New program to fully factor with GMP-ECM

[rogue]

Originally Posted by Andi_HB

Any news to this good work?
Or is this Project dead (R.I.P.) ?

What is your question referring to?

[Andi_HB]

I only want to ask for newer Versions.
For example the Version from last year do not accept numbers like this 12296132992537^223-1
I have to tell the program the complete Number (2920 digits) - maybe that`s a Problem from gmp-ecm.
For my Windows it looks like the output option do not work correctly (allways say "No factors found").

but I don't have any efficient way to stop jobs 1 through 3 when job 4 finds a factor, so I often leave the machine running overnight and find in the morning that I've found a factor twice.

- Do you have news to solve this Problem (implement the code from tmorrow?)

Regards Andi_HB

[rogue]

Originally Posted by Andi_HB

I only want to ask for newer Versions.
For example the Version from last year do not accept numbers like this 12296132992537^223-1
I have to tell the program the complete Number (2920 digits) - maybe that`s a Problem from gmp-ecm.
For my Windows it looks like the output option do not work correctly (allways say "No factors found").

- Do you have news to solve this Problem (implement the code from tmorrow?)

Regards Andi_HB

I see. I doubt that it would be difficult to do, but the program would probably need to be linked with the GMP-ECM expression parser because I wouldn't want to have to replicate the work that an expression parser does.

[Yamato]

Originally Posted by rogue

I see. I doubt that it would be difficult to do, but the program would probably need to be linked with the GMP-ECM expression parser because I wouldn't want to have to replicate the work that an expression parser does.

I have added a stand-alone parser (see parser.c) which accepts every valid combination of the following characters: '0', ..., '9', ' ', '(', ')', '!', '^', '/', '*', '-', '+'

Now it should be possible to type something like

factor "((9*2^5! - 5)*(3^40 + 2) - 8)/3^4"

(don't forget the quotation marks)
I have also created a Win32 binary (after some fiddling): factor_win32.zip

[jasonp]

msieve also contains an expression parser (common/expr_eval.c). Of course, msieve also reproduces all the functionality of Mark's program too :)

[rogue]

Originally Posted by jasonp

msieve also contains an expression parser (common/expr_eval.c). Of course, msieve also reproduces all the functionality of Mark's program too :)

I would recommend using Jason's program as it does a bunch of other checks that mine does not do and of course can use MPQS to deal with larger factorizations where ECM is not as efficient.

[Andi_HB]

Is it possible to implement the B2 and sigma in the output if a factor was found? I use the Windows Version in Dos command prompt on Core 2 Duo.

Regards Andi_HB

[rogue]

Originally Posted by Andi_HB

Is it possible to implement the B2 and sigma in the output if a factor was found? I use the Windows Version in Dos command prompt on Core 2 Duo.

Regards Andi_HB

I have not actively worked on this in a long time. I wasn't aware that anyone even used my code. What is the need for that information?