Hi everyone, what I have here is new password library that supports multiple passwords and multiple users. I didn't really like the current password library mainly because it can only check one password and it doesn't support any usernames. So I made a new one and after extensive testing, I think it should be shared.
Please note, I am still working on it to make it more code efficient, so I welcome any suggestions you may have. This is more of a rough draft version, because I am in the process of linking it with both LiquidCrystal (/ I2C) libraries and the current UTFT library.
saw several things, one that catched my eye:
1)
passTemp[((Number_of_Passwords != 0) ? Number_of_Passwords : 0) + NOP_Inc]
equals
passTemp[Number_of_Passwords + NOP_Inc]
// SetPassword and SetUser are technically not needed functions, because AddPassword takes care of both passwords and users
why add them if they are not needed?
do not understand the detailed working esp addressing of the array elements (didn't dive into it) but I assume strcpy() or strncpy() would eleminate a lot of loops. a bit like
int n = (i % (( Num_of_Users + User_Inc) * Max_Lenght_of_username))/Max_Users;
strncpy(user, usertemp[n], j );
Funny you should say that, because after I posted this, I began to rewrite the library. I got rid of set user and set password and I'm trying to see if I can cut the code down by storing the passwords and usernames in a structure. It works great until I try to add a new password to the list. That's when I'm guessing the memory is leaking.
After much debugging I noticed that if I add four password in the sketch, it works fine. But once I manually enter a password, the password I entered always matches what was stored.
I did a print out of what I entered and got this. These are the four passwords I stored at compile time.
1234 (what I enter) 1234(what is stored)
5678. 5678. = user found
abcd. abcd = user found
Asdf. Asdf =user found
Now I add a new password 8520, It gets stored but now if I enter an actual wrong password it over writes 8520 and says that the password was found every time.
I really don't know how to solve this problem other than to just not have the user be able to store new passwords.
Any thoughts as to why this is happening? There is always a snag.
I figured it out... kinda, I went back to my original version and just modified it but kept the char * arrays instead of using a structure. I still want to understand what the problem was, I'm still thinking it was memory leakage issues.
UPDATE: (Actually finished it almost two weeks ago)
My password library is technically done, but now I'm deciding whether or not I should allow the user to choose how the data is stored. Right now both the passwords and usernames are stored in the heap, rather than the stack. However, I am thinking maybe I should allow the user to have a choice of where they store their data. The heap seems to take up more compile memory, but it also seems more efficient, and less likely to cause memory problems. Whereas the if the data was stored in the stack via 2D arrays, then it would take up less memory, but may cause problems later.
So if anyone thinks its a good idea to implement, let me know, if not then I'll keep the library as is.
Oh, I decided to not implement the LCD or Keypad libraries in my library, mainly because not everyone uses the standard LCD and Keypad libraries. But it will still work with them regardless of the library(s) used.
I've been going through all my past libraries and I decided to add some more things to this library. I added a few new features, that I think many will enjoy.
Just have one password that is 4 characters long and another 3 long and they both will work.
You can even have 1234 and 123, and they will be seen as two different passwords.
Keep in mind the default max password length is 6 characters and the max number of passwords you can have is 10.
If you want longer passwords or more than 10, then you need to go into the .h file and change these numbers accordingly.
For real world applications, I would agree 100% with you, but for something using an arduino to where you and only you can change the code, why would you need to?
When I change Joe's password from abcd to abc in the NewPassword_SetEntry patch, it never accepts the 3 digit password initially. It does accept it the second time and then it inconsistently works while sending the same abc 3 digit password then on. I've been adjusting char Pword[10] = {'\0'}; to no avail.
The SetEntry sketch does accept the 4 digit passwords consistently everytime. If you go between sending 3 and 4 digit passwords back and forth it's extremely inconsistent.
Is there any work around for this? I'm using your code to have my Arduino execute simple pin high and low commands reacting from an already existing stream of 3 or 4 digit data that can not not add a #. It can not be altered at all. This really sucks for me because ManualEntry sketch works perfectly but as you know requires the # at the end.
I hadn't realized I didn't put the newer version of the library when I made it. The one you have is the one that uses the Heap instead of the stack to store the passwords.
It works yes, but it is a very poor way of doing it and it is prone to have errors.
See the new version of setEntry. I added a comment in the sketch, please read it.