|
A minor critisism, but I'd add the following:
typedef std::vector<long> collection_t;
typedef collection_t::value_type value_type;
typedef collection_t::iterator iterator;
typedef collection_t::const_iterator const_iterator;
Then I'd change all methods that return std::vector<long>::iterator to return iterator instead. Also I'd replace the std::vector<long> _Grades; with collection_t. I'd also rename grades_begin to begin and grades_end to end and add const versions which return const_iterators.
With all this inplace:
- The collection type can be changed in one palce.
- Code like the following will work as expected:
Student &s = ....
Student::iterator i = s.begin();
Student::iterator e = s.end();
Steve
|
|
|
|
|
Agreed, but I had to leave something for his academic exercise 
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Humm!! his Lecturer might be Seeing this tooo  .. anyway nice code!
"Opinions are neither right nor wrong. I cannot change your opinion. I can, however, change what influences your opinion." - David Crow
cheers,
Alok Gupta
VC Forum Q&A :- I/ IV
Support CRY- Child Relief and You
|
|
|
|
|
ThatsAlok wrote: his Lecturer might be Seeing this tooo
Which is why I left plenty for him to do. I just wrote the code to read the file (and even it is incomplete at the moment ... several things should be cleaned up a bit to actually meet professional standards). Typically, students will write code that will read all that data in and store it in several unrelated arrays (that they are relating in their head, but not in code) and do whatever the assignment asks by manipulating that data. This makes the code very unreadable and enforces a horrible habit that will only get worse if not corrected (which is the only reason I offered as much code as I did for this post). As a side note, when I was in college, I actually had a few professors who would count an assignment similar to this one as wrong if the code produced was not up to professional standards. Just some things to keep in mind.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Nice example, Zac. It's quite informative.
Zac Howland wrote: _Grades.insert(back_inserter(_Grades), grades.begin(), grades.end());
The compiler complains with:
error C2664: 'void __thiscall std::vector<long,class std::allocator<long=""> >::insert(long *,unsigned int,const long &)' : cannot convert parameter 1 from 'class std::back_insert_iterator<class std::vector<long,class="" std::allocator<long=""> > >' to 'long *'
"Approved Workmen Are Not Ashamed" - 2 Timothy 2:15
"Judge not by the eye but by the heart." - Native American Proverb
|
|
|
|
|
hmm, looks like it is trying to call the single insert instead of the bulk insert. You can change it (without too much efficiency loss) to:
copy(grades.begin(), grades.end(), back_inserter(_Grades));
What STL implementation are you using by chance?
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Zac Howland wrote: What STL implementation are you using by chance?
I do not know. It's the one that ships with VC++ v6.
"Approved Workmen Are Not Ashamed" - 2 Timothy 2:15
"Judge not by the eye but by the heart." - Native American Proverb
|
|
|
|
|
Actually, I think I know why it might have done that. Since the old Dinkumware implementation used pointers as iterators (basically, typdef a pointer to iterator for a given container) for many of the containers, it probably couldn't easily figure out that it should call the bulk instead of the single. I'm not 100% sure on that, but I was able to reproduce it with an older gcc implementation.
Another alternative is to use the insert algorithm which only takes iterators. However, copy works for this case.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Zac,
I was going back over this code and was confused on a few things. In main(), the call to copy() is made once for the entire file. This, in turn, is supposed to call operator>> once for each line of the file. Is that correct? Using the debugger, I see operator>> getting called only for the first line of the file. I'm not seeing the students vector grow at all. What am I missing?
Thanks,
DC
"A good athlete is the result of a good and worthy opponent." - David Crow
"To have a respect for ourselves guides our morals; to have deference for others governs our manners." - Laurence Sterne
|
|
|
|
|
Looking back at it, is a slight bug in my example code. Basically, since the id's and grades are both longs, there is no way to tell them apart. It would probably be helpful to modify the output slightly by putting a grade count in between the student id and the first grade so that you can modify the copy call (for obtaining the grades) to just grab that many items (instead of grabbing everything that is left in the file). Another (possibly simpler alternative) would be to grab the student id, and then use getline to obtain the rest of the line, place that string into a stringstream and then call copy on it (this solution has the added benefit of not having to change the input/output file format).
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
My bad, Zac. I meant to comment on this thread.
"A good athlete is the result of a good and worthy opponent." - David Crow
"To have a respect for ourselves guides our morals; to have deference for others governs our manners." - Laurence Sterne
|
|
|
|
|
No problem. I figured as much 
It wouldn't take much to get that to be working code, but originally it was just meant as a nudge in the right direction.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
q_p wrote: where the first number of each line will be a student ID number and the next five numbers are grades they got on 5 different tests.
Reminds me of a gradebook program I wrote back in college. There's a lot you can do with this, once you get past the basics.
"Approved Workmen Are Not Ashamed" - 2 Timothy 2:15
"Judge not by the eye but by the heart." - Native American Proverb
|
|
|
|
|
When my program blows up, sometimes I get the following error on the next build. To date, the only solution I have found is to close VC++ and restart the computer. This works, but is a bit time consuming. It actually tempts me to get up and wander around while the machine restarts.
LINK : fatal error LNK1104: cannot open file "Debug/tradingdemo.exe"
So, any ideas on how to fix this without restarting?
Thanks
|
|
|
|
|
Oliver123 wrote: So, any ideas on how to fix this without restarting?
Yes
Oliver123 wrote: When my program blows up
Fix the bug that causes the crash. Also you can put exception handling in your code that would notify you of the error without crashing the process.
led mike
|
|
|
|
|
led mike wrote: Also you can put exception handling in your code that would notify you of the error without crashing the process.
I'm a member of the school that thinks this is rarely a good idea. If you're got a crash let it crash; it makes the bug easier to locate and fix. Also it means that JIT debuggers can be used. Also for clients, Dr.Watson will generate a crash dump file (if configured to do so) which can be used for postmortem debugging. In short if a program crashes, the best thing for it to do is crash.
Steve
|
|
|
|
|
Well he is talking about a "developer problem" not a production problem. I use exception handling to debug all the time. I find that a stack trace is extremely helpful in locating bugs. 
led mike
|
|
|
|
|
Oliver123 wrote: When my program blows up...
ALWAYS wear eye protection. I prefer goggles and a flight helmet.
If you're running in the debugger (sounds like no) you can Shift-F5 stop debugging.
Otherwise CTRL-ALT-DEL, find tradingdemo.exe in the list of running processes, and end the
process.
|
|
|
|
|
You forgot the final step:
format c:
led mike
|
|
|
|
|
led mike wrote: You forgot the final step
Yeah I guess a quick format isn't too "Time consuming".
edit: format /Q c:
Mark
|
|
|
|
|
I bet that the "tradingdemo.exe" process is still running. Check in task manager. If it is just kill it and you should be able to rebuild.
Steve
|
|
|
|
|
Hi,
I have a Dialog application. In some cases I am displaying data in the main dialog by accessing data in an array:
str2.Format("%02d\t, m_Parse.mChannel[i].m_SV);
str+=str2;
SetDlgItemText(IDC_SAT, str);
etc.
etc.
What I want to do is to have some of the data displayed in different colors other than black. How can I set the color of some data before displaying it?
Can someone help me out?
Jerry
|
|
|
|
|
I think the only control that would allow you to mix colors in the text would be a rich edit
control. You could use read-only rich edit controls.
If that's way overkill for just colored text, you could render the text yourself in response to
the WM_PAINT message using perhaps some/all of these functions:
SelectObject() to select a NULL brush (NULL_BRUSH) into the DC or possibly another font as well
SetTextColor()
SetROP2() to set the foreground mixing mode (to R2_COPYPEN, for example)
SetBkMode() to set the background mixing mode (to TRANSPARENT, for example)
DrawText() to draw the text
Just two possibilities
Mark
|
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
