Re: GREP pattern to replace the first element with a tab

[John Delacour]

On 21 Mar 2021, at 08:39, samar  wrote:

> The reason you cannot reproduce the error with your file may be that the 
> second column is limited to three different texts (B1, B2, and B3) whereas in 
> mine there are more (up to B7 here, but the script should also work with more 
> than seven):

The script works with all files that have tab-delimited values, whatever those 
values may be.  Your supposition is quite unfounded; the logic of the script is 
quite clear.

> A1B1
> A1B2
> A1B3
> A1B4
> A1B5
...

but if you separate the values with spaces, as you have done here, then the 
file does not meet your own criteria.  As Christopher has remarked, while 
asking you to send your file, the script works with all valid files, and does 
the work not in eight seconds (!!) but in the twinkling of an eye.

If anybody can point out a real flaw in my script, I shall be happy to correct 
it.  If you will send me your file, I shall be glad to explain why you are 
having the problem.

JD




-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/B4793D64-D656-4A5C-A557-2BACED29814C%40gmail.com.

Re: GREP pattern to replace the first element with a tab

[Kaveh Bazargan]

Try removing the ^ at the start of search. This is because once you have
placed the bullet, the next char is no longer at the start of line. now you
will just be searching from the bullet forwards. So search is now:

([^\t]+)\t([^\t]+\r)(?=\1)

I also put the last \r inside the second bracketed text. Works here.

On Mon, 22 Mar 2021 at 12:50, Tim A  wrote:

> Neat.
> Any way to single step thru Find/Replace to watch the the grep pattern
> progress instead of doing a Replace All?
> If I do a Next and Replace it doesn't work properly.
>
> Thanks
>
> On Sunday, March 21, 2021 at 5:42:16 AM UTC-7 Kaveh wrote:
>
>> Here is a two stage regex solution. Assuming no bullets (•) in your file,
>>
>> Search: ^([^\t]+)\t([^\t]+)\r(?=\1)
>> replace: \1\t\2\r•
>>
>> Search: •[^\t]+
>> Replace with empty
>>
>> On Sun, 21 Mar 2021 at 08:47, Christopher Stone 
>> wrote:
>>
>>> On 03/20/2021, at 15:48, John Delacour  wrote:
>>>
>>> That makes no sense to me.  If the current col1 is identical to col1 of
>>> the previous line, then the value will not be printed; that is the clear
>>> logic of the routine.  I cannot reproduce your error.
>>>
>>> --
>>>
>>> Hey Folks,
>>>
>>> I'm with JD on this one.  I ran his script on a sample file, and it
>>> performed as expected.
>>>
>>> Samar – I'd want a zipped copy of your test file that's failing to
>>> test.  Also - you said you're working with a *large* file.  How large?
>>> It makes a difference in how one approaches the problem.
>>>
>>> I like JD's Perl solution for this.  It's neat, clean, fast, and will
>>> handle big files with ease.
>>>
>>> If I *was* to use AppleScript I'd go this route:
>>>
>>> ---
>>> # Auth: Christopher Stone 
>>> # dCre: 2021/03/21 02:56
>>> # dMod: 2021/03/21 02:56
>>> # Appl: BBEdit
>>> # Task: Massage Text of Columns A and B of a Table.
>>> # Libs: None
>>> # Osax: None
>>> # Tags: @Applescript, @Script, @BBEdit
>>> ---
>>>
>>> *set* colAStr *to* *missing value*
>>> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text
>>> item delimiters, tab}
>>>
>>> *tell* *application* "BBEdit" *to* ¬
>>> *set* paragraphList *to* contents *of* *lines* *of* *front*
>>> *document* *where* *its* contents *is* *not* ""
>>>
>>> *repeat* *with* i *in* paragraphList
>>> *if* colAStr = *text item* 1 *of* i *then*
>>> *set* contents *of* i *to* tab & *text item* 2 *of* i
>>> *else*
>>> *set* colAStr *to* *text item* 1 *of* i
>>> *end* *if*
>>> *end* *repeat*
>>>
>>> *set* AppleScript's text item delimiters *to* linefeed
>>> *set* paragraphList *to* paragraphList *as* *text*
>>> *set* AppleScript's text item delimiters *to* oldTIDS
>>>
>>> *tell* *application* "BBEdit" *to* ¬
>>> *set* *text* *of* *front* *document* *to* paragraphList
>>>
>>> ---
>>>
>>> Although it could bog down with very big files.
>>>
>>>
>>> --
>>> Best Regards,
>>> Chris
>>>
>>> --
>>> This is the BBEdit Talk public discussion group. If you have a feature
>>> request or need technical support, please email "sup...@barebones.com"
>>> rather than posting here. Follow @bbedit on Twitter: <
>>> https://twitter.com/bbedit>
>>> ---
>>> You received this message because you are subscribed to the Google
>>> Groups "BBEdit Talk" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to bbedit+un...@googlegroups.com.
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com
>>> 
>>> .
>>>
>>
>>
>> --
>> Kaveh Bazargan PhD
>> Director
>> River Valley Technologies  ● Twitter
>>  ● LinkedIn
>>  ● ORCID
>> 
>> *Accelerating the Communication of Research*
>>
> --
> This is the BBEdit Talk public discussion group. If you have a feature
> request or need technical support, please email "supp...@barebones.com"
> rather than posting here. Follow @bbedit on Twitter: <
> https://twitter.com/bbedit>
> ---
> You received this message because you are subscribed to the Google Groups
> "BBEdit Talk" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to bbedit+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/bbedit/1aaf8e41-f479-431e-a290-777609746a27n%40googlegroups.com
> 
> .
>


-- 
Kaveh Bazargan PhD
Director
River Valley Technologies 

Re: GREP pattern to replace the first element with a tab

[Tim A]

Neat. 
Any way to single step thru Find/Replace to watch the the grep pattern 
progress instead of doing a Replace All? 
If I do a Next and Replace it doesn't work properly.

Thanks

On Sunday, March 21, 2021 at 5:42:16 AM UTC-7 Kaveh wrote:

> Here is a two stage regex solution. Assuming no bullets (•) in your file, 
>
> Search: ^([^\t]+)\t([^\t]+)\r(?=\1)
> replace: \1\t\2\r•
>
> Search: •[^\t]+
> Replace with empty
>
> On Sun, 21 Mar 2021 at 08:47, Christopher Stone  
> wrote:
>
>> On 03/20/2021, at 15:48, John Delacour  wrote:
>>
>> That makes no sense to me.  If the current col1 is identical to col1 of 
>> the previous line, then the value will not be printed; that is the clear 
>> logic of the routine.  I cannot reproduce your error.
>>
>> --
>>
>> Hey Folks,
>>
>> I'm with JD on this one.  I ran his script on a sample file, and it 
>> performed as expected.
>>
>> Samar – I'd want a zipped copy of your test file that's failing to test.  
>> Also - you said you're working with a *large* file.  How large?  It 
>> makes a difference in how one approaches the problem.
>>
>> I like JD's Perl solution for this.  It's neat, clean, fast, and will 
>> handle big files with ease.
>>
>> If I *was* to use AppleScript I'd go this route:
>>
>> ---
>> # Auth: Christopher Stone 
>> # dCre: 2021/03/21 02:56
>> # dMod: 2021/03/21 02:56 
>> # Appl: BBEdit
>> # Task: Massage Text of Columns A and B of a Table.
>> # Libs: None
>> # Osax: None
>> # Tags: @Applescript, @Script, @BBEdit
>> ---
>>
>> *set* colAStr *to* *missing value*
>> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text 
>> item delimiters, tab}
>>
>> *tell* *application* "BBEdit" *to* ¬
>> *set* paragraphList *to* contents *of* *lines* *of* *front* 
>> *document* *where* *its* contents *is* *not* ""
>>
>> *repeat* *with* i *in* paragraphList
>> *if* colAStr = *text item* 1 *of* i *then*
>> *set* contents *of* i *to* tab & *text item* 2 *of* i
>> *else*
>> *set* colAStr *to* *text item* 1 *of* i
>> *end* *if*
>> *end* *repeat*
>>
>> *set* AppleScript's text item delimiters *to* linefeed
>> *set* paragraphList *to* paragraphList *as* *text*
>> *set* AppleScript's text item delimiters *to* oldTIDS
>>
>> *tell* *application* "BBEdit" *to* ¬
>> *set* *text* *of* *front* *document* *to* paragraphList
>>
>> ---
>>
>> Although it could bog down with very big files.
>>
>>
>> --
>> Best Regards,
>> Chris
>>
>> -- 
>> This is the BBEdit Talk public discussion group. If you have a feature 
>> request or need technical support, please email "sup...@barebones.com" 
>> rather than posting here. Follow @bbedit on Twitter: <
>> https://twitter.com/bbedit>
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "BBEdit Talk" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to bbedit+un...@googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com
>>  
>> 
>> .
>>
>
>
> -- 
> Kaveh Bazargan PhD
> Director
> River Valley Technologies  ● Twitter 
>  ● LinkedIn 
>  ● ORCID 
> 
> *Accelerating the Communication of Research*
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/1aaf8e41-f479-431e-a290-777609746a27n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[Bruce Van Allen]

On 21 Mar 2021, at 19:37, Christopher Stone wrote:

On 03/21/2021, at 03:39, samar > wrote:


The reason you cannot reproduce the error with your file may be that 
the second column is limited to three different texts (B1, B2, and 
B3) whereas in mine there are more (up to B7 here, but the script 
should also work with more than seven):




Hey Samar,

Nyet.  The logic of JD's script is quite clear if you know how to read 
Perl, and that shouldn't happen.


Hmm. When I first read Samar’s post - the one with the enclosed screen 
shots - I thought he wanted it to go from something like this:


A1[tab]B1
[tab]B2 
[tab]B3 
[tab]B4 
[tab]B5
A2[tab]B1
[tab]B2 
[tab]B3
[tab]B4
A3[tab]B1   
[tab]B2 
[tab]B3 
[tab]B4 
[tab]B5 
[tab]B6 

to something like this:

A1[tab]B1
A1[tab]B2   
A1[tab]B3   
A1[tab]B4   
A1[tab]B5
A2[tab]B1
A2[tab]B2   
A2[tab]B3
A2[tab]B4
A3[tab]B1   
A3[tab]B2   
A3[tab]B3   
A3[tab]B4   
A3[tab]B5
A3[tab]B6   

If this is correct, JD’s script could be easily changed to do it.

Note that to get the first $x ($a in JD’s script), you have to parse & 
print the first line before looping over the rest.


And this assumes that the first line has a value in both columns.


#!/usr/bin/perl
my ($x, $y) = split "\t", <>;
print "$x\t$y";
while (<>) {
my ($col1, $col2) = split "\t", $_;
if ($col1 and $col1 ne $x) {
$x = $col1;
}
print "$x\t$col2";
}

But maybe I misunderstood the goal.


HTH,

- Bruce

_bruce__van_allen__santa_cruz__ca

--
This is the BBEdit Talk public discussion group. If you have a feature request or need 
technical support, please email "supp...@barebones.com" rather than posting here. 
Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups "BBEdit Talk" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/5903E9EF-FB0E-4D63-BBD5-35A8B56889A2%40cruzio.com.

Re: GREP pattern to replace the first element with a tab

[Christopher Stone]

On 03/21/2021, at 03:39, samar mailto:arnet...@bluewin.ch>> wrote:
> The reason you cannot reproduce the error with your file may be that the 
> second column is limited to three different texts (B1, B2, and B3) whereas in 
> mine there are more (up to B7 here, but the script should also work with more 
> than seven):


Hey Samar,

Nyet.  The logic of JD's script is quite clear if you know how to read Perl, 
and that shouldn't happen.

My best guess is your test file is different in structure than what you've 
given us as an example, and that's throwing off JD's script – but I can't tell 
without being able to reproduce the problem.

I created a 10,000 line test file with 1000 A-groups, and JD's script works 
perfectly in about 1/4 of a second on my old Mid-2010 17" 2.66 GHz Intel Core 
i7 MacBook Pro.

When asking for text processing it's best to give real-world data if at all 
possible, because the devil is in the details – and small anomalies can cause 
big problems.

I've rewritten JD's script to hopefully reduce any chance of error.  It's a bit 
terse, but hopefully the commenting overcomes that.

Why wait 8 seconds when instantaneous is available?

--
Best Regards,
Chris


#!/usr/bin/env perl -sw
# 
-
# Auth: Christopher Stone mailto:scriptmeis...@thestoneforge.com>>
# dCre: 2012/11/27 08:12
# dMod: 2021/03/21 21:13 
# Task: Create Indented Structure from a 2 Column tab-delimited table.
# Tags: @ccstone, @Shell, @Script, @Indented, @Column, @Table
# 
-
use v5.010;

my $column1StartStr = "";# Initialize variable

while (<>) { # Process Lines one at a time.

   m!(^[^\t]+)(\t.+)!m;  # Matches Column 1 & [\t]Column 2 in the 
current line to $1 & $2.

   if ($1 eq $column1StartStr) { # Equality test – start-state of Col A & Col A 
of current line:

  say "$2";  # If equal; print only Col B with \t delim.

   } else {

  say "$1$2";# If NOT equal; print both columns.
  $column1StartStr = $1; # If NOT equal; reset Col A start string.

   }

}

# 
-


-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/EF00E382-AECA-426B-99FD-A49353D18137%40gmail.com.

Re: GREP pattern to replace the first element with a tab

[Kaveh Bazargan]

Glad it helped Samar. Yes, I am generally afraid of lookaheads and sometime
not sure of why we can't just use a pattern and replace it verbatim, but
this is a good case for it because you have to ensure you do not select the
start of the next line in the grep search.

The thing I find hardest is remembering the syntax for lookaheads  and look
behinds so on Mac I use TypeIt4Me
 with shortcuts.

Regards
Kaveh

On Sun, 21 Mar 2021 at 19:04, samar  wrote:

> Very clever! I had not thought of using positive lookahead, but this
> indeed makes GREP work here. The only thing I had to add was a tab
> character before the closing parenthesis in the first search so that
> similar A1 values (those beginning in the same way) will be excluded:
>
> ^([^\t]+)\t([^\t]+)\r(?=\1\t)
>
> Thanks much!
> samar
> On Sunday, March 21, 2021 at 1:42:16 PM UTC+1 Kaveh wrote:
>
>> Here is a two stage regex solution. Assuming no bullets (•) in your file,
>>
>> Search: ^([^\t]+)\t([^\t]+)\r(?=\1)
>> replace: \1\t\2\r•
>>
>> Search: •[^\t]+
>> Replace with empty
>>
>> On Sun, 21 Mar 2021 at 08:47, Christopher Stone 
>> wrote:
>>
>>> On 03/20/2021, at 15:48, John Delacour  wrote:
>>>
>>> That makes no sense to me.  If the current col1 is identical to col1 of
>>> the previous line, then the value will not be printed; that is the clear
>>> logic of the routine.  I cannot reproduce your error.
>>>
>>> --
>>>
>>> Hey Folks,
>>>
>>> I'm with JD on this one.  I ran his script on a sample file, and it
>>> performed as expected.
>>>
>>> Samar – I'd want a zipped copy of your test file that's failing to
>>> test.  Also - you said you're working with a *large* file.  How large?
>>> It makes a difference in how one approaches the problem.
>>>
>>> I like JD's Perl solution for this.  It's neat, clean, fast, and will
>>> handle big files with ease.
>>>
>>> If I *was* to use AppleScript I'd go this route:
>>>
>>> ---
>>> # Auth: Christopher Stone 
>>> # dCre: 2021/03/21 02:56
>>> # dMod: 2021/03/21 02:56
>>> # Appl: BBEdit
>>> # Task: Massage Text of Columns A and B of a Table.
>>> # Libs: None
>>> # Osax: None
>>> # Tags: @Applescript, @Script, @BBEdit
>>> ---
>>>
>>> *set* colAStr *to* *missing value*
>>> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text
>>> item delimiters, tab}
>>>
>>> *tell* *application* "BBEdit" *to* ¬
>>> *set* paragraphList *to* contents *of* *lines* *of* *front*
>>> *document* *where* *its* contents *is* *not* ""
>>>
>>> *repeat* *with* i *in* paragraphList
>>> *if* colAStr = *text item* 1 *of* i *then*
>>> *set* contents *of* i *to* tab & *text item* 2 *of* i
>>> *else*
>>> *set* colAStr *to* *text item* 1 *of* i
>>> *end* *if*
>>> *end* *repeat*
>>>
>>> *set* AppleScript's text item delimiters *to* linefeed
>>> *set* paragraphList *to* paragraphList *as* *text*
>>> *set* AppleScript's text item delimiters *to* oldTIDS
>>>
>>> *tell* *application* "BBEdit" *to* ¬
>>> *set* *text* *of* *front* *document* *to* paragraphList
>>>
>>> ---
>>>
>>> Although it could bog down with very big files.
>>>
>>>
>>> --
>>> Best Regards,
>>> Chris
>>>
>>> --
>>> This is the BBEdit Talk public discussion group. If you have a feature
>>> request or need technical support, please email "sup...@barebones.com"
>>> rather than posting here. Follow @bbedit on Twitter: <
>>> https://twitter.com/bbedit>
>>> ---
>>> You received this message because you are subscribed to the Google
>>> Groups "BBEdit Talk" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to bbedit+un...@googlegroups.com.
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com
>>> 
>>> .
>>>
>>
>>
>> --
>> Kaveh Bazargan PhD
>> Director
>> River Valley Technologies  ● Twitter
>>  ● LinkedIn
>>  ● ORCID
>> 
>> *Accelerating the Communication of Research*
>>
> --
> This is the BBEdit Talk public discussion group. If you have a feature
> request or need technical support, please email "supp...@barebones.com"
> rather than posting here. Follow @bbedit on Twitter: <
> https://twitter.com/bbedit>
> ---
> You received this message because you are subscribed to the Google Groups
> "BBEdit Talk" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to bbedit+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> 

Re: GREP pattern to replace the first element with a tab

[samar]

Very clever! I had not thought of using positive lookahead, but this indeed 
makes GREP work here. The only thing I had to add was a tab character 
before the closing parenthesis in the first search so that similar A1 
values (those beginning in the same way) will be excluded:

^([^\t]+)\t([^\t]+)\r(?=\1\t)

Thanks much!
samar
On Sunday, March 21, 2021 at 1:42:16 PM UTC+1 Kaveh wrote:

> Here is a two stage regex solution. Assuming no bullets (•) in your file, 
>
> Search: ^([^\t]+)\t([^\t]+)\r(?=\1)
> replace: \1\t\2\r•
>
> Search: •[^\t]+
> Replace with empty
>
> On Sun, 21 Mar 2021 at 08:47, Christopher Stone  
> wrote:
>
>> On 03/20/2021, at 15:48, John Delacour  wrote:
>>
>> That makes no sense to me.  If the current col1 is identical to col1 of 
>> the previous line, then the value will not be printed; that is the clear 
>> logic of the routine.  I cannot reproduce your error.
>>
>> --
>>
>> Hey Folks,
>>
>> I'm with JD on this one.  I ran his script on a sample file, and it 
>> performed as expected.
>>
>> Samar – I'd want a zipped copy of your test file that's failing to test.  
>> Also - you said you're working with a *large* file.  How large?  It 
>> makes a difference in how one approaches the problem.
>>
>> I like JD's Perl solution for this.  It's neat, clean, fast, and will 
>> handle big files with ease.
>>
>> If I *was* to use AppleScript I'd go this route:
>>
>> ---
>> # Auth: Christopher Stone 
>> # dCre: 2021/03/21 02:56
>> # dMod: 2021/03/21 02:56 
>> # Appl: BBEdit
>> # Task: Massage Text of Columns A and B of a Table.
>> # Libs: None
>> # Osax: None
>> # Tags: @Applescript, @Script, @BBEdit
>> ---
>>
>> *set* colAStr *to* *missing value*
>> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text 
>> item delimiters, tab}
>>
>> *tell* *application* "BBEdit" *to* ¬
>> *set* paragraphList *to* contents *of* *lines* *of* *front* 
>> *document* *where* *its* contents *is* *not* ""
>>
>> *repeat* *with* i *in* paragraphList
>> *if* colAStr = *text item* 1 *of* i *then*
>> *set* contents *of* i *to* tab & *text item* 2 *of* i
>> *else*
>> *set* colAStr *to* *text item* 1 *of* i
>> *end* *if*
>> *end* *repeat*
>>
>> *set* AppleScript's text item delimiters *to* linefeed
>> *set* paragraphList *to* paragraphList *as* *text*
>> *set* AppleScript's text item delimiters *to* oldTIDS
>>
>> *tell* *application* "BBEdit" *to* ¬
>> *set* *text* *of* *front* *document* *to* paragraphList
>>
>> ---
>>
>> Although it could bog down with very big files.
>>
>>
>> --
>> Best Regards,
>> Chris
>>
>> -- 
>> This is the BBEdit Talk public discussion group. If you have a feature 
>> request or need technical support, please email "sup...@barebones.com" 
>> rather than posting here. Follow @bbedit on Twitter: <
>> https://twitter.com/bbedit>
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "BBEdit Talk" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to bbedit+un...@googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com
>>  
>> 
>> .
>>
>
>
> -- 
> Kaveh Bazargan PhD
> Director
> River Valley Technologies  ● Twitter 
>  ● LinkedIn 
>  ● ORCID 
> 
> *Accelerating the Communication of Research*
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/9531484b-8ac0-4e99-b63a-0662d5b5c654n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[samar]

Wow, thank you, Christian, that works exactly as expected! (I only needed 
to change the closing quotation mark for the script to run.)

This is very helpful – my 8,000-line file was magically modified within 8 
seconds.

samar
On Saturday, March 20, 2021 at 8:13:30 PM UTC+1 ChristianBoyce wrote:

> You asked for GREP, and this isn’t GREP, but it still solves your problem. 
> I used AppleScript. 
>
> Note: this script is written in order to be as easy to follow as possible. 
> It could be improved of course. 
>
> Basically I’m going down your document a line at a time, comparing what’s 
> on the left side of the tab to the value that was in Column One the last 
> time it changed. If it’s the same as that, I don’t write it— I just write 
> tab and Column Two to a variable called newText. If it’s different, I write 
> Column One and Column Two to newText. Finally I set the text of the 
> document to the contents of the variable newText.
>
> Watch out fo the line that sets AppleScript’s text item delimiters to a 
> tab. The line actually looks like this before compiling:
>
> *set* AppleScript's text item delimiters *to* {"\t”}
>
> But when you compile the slash-t is changed to a tab, which is invisible. 
>
> --
>
> *use* AppleScript version "2.4" -- Yosemite (10.10) or later
> *use* *scripting additions*
>
> *tell* *application* "BBEdit"
> *set* myDoc *to* *document* 1
> *set* originalText *to* *text* *of* myDoc
> *set* newText *to* ""
> *set* ColumnOne *to* ""
> *set* ColumnTwo *to* ""
> --
> *set* theParagraphs *to* *paragraphs* *of* originalText *as* *list*
> *repeat* *with* aParagraph *in* theParagraphs
> *set* saveTID *to* AppleScript's text item delimiters
> *set* AppleScript's text item delimiters *to* {" “}— that’s a tab in 
> there— you can’t tell but it is. Use “\t” as I’ve written 
> — on the next line
> -- *set* AppleScript's text item delimiters *to* {"\t”}— commented out so 
> you can see how it looks before compiling
>
> *set* thisColumnOne *to* *text item* 1 *of* aParagraph
> *set* thisColumnTwo *to* *text item* 2 *of* aParagraph
> *if* thisColumnOne *is* ColumnOne *then*
> *set* thisOutput *to* tab & thisColumnTwo & return
> *else*
> *set* thisOutput *to* thisColumnOne & tab & thisColumnTwo & return
> *set* ColumnOne *to* thisColumnOne
> *end* *if*
> *set* newText *to* newText & thisOutput
> *end* *repeat*
> *set* AppleScript's text item delimiters *to* saveTID
> *set* *text* *of* myDoc *to* newText
> *end* *tell*
>
> On Mar 20, 2021, at 4:30 AM, samar  wrote:
>
> Hi
>
> This is the first time a post a message here.
>
> I'm looking for something which I assume is pretty easy to accomplish with 
> GREP but I fail to see how.
>
> I have a large text file with entries sorted in this way:
>
> 
>
> That is, each line has two elements (represented here by A\d and B\d) 
> separated by one tab character. The length of the two elements is between 1 
> and 100 characters each.
>
> Now I need a GREP Find/Replace string to make it look like this:
>
> 
>
> That is, the first element is replaced by a tab character *if* the first 
> element is a repetition of the first element of the previous line(s). The 
> sort order remains unchanged.
>
> Sounds simple. And yet all I manage to GREP for is this:
>
> 
> The result looks promising but in now way satisfactory:
>
> 
>
> The tab characters are inserted correctly, but the problem has to do with 
> repetition ... Does anyone see how this can be resolved?
>
> Thanks
> samar
>
>
>
> —
>
> Socially distantly yours,
>
> Christian Boyce
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/2846c38d-5e24-4501-9cdf-5126cfddc9b7n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[Kaveh Bazargan]

Here is a two stage regex solution. Assuming no bullets (•) in your file,

Search: ^([^\t]+)\t([^\t]+)\r(?=\1)
replace: \1\t\2\r•

Search: •[^\t]+
Replace with empty

On Sun, 21 Mar 2021 at 08:47, Christopher Stone 
wrote:

> On 03/20/2021, at 15:48, John Delacour  wrote:
>
> That makes no sense to me.  If the current col1 is identical to col1 of
> the previous line, then the value will not be printed; that is the clear
> logic of the routine.  I cannot reproduce your error.
>
> --
>
> Hey Folks,
>
> I'm with JD on this one.  I ran his script on a sample file, and it
> performed as expected.
>
> Samar – I'd want a zipped copy of your test file that's failing to test.
> Also - you said you're working with a *large* file.  How large?  It makes
> a difference in how one approaches the problem.
>
> I like JD's Perl solution for this.  It's neat, clean, fast, and will
> handle big files with ease.
>
> If I *was* to use AppleScript I'd go this route:
>
> ---
> # Auth: Christopher Stone 
> # dCre: 2021/03/21 02:56
> # dMod: 2021/03/21 02:56
> # Appl: BBEdit
> # Task: Massage Text of Columns A and B of a Table.
> # Libs: None
> # Osax: None
> # Tags: @Applescript, @Script, @BBEdit
> ---
>
> *set* colAStr *to* *missing value*
> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text
> item delimiters, tab}
>
> *tell* *application* "BBEdit" *to* ¬
> *set* paragraphList *to* contents *of* *lines* *of* *front* *document*
> *where* *its* contents *is* *not* ""
>
> *repeat* *with* i *in* paragraphList
> *if* colAStr = *text item* 1 *of* i *then*
> *set* contents *of* i *to* tab & *text item* 2 *of* i
> *else*
> *set* colAStr *to* *text item* 1 *of* i
> *end* *if*
> *end* *repeat*
>
> *set* AppleScript's text item delimiters *to* linefeed
> *set* paragraphList *to* paragraphList *as* *text*
> *set* AppleScript's text item delimiters *to* oldTIDS
>
> *tell* *application* "BBEdit" *to* ¬
> *set* *text* *of* *front* *document* *to* paragraphList
>
> ---
>
> Although it could bog down with very big files.
>
>
> --
> Best Regards,
> Chris
>
> --
> This is the BBEdit Talk public discussion group. If you have a feature
> request or need technical support, please email "supp...@barebones.com"
> rather than posting here. Follow @bbedit on Twitter: <
> https://twitter.com/bbedit>
> ---
> You received this message because you are subscribed to the Google Groups
> "BBEdit Talk" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to bbedit+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com
> 
> .
>


-- 
Kaveh Bazargan PhD
Director
River Valley Technologies  ● Twitter
 ● LinkedIn
 ● ORCID

*Accelerating the Communication of Research*

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/CAJ2R9phrfyDckhU2d_Vp6p4XVb4npEcB7Z0DvOUtzMDzBac9Mg%40mail.gmail.com.

Re: GREP pattern to replace the first element with a tab

[samar]

Hi Chris

(For some reason my messages here occur only several hours after I've sent 
them ...)

Thank you very much indeed – your AppleScript works perfectly well here (8 
seconds when applied to my file of 8,000 lines), the result is the same as 
when I run Christian's script. So I can even choose between the two!

samar
On Sunday, March 21, 2021 at 9:47:34 AM UTC+1 listmei...@gmail.com wrote:

> On 03/20/2021, at 15:48, John Delacour  wrote:
>
> That makes no sense to me.  If the current col1 is identical to col1 of 
> the previous line, then the value will not be printed; that is the clear 
> logic of the routine.  I cannot reproduce your error.
>
> --
>
> Hey Folks,
>
> I'm with JD on this one.  I ran his script on a sample file, and it 
> performed as expected.
>
> Samar – I'd want a zipped copy of your test file that's failing to test. 
>  Also - you said you're working with a *large* file.  How large?  It 
> makes a difference in how one approaches the problem.
>
> I like JD's Perl solution for this.  It's neat, clean, fast, and will 
> handle big files with ease.
>
> If I *was* to use AppleScript I'd go this route:
>
> ---
> # Auth: Christopher Stone 
> # dCre: 2021/03/21 02:56
> # dMod: 2021/03/21 02:56 
> # Appl: BBEdit
> # Task: Massage Text of Columns A and B of a Table.
> # Libs: None
> # Osax: None
> # Tags: @Applescript, @Script, @BBEdit
> ---
>
> *set* colAStr *to* *missing value*
> *set* {oldTIDS, AppleScript's text item delimiters} *to* {AppleScript's text 
> item delimiters, tab}
>
> *tell* *application* "BBEdit" *to* ¬
> *set* paragraphList *to* contents *of* *lines* *of* *front* *document* 
> *where* *its* contents *is* *not* ""
>
> *repeat* *with* i *in* paragraphList
> *if* colAStr = *text item* 1 *of* i *then*
> *set* contents *of* i *to* tab & *text item* 2 *of* i
> *else*
> *set* colAStr *to* *text item* 1 *of* i
> *end* *if*
> *end* *repeat*
>
> *set* AppleScript's text item delimiters *to* linefeed
> *set* paragraphList *to* paragraphList *as* *text*
> *set* AppleScript's text item delimiters *to* oldTIDS
>
> *tell* *application* "BBEdit" *to* ¬
> *set* *text* *of* *front* *document* *to* paragraphList
>
> ---
>
> Although it could bog down with very big files.
>
>
> --
> Best Regards,
> Chris
>
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/f552240a-d802-4227-9311-8d441c01ac61n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[samar]

The reason you cannot reproduce the error with your file may be that the 
second column is limited to three different texts (B1, B2, and B3) whereas 
in mine there are more (up to B7 here, but the script should also work with 
more than seven):

A1B1
A1B2
A1B3
A1B4
A1B5
A1B6
A1B7
A2B1
A2B2
A2B3
A2B4
A2B5
A2B6
A3B1
A3B2
A3B3
A3B4
A3B5
A3B6
A3B7

On Saturday, March 20, 2021 at 9:48:26 PM UTC+1 John R M. Delacour wrote:

>
>
> On 20 Mar 2021, at 16:33, samar  wrote:
>
> ...when I run the text filter, not all occurrences of the first element 
> get replaced when necessary:
>
>
> 
>
> Here a tab should also replace A1 in line 5, A2 in line 12, and A3 in line 
> 18.
>
>
> That makes no sense to me.  If the current col1 is identical to col1 of 
> the previous line, then the value will not be printed; that is the clear 
> logic of the routine.  I cannot reproduce your error.
>
>
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/007bd5f4-3616-41d5-8cfe-98bbf550efe8n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[Christopher Stone]

On 03/20/2021, at 15:48, John Delacour mailto:johndelac...@gmail.com>> wrote:
> That makes no sense to me.  If the current col1 is identical to col1 of the 
> previous line, then the value will not be printed; that is the clear logic of 
> the routine.  I cannot reproduce your error.


Hey Folks,

I'm with JD on this one.  I ran his script on a sample file, and it performed 
as expected.

Samar – I'd want a zipped copy of your test file that's failing to test.  Also 
- you said you're working with a large file.  How large?  It makes a difference 
in how one approaches the problem.

I like JD's Perl solution for this.  It's neat, clean, fast, and will handle 
big files with ease.

If I was to use AppleScript I'd go this route:

---
# Auth: Christopher Stone mailto:scriptmeis...@thestoneforge.com>>
# dCre: 2021/03/21 02:56
# dMod: 2021/03/21 02:56 
# Appl: BBEdit
# Task: Massage Text of Columns A and B of a Table.
# Libs: None
# Osax: None
# Tags: @Applescript, @Script, @BBEdit
---

set colAStr to missing value
set {oldTIDS, AppleScript's text item delimiters} to {AppleScript's text item 
delimiters, tab}

tell application "BBEdit" to ¬
set paragraphList to contents of lines of front document where its contents 
is not ""

repeat with i in paragraphList
if colAStr = text item 1 of i then
set contents of i to tab & text item 2 of i
else
set colAStr to text item 1 of i
end if
end repeat

set AppleScript's text item delimiters to linefeed
set paragraphList to paragraphList as text
set AppleScript's text item delimiters to oldTIDS

tell application "BBEdit" to ¬
set text of front document to paragraphList

---

Although it could bog down with very big files.


--
Best Regards,
Chris

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/B75B0684-484C-4034-AF69-D0731A9FE590%40gmail.com.

Re: GREP pattern to replace the first element with a tab

[John Delacour]

On 20 Mar 2021, at 16:33, samar mailto:arnet...@bluewin.ch>> wrote:

> ...when I run the text filter, not all occurrences of the first element get 
> replaced when necessary:
> 
> 
> 
> Here a tab should also replace A1 in line 5, A2 in line 12, and A3 in line 18.

That makes no sense to me.  If the current col1 is identical to col1 of the 
previous line, then the value will not be printed; that is the clear logic of 
the routine.  I cannot reproduce your error.


-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/65002B2A-D433-44C3-A303-1E84D8A91EE9%40gmail.com.

Re: GREP pattern to replace the first element with a tab

[Christian Boyce]

You asked for GREP, and this isn’t GREP, but it still solves your problem. I 
used AppleScript. 

Note: this script is written in order to be as easy to follow as possible. It 
could be improved of course. 

Basically I’m going down your document a line at a time, comparing what’s on 
the left side of the tab to the value that was in Column One the last time it 
changed. If it’s the same as that, I don’t write it— I just write tab and 
Column Two to a variable called newText. If it’s different, I write Column One 
and Column Two to newText. Finally I set the text of the document to the 
contents of the variable newText.

Watch out fo the line that sets AppleScript’s text item delimiters to a tab. 
The line actually looks like this before compiling:

set AppleScript's text item delimiters to {"\t”}

But when you compile the slash-t is changed to a tab, which is invisible. 

--

use AppleScript version "2.4" -- Yosemite (10.10) or later
use scripting additions

tell application "BBEdit"
set myDoc to document 1
set originalText to text of myDoc
set newText to ""
set ColumnOne to ""
set ColumnTwo to ""
--
set theParagraphs to paragraphs of originalText as list
repeat with aParagraph in theParagraphs
set saveTID to AppleScript's text item delimiters
set AppleScript's text item delimiters to {"“}— that’s a 
tab in there— you can’t tell but it is. Use “\t” as I’ve written 
— on the next line
--  set AppleScript's text item delimiters to {"\t”}— commented out so you 
can see how it looks before compiling

set thisColumnOne to text item 1 of aParagraph
set thisColumnTwo to text item 2 of aParagraph
if thisColumnOne is ColumnOne then
set thisOutput to tab & thisColumnTwo & return
else
set thisOutput to thisColumnOne & tab & thisColumnTwo & 
return
set ColumnOne to thisColumnOne
end if
set newText to newText & thisOutput
end repeat
set AppleScript's text item delimiters to saveTID
set text of myDoc to newText
end tell

> On Mar 20, 2021, at 4:30 AM, samar  wrote:
> 
> Hi
> 
> This is the first time a post a message here.
> 
> I'm looking for something which I assume is pretty easy to accomplish with 
> GREP but I fail to see how.
> 
> I have a large text file with entries sorted in this way:
> 
> 
> 
> That is, each line has two elements (represented here by A\d and B\d) 
> separated by one tab character. The length of the two elements is between 1 
> and 100 characters each.
> 
> Now I need a GREP Find/Replace string to make it look like this:
> 
> 
> 
> That is, the first element is replaced by a tab character *if* the first 
> element is a repetition of the first element of the previous line(s). The 
> sort order remains unchanged.
> 
> Sounds simple. And yet all I manage to GREP for is this:
> 
> 
> The result looks promising but in now way satisfactory:
> 
> 
> 
> The tab characters are inserted correctly, but the problem has to do with 
> repetition ... Does anyone see how this can be resolved?
> 
> Thanks
> samar
> 
> 

—

Socially distantly yours,

Christian Boyce

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/2441499A-F563-4AF4-B63C-8B5B9B5B47DC%40christianboyce.com.

Re: GREP pattern to replace the first element with a tab

[samar]

Thank you! This is helpful, and the result better than mine was.

However, when I run the text filter, not all occurrences of the first 
element get replaced when necessary:

[image: tab3.png]

Here a tab should also replace A1 in line 5, A2 in line 12, and A3 in line 
18.

Thanks
samar
On Saturday, March 20, 2021 at 4:41:36 PM UTC+1 John R M. Delacour wrote:

>
> On 20 Mar 2021, at 11:30, samar  wrote:
>
> I'm looking for something which I assume is pretty easy to accomplish with 
> GREP but I fail to see how.
>
> I have a large text file with entries sorted in this way:
>
> ...That is, the first element is replaced by a tab character *if* the 
> first element is a repetition of the first element of the previous line(s). 
> The sort order remains unchanged.
>
>
> Create a text filter "~/Library/Application Support/BBEdit/Text 
> Filters/SAMAR.pl" and run it from the Text Filters palette (opened from 
> Window menu).  The script will affect the whole file, or just the selection 
> if one is made:
>
>
> #! /usr/bin/perl
> my $a;
> while (<>) {
> my ($col1, $col2) = split "\t", $_;
> if ($col1 eq $a) {
> print "\t$col2"
> } else {
> print "$col1\t$col2";
> $a = $col1;
> }
> }
>
>

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/71112be2-d039-4014-9daf-38608374fb49n%40googlegroups.com.

Re: GREP pattern to replace the first element with a tab

[John Delacour]

On 20 Mar 2021, at 11:30, samar  wrote:

> I'm looking for something which I assume is pretty easy to accomplish with 
> GREP but I fail to see how.
> 
> I have a large text file with entries sorted in this way:
> 
> ...That is, the first element is replaced by a tab character *if* the first 
> element is a repetition of the first element of the previous line(s). The 
> sort order remains unchanged.

Create a text filter "~/Library/Application Support/BBEdit/Text 
Filters/SAMAR.pl" and run it from the Text Filters palette (opened from Window 
menu).  The script will affect the whole file, or just the selection if one is 
made:


#! /usr/bin/perl
my $a;
while (<>) {
my ($col1, $col2) = split "\t", $_;
if ($col1 eq $a) {
print "\t$col2"
} else {
print "$col1\t$col2";
$a = $col1;
}
}

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/4EE3BAFE-E482-4CDD-A428-EAF083608F4A%40gmail.com.

GREP pattern to replace the first element with a tab

[samar]

Hi

This is the first time a post a message here.

I'm looking for something which I assume is pretty easy to accomplish with 
GREP but I fail to see how.

I have a large text file with entries sorted in this way:

[image: tab1.png]

That is, each line has two elements (represented here by A\d and B\d) 
separated by one tab character. The length of the two elements is between 1 
and 100 characters each.

Now I need a GREP Find/Replace string to make it look like this:

[image: tab2.png]

That is, the first element is replaced by a tab character *if* the first 
element is a repetition of the first element of the previous line(s). The 
sort order remains unchanged.

Sounds simple. And yet all I manage to GREP for is this:

[image: Screenshot 2021-03-20 at 12.26.11.png]
The result looks promising but in now way satisfactory:

[image: Screenshot 2021-03-20 at 12.22.46.png]

The tab characters are inserted correctly, but the problem has to do with 
repetition ... Does anyone see how this can be resolved?

Thanks
samar

-- 
This is the BBEdit Talk public discussion group. If you have a feature request 
or need technical support, please email "supp...@barebones.com" rather than 
posting here. Follow @bbedit on Twitter: 
--- 
You received this message because you are subscribed to the Google Groups 
"BBEdit Talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to bbedit+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/bbedit/408db98a-d7c7-44b9-9689-ed8f9fe8b4d0n%40googlegroups.com.