Should have written this since day one. Not sure how many have given up using it but here it is.
If you have tried to do search with Breeze and given up due to this not being written please try again

https://github.com/tomvita/Breeze-Beta/wiki/Using-Search-Manager

Now that you know how to use Search Manager, this next one may finally be useful to you
https://github.com/tomvita/Breeze-Beta/wiki/Basics-of-Search

You may want to check this one out too
https://github.com/tomvita/Breeze-Beta/wiki/Setup-Profile-to-launch-Breeze-(optional)

 

[dsrules]

There are two problem. One is the extra buffer size that I allocate, it's only 0x200 which can only support distance of 0x100 bytes. The other is the start of you target is very near the start of segment, I suppose when the distance is bigger than that it won't work.

yeah, it's behaving strangely, on another game, if I used 0xb0 distance it could find the 1 result, but if I use 0x7C , it returns many results whether 0x7C matches ValueB or not, Edizon SE always found the only 1 result using multisearch

anyways, here are my test results for beta79b

1. Toggle Skip still crashes

2. with custom shortcut on, Progress Information window, X End Search and B Pause Search could not be programmable, also push the default shortcuts X and B have no effect

3. Advance Cheat menu > More > all the "Choose any file, Load file" options will crash if you push A in an empty folder

 

[TomSwitch]

yeah, it's behaving strangely, on another game, if I used 0xb0 distance it could find the 1 result, but if I use 0x7C , it returns many results whether 0x7C matches ValueB or not, Edizon SE always found the only 1 result using multisearch

anyways, here are my test results for beta79b

This is the code
"
#define tvr(t) options.tvp_distance *sizeof(t)
#define _TVP(t) if (_value1(t) == _Memory(t)) { s64 save_i = i; for (i = save_i - tvr(t); i <= save_i + tvr(t); i += sizeof(t)) if (_value2(t) == _Memory(t) && i != save_i) {i = save_i; _add_address _add_data(t); break;}; i = save_i;}
"

Anything within this search range that match will give you a candidate:
for (i = save_i - tvr(t); i <= save_i + tvr(t); i += sizeof(t))

As to why 0x7C has less candidate it is probably because some of the candidates is closer than 0x7C from the start of the segment so the read that try to read beyond what is readable failed.

For speed I read a block of memory when there are candidates that are within my allocated buffer size so the 0x100 limit it only applies when the candidate is at the beginning or the end of this group.
Here is the code:
"
rc = dmntchtReadCheatProcessMemory(addr - extra_buffer_pre, buffer, len + extra_buffer_pre + extra_buffer_post); if (R_FAILED(rc)) { \
this->menu->LogText("break1 rc= %x \n", (int)rc); \
return rc; \
};
"
When the read fail the all the candidates that falls within this range is dropped

I will think about whether to improve this algorithm. This starts to become problematic when the distance increases but I think when the distance increase the usefulness of this search reduces as the large amount of candidates becomes too much for human to process

Post automatically merged: Jul 25, 2023

As to why 0x7C has less candidate it is probably because some of the candidates is closer than 0x7C from the start of the segment so the read that try to read beyond what is readable failed.

For speed I read a block of memory when there are candidates that are within my allocated buffer size so the 0x100 limit it only applies when the candidate is at the beginning or the end of this group.
Here is the code:
"
rc = dmntchtReadCheatProcessMemory(addr - extra_buffer_pre, buffer, len + extra_buffer_pre + extra_buffer_post); if (R_FAILED(rc)) { \
this->menu->LogText("break1 rc= %x \n", (int)rc); \
return rc; \
};
"
When the read fail the all the candidates that falls within this range is dropped

I will think about whether to improve this algorithm. This starts to become problematic when the distance increases but I think when the distance increase the usefulness of this search reduces as the large amount of candidates becomes too much for human to process

Actually this was already taken care of ( so I think ) 15 months ago when I wrote this code.
Do you see any "break1 rc=" message?

Post automatically merged: Jul 25, 2023

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79c

toggle skip don't crash for me, if you have no target you will crash when you start the search and there is still no guard against that, is that the reason? target is indicated by a house icon, if you skip the target you have to designate another before you start the search

With the 0x100 byte extended to 0x400 byte your 7C should be within range for all candidates now

 

[dsrules]

This is the code
"
#define tvr(t) options.tvp_distance *sizeof(t)
#define _TVP(t) if (_value1(t) == _Memory(t)) { s64 save_i = i; for (i = save_i - tvr(t); i <= save_i + tvr(t); i += sizeof(t)) if (_value2(t) == _Memory(t) && i != save_i) {i = save_i; _add_address _add_data(t); break;}; i = save_i;}
"

Anything within this search range that match will give you a candidate:
for (i = save_i - tvr(t); i <= save_i + tvr(t); i += sizeof(t))

As to why 0x7C has less candidate it is probably because some of the candidates is closer than 0x7C from the start of the segment so the read that try to read beyond what is readable failed.

For speed I read a block of memory when there are candidates that are within my allocated buffer size so the 0x100 limit it only applies when the candidate is at the beginning or the end of this group.
Here is the code:
"
rc = dmntchtReadCheatProcessMemory(addr - extra_buffer_pre, buffer, len + extra_buffer_pre + extra_buffer_post); if (R_FAILED(rc)) { \
this->menu->LogText("break1 rc= %x \n", (int)rc); \
return rc; \
};
"
When the read fail the all the candidates that falls within this range is dropped

I will think about whether to improve this algorithm. This starts to become problematic when the distance increases but I think when the distance increase the usefulness of this search reduces as the large amount of candidates becomes too much for human to process

Post automatically merged: Jul 25, 2023

Actually this was already taken care of ( so I think ) 15 months ago when I wrote this code.
Do you see any "break1 rc=" message?

Post automatically merged: Jul 25, 2023

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79c

toggle skip don't crash for me, if you have no target you will crash when you start the search and there is still no guard against that, is that the reason? target is indicated by a house icon, if you skip the target you have to designate another before you start the search

With the 0x100 byte extended to 0x400 byte your 7C should be within range for all candidates now

yeah, I just open Breeze and go straight to Toggle Skip for no reason lol

with 79c, push A in an empty folder still crashes for me, Select File disappears when push A then crash screen after a few seconds

and Distance A,,B search is still the same
https://gbatemp.net/attachments/n1-jpg.384830/
like, I am using the value 0x1032AAAA for valueA, and +4 or +8 for ValueB, it could found the 1 result, but if I use +0x10 for ValueB then no result found

 

[TomSwitch]

yeah, I just open Breeze and go straight to Toggle Skip for no reason lol

with 79c, push A in an empty folder still crashes for me, Select File disappears when push A then crash screen after a few seconds

and Distance A,,B search is still the same
https://gbatemp.net/attachments/n1-jpg.384830/
like, I am using the value 0x1032AAAA for valueA, and +4 or +8 for ValueB, it could found the 1 result, but if I use +0x10 for ValueB then no result found

Silly mistake and I was so confident that I didn't even test it. Now it's fixed and I tested it.
release 79c is updated with this fix

 

[dsrules]

View attachment 385059
Silly mistake and I was so confident that I didn't even test it. Now it's fixed and I tested it.
release 79c is updated with this fix

cool! will try it out later

 

[TomSwitch]

79c updated again, this time perhaps you can do long distance

 

[dsrules]

79c updated again, this time perhaps you can do long distance

for u32 ValueA +0xC =ValueB, I need to use distance = 3 to get a good result,
it seems that it searches for all results within a certain range , if distance = 100 , it returns all results from no matter ValueB is at +4,+0x20, -0×18, or -0xc, 0x40, -0x60 instead of aiming at the exact ValueB set distance like Edizon SE
was that how it was intended to do?

 

[TomSwitch]

for u32 ValueA +0xC =ValueB, I need to use distance = 3 to get a good result,
it seems that it searches for all results within a certain range , if distance = 100 , it returns all results from no matter ValueB is at +4,+0x20, -0×18, or -0xc, 0x40, -0x60 instead of aiming at the exact ValueB set distance like Edizon SE
was that how it was intended to do?

yes A,,B means B can be anywhere that is infront or behind A by a distance of 100 * size of the data type you search

if you know the distance where B is at from A you should use advance search with the precise gap

EdiZon SE's multi search is designed to find items you already found before, you need to know the data structure already as well as it's alignment. ( most of the time data are 128 bit aligned so that is the default, you can change it if needed )

Breeze's advance search is a more flexible.

 

[dsrules]

yes A,,B means B can be anywhere that is infront or behind A by a distance of 100 * size of the data type you search

if you know the distance where B is at from A you should use advance search with the precise gap

EdiZon SE's multi search is designed to find items you already found before, you need to know the data structure already as well as it's alignment. ( most of the time data are 128 bit aligned so that is the default, you can change it if needed )

Breeze's advance search is a more flexible.

yeah, I misunderstood how it's suppose to work
Thanks for the explanations, it's all good now

 

[dsrules]

@TomSwitch , just noticed that in betas7x, it doesn't show the full 10 digits address like they did in pre7x versions
not sure if this is a bug or not

 

[TomSwitch]

@TomSwitch , just noticed that in betas7x, it doesn't show the full 10 digits address like they did in pre7x versions
not sure if this is a bug or not
View attachment 387153
View attachment 387152

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79d

 

[dsrules]

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79d

cool, that was quick!

Post automatically merged: Aug 8, 2023

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79d

minor issue
when searched in Hex then Show Candidates, it will still only shows 8 digits instead of 10, need to use Mode toggle to display 10 digits

also, it is possible to keep Hex mode search after exit Breeze? like Start Search in Decimal will still be Decimal after reopen Breeze , float will still be float, but Hex becomes Decimal

 

[dsrules]

@TomSwitch, maybe the previous message went unnoticed because of auto merged
79d

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79d

minor issue
when searched in u32 Hex then Show Candidates, it will still only shows 8 digits instead of 10, need to use Mode toggle to display 10 digits

also, it is possible to keep Hex mode search after exit Breeze? like Start Search in Decimal will still be Decimal after reopen Breeze , float will still be float, but Hex becomes Decimal

 

[TomSwitch]

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79e

 

[dsrules]

https://github.com/tomvita/Breeze-Beta/releases/tag/beta79e

Thanks, now everything seems to be working, nicely

 

[dsrules]

@TomSwitch , could you test and see if you could replicate this or not on beta80?
setup search as u32 == in Search Manager then close Breeze, open Breeze again and goto Gen2 Action menu once then go to Search Manager , u32 == becomes A..B
It happens only when goto Gen2 Action, all other menus no problem

 

[TomSwitch]

@TomSwitch , could you test and see if you could replicate this or not on beta80?
setup search as u32 == in Search Manager then close Breeze, open Breeze again and goto Gen2 Action menu once then go to Search Manager , u32 == becomes A..B
It happens only when goto Gen2 Action, all other menus no problem

I notice A..B too. Thanks for nailing it down to gen2 menu. Will investigate

Post automatically merged: Aug 20, 2023

https://github.com/tomvita/Breeze-Beta/releases/tag/beta80a

 

[Visntse]

hey, i use beta81a with 16.1.0, fresh install, not overwrite old version.
and use the "010000000000000D" folder.(if no this folder, breeze can't be run.)

i install the gen2 and profileshortcut, but when i click profile, it's not show breeze, it's show look like album menu, and must click breeze again.
click profile on 16.0.3 with beta80 it's work fine, i don't know that is 16.1.0's bug?
or maybe you can fix this?
please, thx so much.

 

[TomSwitch]

hey, i use beta81a with 16.1.0, fresh install, not overwrite old version.
and use the "010000000000000D" folder.(if no this folder, breeze can't be run.)

i install the gen2 and profileshortcut, but when i click profile, it's not show breeze, it's show look like album menu, and must click breeze again.
click profile on 16.0.3 with beta80 it's work fine, i don't know that is 16.1.0's bug?
or maybe you can fix this?
please, thx so much.

OK, I guess you gave me one good reason to update to 16.1

Post automatically merged: Aug 23, 2023

It has been reported to be working. Did you go to setting and toggle profile short cut to on?

PS: there is a profilehbm.zip inside Breeze.zip and if you unzip that to the root of SD you will get profile launching hbmenu

 

[Visntse]

yes, that's my use all 81a files.
i install gen2 and turn on shortcut in the breeze, and reboot console.
please u check what problem?