrfpevorYbpngvba}
)

Selectively decode the text

Code: Select all

jrepl "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"^
      "nopqrstuvwxyzabcdefghijklmNOPQRSTUVWXYZABCDEFGHIJKLM"^
      /t "" /p "{[^}]*}" /m /f obfuscated.txt

--RESULT--

Code: Select all

:: Note that { obfuscated text
:: can span lines } Pretty cool!
::{ Activate dwarves when Hall of Mists is reached }
if !#{dwarfState}! equ 0 (
  if !#{loc}! geq 15 set /a #{dwarfState}=1
  goto :{describeLocation}
)

Normal line based examples (back to the events.txt source):

Code: Select all

>:: Selectively mask all Session and Event titles
>jrepl "." "*" /p "- .*" /f events.txt
Session 1 ************************
07:00 AM  Event 1 *************
08:00 AM  Event 2 ******************
09:00 AM  Event 3 **************************
Session 2 *****************************
10:00 AM  Event 4 *************************
11:00 AM  Event 5 *************************
12:00 PM  Event 6 *****************************
Session 3 ******************************
01:00 PM  Event 7 **********************
02:00 PM  Event 8 ******************
03:00 PM  Event 9 *****************

>:: /INC or /EXC (performed first) can be combined with /P (performed second)
>:: Only mask titles for Session 3
>:: Note how the /INC block continues to the end of file if the RangeEnd is not found
>jrepl "." "*" /inc "/^Session 3/:/^Session/" /p "- .*" /f events.txt
Session 1 - The mysteries of magic
07:00 AM  Event 1 - Abracadabra
08:00 AM  Event 2 - Where did it go?
09:00 AM  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 AM  Event 4 - Ouch, that's gotta hurt
11:00 AM  Event 5 - The strongest man alive
12:00 PM  Event 6 - How can she bend like that?
Session 3 ******************************
01:00 PM  Event 7 **********************
02:00 PM  Event 8 ******************
03:00 PM  Event 9 *****************

4) Improved Performance

In prior versions, many options made use of the JScript eval() function to execute user supplied code. This is comparatively quite slow, and especially has a large impact when it is used for each and every match.

I have managed to eliminate the use of eval() in any repetitive match code. I did this by dynamically defining a function with eval() only once, at the beginning, before any looping commences.

This gives a major performance boost to /T, /INC, and /EXC.

The code evaluated by eval() may contain many statements. The eval() function uses the value of the last statement (expression) as the return value. The /J and /JMATCH options rely on this behavior of eval(), so I could not modify how they are implemented.

So I added two new "Quick" options with similar functionality that take advantage of the new design.

/JQ - Functions the same as /J, except the code must store the return value in a variable named $txt.

/JMATCHQ - Functions the same as /JMATCH, except the code must store the return value in a variable named $txt.

Examples:

Code: Select all

>:: Old slow way with /J
>jrepl " [AP]M " "$0.toLowerCase()" /j /f events.txt
Session 1 - The mysteries of magic
07:00 am  Event 1 - Abracadabra
08:00 am  Event 2 - Where did it go?
09:00 am  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 am  Event 4 - Ouch, that's gotta hurt
11:00 am  Event 5 - The strongest man alive
12:00 pm  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 pm  Event 7 - I can't bear to look
02:00 pm  Event 8 - Now that's gross
03:00 pm  Event 9 - Hold your nose!

>:: New Quick way with /JQ
>jrepl " [AP]M " "$txt=$0.toLowerCase()" /jq /f events.txt
Session 1 - The mysteries of magic
07:00 am  Event 1 - Abracadabra
08:00 am  Event 2 - Where did it go?
09:00 am  Event 3 - How could you know that?
Session 2 - You won't believe your eyes
10:00 am  Event 4 - Ouch, that's gotta hurt
11:00 am  Event 5 - The strongest man alive
12:00 pm  Event 6 - How can she bend like that?
Session 3 - Don't say you weren't warned
01:00 pm  Event 7 - I can't bear to look
02:00 pm  Event 8 - Now that's gross
03:00 pm  Event 9 - Hold your nose!

>:: Old slow way with /JMATCH
>jrepl "Event .* - (.*)" "$1.toUpperCase()" /jmatch /f events.txt
ABRACADABRA
WHERE DID IT GO?
HOW COULD YOU KNOW THAT?
OUCH, THAT'S GOTTA HURT
THE STRONGEST MAN ALIVE
HOW CAN SHE BEND LIKE THAT?
I CAN'T BEAR TO LOOK
NOW THAT'S GROSS
HOLD YOUR NOSE!

>:: New Quick way with /JMATCHQ
>jrepl "Event .* - (.*)" "$txt=$1.toUpperCase()" /jmatchq /f events.txt
ABRACADABRA
WHERE DID IT GO?
HOW COULD YOU KNOW THAT?
OUCH, THAT'S GOTTA HURT
THE STRONGEST MAN ALIVE
HOW CAN SHE BEND LIKE THAT?
I CAN'T BEAR TO LOOK
NOW THAT'S GROSS
HOLD YOUR NOSE!

Have fun with the new toys

Dave Benham

[foxidrive]

I like your style, and your continued improvements.

[dbenham]

Thanks

But BAAAH I discovered a bug introduced by v6.0 and had to update my prior post to v6.1.

I had wanted /JBEG and /JLIB to always have access to the line count (cnt variable) if the /C option is in force. But the /INC and /EXC options can now implicitly set /C if a negative line number is used. So I moved the code where I loaded /JBEG and /JLIB to after where I load /INC and /EXC. But that completely broke /JBEG and /JLIB. Arrgh.

So I put /JBEG and /JLIB processing back where it belongs in v 6.1, and everything works again. The only downside is the cnt variable will not be available to /JBEG or /JLIB code if you rely on it being set implicitly by /INC or /EXC. Moral of the story - if you need cnt in /JBEG or /JLIB, then always explicitly set /C.


Dave Benham

[dbenham]

I updated my earlier v6 post to v6.2 - I fixed a bug that prevented /J, /JQ, /JMATCH, and /JMATCHQ from working with /P.

===================

More importantly, here is new version. V6.4, which is functionally identical to v6.2, but it just has the potential to run faster, depending on the chosen options.

JREPL6.4.zip

(17.12 KiB) Downloaded 1091 times

After developing v6.2, I ran some performance tests, and discovered that v1.0 was significantly faster than v6.x when minimal options were used. The issue is my main loop had lots of conditional code that would only fire if certain options were chosen. The condition checks took some time.

So I developed v6.3 to optimize the main loop by dynamically building the loop to meet the needs of the chosen options. I managed to get a simple find/repl with no options to perform faster than v1.0
edit - v6.4 fixed a bug introduced by v6.3 where /EXC result was inverted.

As always, let me know if you find bugs. I wouldn't be surprised if there were a few, given that I completely rebuilt the main loop, trying to maintain the logic. If you find v6.3 bugs that prevent you from accomplishing a task, then I recommend falling back to v6.2. If that still does not work, then I recommend v4.5.


Dave Benham

[foxidrive]

Thanks Dave.

[brinda]

thanks for continuous update dave

[foxidrive]

thanks for continuous update dave

What'd be good is a counter for thanks and appreciation that people can click. I feel it would be pretty darn high with many clicks.

I'd like to see a counter for Aacini as well. He's given just as much effort as Dave, in many areas.

There are many people who have contributed a great deal here, and elsewhere, and they are just as worthy of praise.

I single out Dave and Antonio here for their contributions of native batch files that give so many functions that were once very tedious to do in vanilla batch code. I admire them both for their generosity to the batch file community in providing many batch tools, and for improving them, and for remaining part of the community.



So yeah. Thanks guys. And thanks to all the contributors.



===============================================
2021-03-03 Edit from dbenham:

It breaks my heart every time I see this post, knowing that foxidrive passed away just 14 days after posting this. I think he knew he might die soon when he wrote it.

God bless your soul foxi. You will not be forgotten.

[Aacini]

You are welcome, foxi, and cheers!

Antonio