It's work, sigh, but you can write a routine to read the path in and parse into the individual directories then search. Not sure you could call the shell's dir function to take advantage *.XYZ though. Writing the whole thing would suck but somebody did so for win32. Somebody must have done the *.XYZ or *.* bit for FreeDos???
Go for it. Wirte the bits for FreeDos FreeDos would be one function better than the last MSDOS On 6/14/14, Tom Ehlert <te@...> wrote: > >> is there an int21h function or something provided by freedos for >> doing a filepath and wildcard filespec match? >> I know win32 has such a function. the win32 function is at >> http://msdn.microsoft.com/en-us/library/windows/desktop/bb773727%28v=vs.85%29.aspx > there is no such function in DOS > >> and the filesystem algorithm spec is at >> http://msdn.microsoft.com/en-us/library/ff469270.aspx I already have, but it's a part of my commercial library I wanted to make a living on, and I hesitate to produce it. I did not want to go the donations route to make a living, with audits, audit trails, and the whole thing. If I give it to you, and you GPL it, then I have nothing to use in my commercial programs and I am out my code (and I think I do have a use for that function). mine is simply a wildcard string match function. it's about 130 lines of C++, so it's not real huge. there may be appropriate functionality provided in C++14 and Boost with the filesystem class, which handles filenames, paths, mungeing of same. AND it's OS-portable. however, that filesystem algorithm spec does have some strangeties in it, like some stuff that sounds like it was made for commandline argument processors in command.com for instance, because some of it seems unrelated to doing a filespec... makes use of .find_first_of .substr() .size() .compare() .icompare() (I had to write this) and that's it for string stuff, all of which can be done in C, and probably also implementable in assembler, although find_first_of is used simply to iterate through a list of characters to search for and find whichever it finds first, and these are the different wildcard chars, so this is implementable with a nested for loop in assy. there is no string finding, so you are off the hook there. substr is usually very quickly implementable with C strings in this instance by simply adding to the string pointer (adding to an address offset). but there are some .find() instances where a length is required, so you would have to do an ECX loop string copy where ECX contains one of certain calculated values. NUMA I think refers to Non-Uniform Memory Access you can set processor affinity, etc. http://msdn.microsoft.com/en-us/library/windows/desktop/aa363804%28v=vs.85%29.aspx regarding USB mice and keyboards: sometimes I have multiple sets plugged in for various reasons. so simply polling a keyboard for a specific mouse port could work, or you could on init scan the usb root hubs for a list of HIDs, for th emice, make a list, and poll them on timer interrupt, and if ANY one of them makes a change, report it. same goes with the keyboard(s). you obviously can't connect too many devices simultaneously because at some point, older, slower boxes might have a problem with the interrupt handler running too long (?), not sure if there is a way around this. also, there needs to be a way to have a command to tell the driver to re-init (re-scan the root hub) in case something goes wrong or for whatever reason a user must change his USB devices around (like one of them goes out, rearrangement, whatever). I have had to a numbewr of times. Are the FreedDOS folk making a 64-bit multitasking DOS with EFI+GPT support, or what is going to be attempted for the next version? I thought I had rad something that said it was being worked on, but maybe I just got too excited over something I read. or do I just wait-and-see? :-) regarding SpeedStep, I have a 150W proc with an underperforming 240mm liquid cooler (maybe the paste doesn't done right when I bought the box, this can make cpus run too hot). My only concern is that when all threads are used concurrently like in a burn/stress test such as prime95, the I mentioned this before. How I mitigate this problem is I scale processor usage "Max Processor State" back to 79%, which is a power setting. it would be really interesting if you could query the cpu's TCASE temperature and make sure that the package temp doesn't go beyond that temperature TCASE-7.8degC to TCASE-10degC through monitoring. seems like just around TCASE-5.8degC it either stays where it is, or just goes hotter and hotter. I want to monitor ahead of time and actively mitigate any problems by preventing package temp from going beyond that point by scaling max processor usage percentage, rather than simply setting a max processor usage percentage in a .ini file or in the registry (well, handy for building management, but not for general computer users)? the OS should do that. you can query the cpu package temperature. why not have the ability to query TCASE as well? this would solve the problem of laptop cpu overheating as well. cooling management with these higher-wattage procs like the Intel Extreme series, or the Xeon procs lead me in this direction. ------------- Jim Michaels jmich...@yahoo.com j...@renewalcomputerservices.com http://RenewalComputerServices.com http://JesusnJim.com (computer repair info, programming) ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ Freedos-devel mailing list Freedos-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-devel
