doc: Add some better examples for scripted diff #17411

The current example isn't too great, for example it uses find instead of git ls-files. Add a subsection with suggestions and examples.

Feel free to propose some other great examples to add.

Concept ACK.

You have to update the TOC.

Concept ACK -- git ls-files and git grep are both great! :)

BTW: what is the best way to do a re-apply of a conflicted scripted diff when rebasing? I did the following to rebase #17410 and it seems it could be more efficient, and if so, we'd want to add it to the documentation:

$ git rebase master
First, rewinding head to replay your work on top of it...                                      
Applying: Rename `db` log category to `walletdb` (like `coindb`)                                          
Applying: scripted-diff: Change `BCLog::DB` to `BCLog::WALLETDB`                                          Using index info to reconstruct a base tree...                                                            
M       src/wallet/db.cpp                                                                          
Falling back to patching base and 3-way merge...                                        
Auto-merging src/wallet/db.cpp                                                                 
CONFLICT (content): Merge conflict in src/wallet/db.cpp                         
error: Failed to merge in the changes.                                                                    
Patch failed at 0002 scripted-diff: Change `BCLog::DB` to `BCLog::WALLETDB`               
Use 'git am --show-current-patch' to see the failed patch               
                                                                                                          
Resolve all conflicts manually, mark them as resolved with                                 
"git add/rm <conflicted_files>", then run "git rebase --continue".                             
You can instead skip this commit: run "git rebase --skip".                                                
To abort and get back to the state before "git rebase", run "git rebase --abort". 

do hard reset, to get rid of git's attempt at re-applying, and conflict markers

$ git reset --hard

manually paste script steps

$ git grep -l "BCLog::DB" src | xargs sed -i "s/BCLog::DB/BCLog::WA
LLETDB/g"
$ sed -i "s/DB          =/WALLETDB    =/g" src/logging.h
$ git status
 M src/logging.cpp                            
 M src/logging.h                               
 M src/wallet/db.cpp                              
 M src/wallet/walletdb.cpp

mandually git add changed objects, and roll rebase forward

$ git add src/logging.cpp src/logging.h src/wallet/db.cpp src/wallet/walletdb.cpp
$ git rebase --continue

I have a one-line script-get bash function would let you replace the manual paste step:

bash -c "$(script-get HEAD)"

Probably more of this could be scripted in general, though

Thanks!

Indeed, my idea was that this is pretty much an alternative merge strategy: ignore git's own attempt (which could cause conflicts or miss new occurences) in favor of re-running the script. It won't have any merge conflicts, by definition, though obviously the result still has to be reviewed.

ACK d8ae24fae2fce7f39a9eb7fa7c9b21fac79e82ac, I have reviewed the code and it looks OK, I agree it can be merged.

Updated toc, replaced occurrences of "it's" with "it is".

re-ACK adbe15504713ddba6e9c024c59d977675d49e350