| |
| |
| |
|
Page: 1 2 3
Comments:
<0> hello
<0> does somebody here knows if there's a way to find out the creation date of a file?
<0> file atime returns the last access time file mtime returns the last modification time
<1> yes
<1> u need a MYSQL databse to do this first
<0> why??
<2> Mythox: use "file stat"
<2> file stat $path foo; puts $foo(ctime)
<0> oh ok thanks
<0> is there a way to set that value too?
<2> nope
<3> yeah you can
<2> with an external bin, yes
<3> naw
<3> you can use file atime ... or file ctime
<2> there is no file ctime command
<2> you're thinking of atime and mtime
<3> hrm...guess so i swore there was an atime/ctime/mtime
<3> guess i never noticed that
<3> you would think if you can set access and modification you could set creation
<3> Mythox, just exec the 'touch' bin standard on all *nix systems
<2> most nix filesystems don't have a creation time for inodes
<2> which is why there is no 'file ctime' command
<3> bsd does
<2> freebsd's UFS records the birthdates (which is the same thing)
<2> yeah
<2> ufs in general
<3> right
<3> all bsds are ufs based
<3> free/net/open/etc...
<3> ffs/ufs are pretty closely related as well
<3> and besides if the file command supports switches for mac osx and darwin...dunno why it wouldn't support ctime for *bsd filesystems
<4> =)
<5> ctime is inode modification time
<6> mh
<7> if i want to store a two paired value like <name>, <id> would i use a dictionary?
<8> if -> http://www.tcl.tk/man/tcl8.4/TclCmd/if.htm
<7> i would just use an array
<7> but then what are dictionaries for?
<0> just to store and search lists of plenty of words
<7> ah
<7> and thats faster then a nested list?
<7> anyway array is good for my purpose then i guess
<7> so now i just need to parse the file to get name and id i wonder whats faster loading the file into memory first and then parse it or directry parse it by going through it line by line
<9> the latter
<9> since loading it first does not yield any advantage
<9> you traverse the file and store its contents in O(n)
<7> well going through it while in memory is probably faster, no since reading whole file is faster then many little reads
<7> ; string range
<10> Gotisch: #878 (1135 clicks) Tcl error: wrong # args: should be "string range string first last"
<7> mh something like that
<7> while {[gets $file line] >= 0} { if {[string range $line 0 4] == "GUID"} { set a [string range $line 6 end] } }
<7> where a would then be the guid
<7> just need name then
<9> wrong
<7> what?
<9> reading a file line by line is very fast
<9> your libc reads the file chunk-wise anyway
<9> reading the file to memory first just doubles your mem requirements
<9> just use a line-wise read primitive like gets
<7> ye seems fast enough
<7> ill test it with inmemory too
<9> otoh - with your 100 line testcase you wont see any difference
<9> Gotisch: you ever written a scanner or parser? or heard a lecture on it ?
<7> mh the file im parsing is 30mb
<11> dict ?
<10> dict -> http://Tcl.Tk/man/tcl8.5/TclCmd/dict.htm
<7> ok testfile is just 1mb
<7> but still
<7> 120000 lines
<9> Gotisch: ok, for that kind of size you _want_ to read/parse the file on the fly
<9> it's _meant_ to be done that way
<7> i can imagine :)
<7> well the fastes i found was just sourcing the file
<7> but its kinda hard since its not in the right format :] and i dont want to rewrite the unkown proc
<9> that's because the primitive tcl-loader that is nativly implemented parse the file
<9> if you can do that - go ahead
<7> ill stick with line by line
<7> seems to be fast enough
<9> hehe... "fast enough" funny. it's the fastest there is
<9> again: the libc reads the file chunk-wise anyway
<9> probably at 4k or 8k chunks
<9> or even mmap()s it and leaves it to the kernel to fill in
<7> i mean fast enough to do it in tcl
<7> else i would have to parse it in c and then just p*** it to interpreter
<11> how are you treating the file since you dont do a read, Gotisch?
<9> yeah
<7> treating?
<7> gets
<11> i mean, what are you doing with the data you get?
<7> saving it in an array
<11> hrm
<7> actually it will be array(name) -> id
<11> a textfile you need to parse or what?
<7> yes
<7> i need to get name and id out of it
<11> oh, right
<7> because in the db is only saved the id
<7> so ill allow searching by name
<7> by looking up right id for it
<11> wouldn't 'set dat [split [read [set fid [open your_file.txt r]]] \n][close $fid]; foreach ln $dat { do_your_stuff }' be a proper alternative?
<7> mh
<7> ye i tested it
<7> % time {readandparsedirectly}
<7> 1123802 microseconds per iteration
<7> % time {readinmemorythenparse}
<7> 882958 microseconds per iteration
<7> seems reading it into memory first is faster
<7> but not by much
<7> gotta test it with bigger file
<11> 1mb is not a lot to keep in mem these days
<7> ill try with the 30mb file
<11> 30 MB of text?
<11> that's crazy ;>
<7> yes
<7> dont get me started :]
<9> Gotisch: did you account for the file reading time, too?
<7> yes
<7> ill show you the procs
<9> include the loading and time everything together
<11> from what he showed they were divided into two different procedures, so most likely they were stand-alone procedures
<7> http://paste.tclhelp.net/?id=5du
<11> i gotta eat
<9> Gotisch: depressing
<9> now, also measure the memory requirements :)
<7> :]
<7> tell me a good way to do so :p
<9> umh
<9> i dont do tcl :)
<7> you cant measure it in tcl imho
<7> and if you can you need debug enabled
<7> i think i dont have it enabled
<9> isnt there some global info command for the runtime env ?
<7> ; memory
<10> Gotisch: #964 (3141 clicks) Tcl error: invalid command name "memory"
<7> you need tcl with debug enabled to get memory
<9> ;info mem
<10> Raimi: #966 (1192 clicks) Tcl error: bad option "mem": must be args, body, cmdcount, commands, complete, default, exists, functions, globals, hostname, level, library, loaded, locals, nameofexecutable, patchlevel, procs, script, sharedlibextension, tclversion, or vars
<7> http://www.tcl.tk/man/tcl8.4/TclCmd/memory.htm
<7> lets hope my laptop can handle that file
<7> i dont have that much ram
<9> heehee
<7> % time {readandparsedirectly}
<7> 14944012 microseconds per iteration
<7> % time {readinmemorythenparse}
<7> 19956040 microseconds per iteration
<7> haha :]
<7> i guess i would need to do it like a hundret times to be sure
Return to
#tcl
or
Go to some related
logs:
#networking
theenglishmansion gallery
expr[llength...] in tcl
wetgorilla
#computers
ubuntu oidentd Unable to setup listening socket
#debian
#beginner
#gamedev
asians wear
|
|