Fastest String Search Algorithm.

submited by

Style Pass

2021-06-10 12:30:05

Previous message (by thread): Fastest String Search Algorithm. Next message (by thread): Fastest String Search Algorithm. Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] On Thu, Jun 10, 2021, 8:54 AM Dan Raymond <draymond@foxvalley.net> wrote: > Amit, here is a collection of quotes from you in this thread: > > "I have invented the fastest string search algorithm. It is called > "Choudhary Algorithm"" > > "I hope that people are not of the view that Indians can't invent or > innovate" > > "for 60 KB text and 3.2 KB pattern, strstr() is faster than my > algorithm" > > "for 60 KB text and 4.4 KB pattern, my algorithm is faster than > strstr()" > > "The current tests are skewed in favor of strstr()..." > > "but I don't have time to fix strstr() benchmarks" > > "people who maintain open source software will try defending what > they do even when it is obvious that what they are doing is not correct" > > The reason you are not getting traction here is not because the > community is racist, defensive, and unwilling to accept their mistakes. > It is because you have failed to substantiate your claim that your > algorithm is the "fastest string search algorithm". By your own > admission the existing implementation is faster when using existing > benchmarks. You claim that the benchmarks are flawed but when you were > asked to submit improvements you refused. You claim that "real world" > scenarios favor your algorithm but then you state that your algorithm > requires a 4.4 KB search string to demonstrate an advantage. How is 4.4 > KB a "real world" scenario? I'm sure there are applications that will > search for a string that long but I've never encountered one. > First of all, I can give more proofs by testing from my end with different haystacks / needles but I am not going to fix your benchmark issues. I am only submitting my algorithm but you are putting extra work on me by asking me to fix your benchmark issues. This is not my issue, this is the issue of the person who designed and implemented the benchmark tests. If you do like this, that is, if someone submits some work and then you ask him to fix existing issues also then no one will be interested in submitting their work because they may not want to fix your existing issues in benchmark tests, etc. Also, when I thought more, I realized that no one uses strstr() in real world. C is now-a-days used only in embedded software / kernel, etc. and I have programmed a lot in C but I don't ever remember using strstr(). You people think that you are the smartest people on this earth so let me tell you something about myself. 1. I am programming since 1998. 2. I have programmed in C, C++, JAVA, PHP, PYTHON, HTML, PERL, JAVASCRIPT, ETC. 3. I had invented a new language and written its interpreter in 2002 when I was an intern in a company. 4. I had written and published "How to implement system call". It can be found here: https://tldp.org/HOWTO/html_single/Implement-Sys-Call-Linux-2.6-i386/ 5. I have contributed many patches to linux kernel but left contributing more when linux kernel developers became defensive. 6. I have worked in Cisco Systems and Juniper Networks in USA. 7. Currently, I am working for a networking company and writing code for 4G / 5G networks. 8. I have done B. Tech. in Computer Science from IIT - Varanasi after getting a rank in IIT-JEE. IIT-JEE is one of the toughest exams in this world. 9. I have done Masters in Computer Networking from North Carolina State University, NC, USA. Regards, Amit