@MechMK1 If you do it anonymously, would them getting mad matter?

@nobody question might not go great unless you roughly know exactly what you fucked up...

Other than that, meeeeh, It seems shitty they're like "you failed, but we're not going to tell you why"

@JourneymanGeek I mean, I know why

I know exactly what I was supposed to do, but I couldn't figure out how to get it to work

ah!

that sounds like a question, especially if you kept track of how it failed

Someone could literally tell me something like "null-character injection" and I'd be like OH MY GOD YES!

o0

I tried that, no worries

@MechMK1 those are the moments of enlightenment

Exactly

Not the EURAKA mouments, the "....Oh..." moments

I won't go into too much detail, but essentially, input was verified if it was an integer, and afterwards used in a raw SQL query

So I had to find some input that both validated as an integer, and which was useful for SQL injection

I tried commenting it out (just to "assume" I managed to evade somehow) and tried something like 1 and 53=ascii(substr(select/**/version(),1,1)) and that worked

so if I had somehow managed to evade the filter, I could have written a script to extract the relevant information from the database and proceed with the further steps, which by the way, I all figured out

Getting info from the DB was the only step I failed at

Hello. Sorry for the question that seems simple. Recently, concerned about security, I used a regular expression [a-zA-Z0-9] in the registration system, in the name field. I realized that this limits a lot, because maybe the user can use a totally different alphabet, as in the case of Asia.

So I thought, "If my fear is SQL injections. I don't need this regular expression that will limit some users. All I need is to use parameterized statements, which are properly designed to protect against SQL statements. Also, filter the input from user is also a plus, against XSS."

Therefore, I would like to know if anyone can tell me if I'm correct about the conclusion I reached?

@Warlock The goal of a user-registration system can be more than to just limit SQL injection. For example, twitter user handles only allow alphanumeric symbols and a handful of other symbols, but no kanji.

While indeed, correctly implemented prepared statements would make SQL injection unlikely, you may also want usernames that are readable to everyone

For example, you would not want someone to register as @​JoeBiden or @Jo​eB​iden or @J​​​​​​​​​​​oeBid​​​​​en or @​​​Joe​​​​​B​​​​iden. I'll leave it up to you to figure out how all of these differ

As a result, it makes sense to only allow a very limited subset of characters, which are recognizable in most of the world. Even in areas where the latin alphabet is not native, it is still commonly used and understood

@MechMK1 Good. Thanks for the tips