| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8
Comments:
<0> could someone propose how i could test for the end of file with using the above?
<0> ok found it, if anyone is interested you can use the peek method
<1> uh, no
<1> i told you
<1> test your reading for failure
<1> if( some_stream >> some_var ) { /*read success*/ } else { /*unsucessful*/ }
<0> if i use read the help says that if an eof occurs the function calls setstate(failbit). But it doesn't say how i can check if it is set
<1> what does read() return?
<0> this
<1> there ya go
<0> *this in any case
<1> if( thing.read( ... ) ){ /*read success*/ } else { /*unsucessful*/ }
<1> .eof() exists in vc++2005
<0> for fstream it doesn't
<1> it's part of the standard, i'm not sure why you think it 'doesn't exist'
<1> yes it does
<0> try to type fstream mystream; mystream.eof()
<1> std::fstream fs; fs.eof(); //compiles just fine
<1> why are you using an fstream and not an ifstream / ofstream ?
<0> because i don't know from the beginning what it will be
<1> and you still didn't paste the error message, so *shrug*
<1> irregardless
<0> error C2039: '_eof' : is not a member of 'std::basic_fstream<_Elem,_Traits>'
<1> check your read for success/failure
<1> _eof ?
<0> ok i tried also _eof
<0> in the first i tried eof
<1> you did include <fstream> right?
<0> yes
<1> shrug
<1> maybe you messed up your include paths or something
<0> i tried what you said about std::fstream fs; fs.eof; but it still doesn't exist, i get the same error
<0> fresh installation
<1> fs.eof();
<0> yes
<1> well, it doesn't matter
<1> eof() isn't used to determine if a read was successful or not
<1> why are you using .read(), anyway?
<1> what exactly are you reading?
<0> what should i use?
<1> depends on what you're reading
<0> an ascii file
<1> i mean, specifically what you're reading
<1> what's in the file?
<0> numbers,and some characters inbetween, it's like a coding so i check every character one by one
<1> line by line, or do you read the entire file character by character?
<0> character by char
<0> is there a better way to read?
<1> .get() gets a single character
<1> although
<1> character by character is horribly inefficient
<1> I would read a bunch into a vector<char> and then do processing until you need more chars
<1> how big is the file?
<0> can be 400mb
<1> k
<1> then certainly read a bunch at a time
<1> maybe a meg for a single read or so
<1> character by character will crawl
<0> you mean when i do the read(char, 1000) or something like that?
<1> well
<1> you can do this...
<1> vector<char> buffer( 1000000 ); if( f.read( &buffer[0], buffer.size() ) ) { //process the data } else { //we can't read that much data.. buffer.resize(0); f.clear(); std::copy( std::istream_iterator<char>( f ), std::istream_iterator<char>(), std::back_inserter( buffer ) ); //process buffer.size() bytes of data }
<1> I think that would work fine
<1> when you hit the 'else' block there you can ***ume that it will be the last read done
<0> well it's a nice idea you gave me, i think i will reconsider the way i thought about it
<0> thanx for the help rdragon
<1> sure thing
<0> one thing i cannot seem to figure out is how to check read if it failed, i mean it says that in any case it returns *this and on eof it does setstate(failbit);
<0> hmm i don't know what happened but now it recognises .fail(), ok forget about it
<1> if( f.read(...) )
<1> that's how you check
<1> just like if( cin >> some_var ){ ... }
<0> although read fails, it stores what it read to the vector, is there a way to know how many elements of the vector are blank, or should i iterate throught it?
<1> nope
<1> thats what the else{ block handles in my code
<0> what do you mean when you say process buffer.size()?
<1> that tells you how much was read
<0> but when you resize the buffer you set it to zero, how will the next statement work? actually i tried it but probably the statement with the std::copy doesn't work
<1> copy copies elements from the stream until it hits the end of stream
<1> and it copies into the vector using the back_inserter
<1> which will push_back() the elements into the vector
<0> this means i should rewind the stream before doing the copy?
<1> no
<0> if i rewind it, it works, if i don't it stores nothing in the buffer, so i just use tellg() and seekg() and it works just fine:)
<1> oh, shrug
<0> it's ok, no worries, it works at the end:)
<2> !seen Elmpie
<2> o.o
<3> http://www.lpboulder.com/quotes/
<4> Does any STL container allow for shifting an element to the end?
<5> could you be a little more specific, what does "shift and element to the end" mean?
<6> I think he's refering about rotating the collection
<5> threat ??? you gonna answer or are we supposed to sit around all night guessing what you want
<4> vawjr, back
<5> so what's the actual problem?
<4> vawjr, what I want to do is insert int's in to a container (no great than a 1000). If an int being inserted is the same as an int already inthe container then I dont want to add the int in but instead shift / move the int inside the container so I have all of the mostly inserted values at one end and the least inserted at the other
<4> eg I insert 2, 3, 4, 5, 2, 4, 6, 7, inside the container I should have something like 2, 4, 3, 5, 6, 7
<5> <shrug> ok
<7> std::set, kthx
<4> yes, dont ask me why I want this :) I just do
<7> use std::set, that is
<5> no
<4> Cowmoo, yes but doesnt a set sort it?
<5> that will leave them sorted
<7> isn't that what you want?
<7> eer
<5> NO
<4> Cowmoo, I want it sorted to some extend, but I dont want it sorted on the value of the int
<7> looks like I miunderstood you..
<4> Cowmoo, I want it sorted based on how many of the same value has been inserted
<4> Cowmoo, like my example I gave above :)
<7> iohhh
<7> k
<8> can someone help me with a sipmle rectangle program thing? my rectangle isn't printing out right
<5> and yes you can rotate the sequence containers
<4> Cowmoo, inserting 3, 8, 9, 2, 5, 2, 1, 2, 3 should result in 2, 3, 8, 9, 5, 1 or something simular
<7> right
<8> http://rewtguy.sweon.net/programming/c++/code/rectangle.cpp
<4> vawjr, ok, so I can shift things around within, say, a vector?
<6> don't use a vector for this thing
<6> a list
<4> ok
<4> I will use a list
<7> no
<5> threat, of course.... sounds like you need book on the library
<6> tho, I don't know how to shift it, but you can just remove it and then push it back
<4> sk8ing, yeah, I was thinking of doing something like that
<4> sk8ing, I was just asking for suggestions before I try and hack it out :)
<4> sk8ing, I dont really want to extend on the list STL or anything, just need to implement something quick :)
<6> the thing with the list is that you should be able to move the element to the end very quickly
<6> with out removing it
<6> but I don't know how
<4> oh
<4> something to do with pointers? :)
<6> I don't think so
<6> it's a linked list so you basicaly would have to specify a new position
<5> seems like an odd thing to want
<8> NEVERMIND I SOLVED IT
<5> Riddle_B0x good
<8> not like anybody bothered to help though ^_^
<4> sk8ing, hmmm, shouldn't I use a vector instead of a list?
<6> no
<4> why?
<4> vector is designed for random access?
<6> yeah but not for random moving of elements
<6> if you'll want to move one element to the end you'll have to move all the other elements after it one position -
<5> well, all of the elements between it and the end
<4> ok
<4> and with a list I can just put it where ever?
<5> you could, but you said "one end"
<5> so what's the _real_ problem?
Return to
#c++
or
Go to some related
logs:
#AllNiteCafe
was the airfield where Charles Lindbergh landed to end his historic solo trans a
#linux
#Linuxhelp banned undernet
#MissKitten
4chan teens rules
#AllNiteCafe
site:www.quotesdb.info g0thicangel
#AllNiteCafe
#autumn undernet
|
|