Comments by k5054 - CodeProject

k5054 5 days ago View

My 5
To to OP: you can confirm this by changing your printf statement to:
printf("strlen(%s) is %d\n", a, strlen(a))"
which will then print out:

strlen(abc
) is 4

k5054 29-Apr-24 18:50pm View

void functions should not return a value.
You should also avoid "magic numbers" (e.g. MaxDateLength = 11). If you change the size of dateout you then need to find everywhere that you've used the "magic" 11. See Rick's example, where the size of the string is a #define. If you need to change the size, you only need to change the value of the #define, and recompile, and no overflows/underflows.

k5054 25-Mar-24 10:17am View

I missed the part where you were looking for a Windows version. There's no mention of windows AFAICS on the web site, and the Wikipedia page https://en.wikipedia.org/wiki/NaCl_(software) only shows that the supported OS is "Unix Like". So it seems that it's unlikely you're going to find Windows specific source. However, the source is available. You should be able to download the tar file from the web site. I'm not up on how well Windows does with bzipped tar files, but there must be loads of solutions out there. Or you could install WSL and play with Linux inside your windows environment.

k5054 21-Mar-24 16:50pm View

I assume you mean
formula = 5/(1+k) + 9/(2+k) + 13/(3+k) + 17/(4+k) - 16 ?

k5054 19-Mar-24 11:19am View

I can confirm that even as far back as gcc 2.95, compilation ends in an error "too few arguments to function".

k5054 18-Feb-24 10:13am View

My 5. Intuiting that the OP was trying to redirect the system() output into a variable is worth every point.

k5054 4-Feb-24 18:28pm View

K&R style function definitions have been removed in C23. But the OP is probably not being forced into C23.

k5054 20-Jan-24 12:56pm View

And your question is?

k5054 14-Jan-24 11:31am View

Go back and check the code. buf is the target. The source is a static string.

k5054 12-Jan-24 11:48am View

Sorta looks like hooks for a doc generator, something like Doxygen https://www.doxygen.nl/index.html
Perhaps generated automatically by your IDE?

k5054 19-Dec-23 12:33pm View

Suppositions:
char has no alignment requirements
int must be on aligned on 4 bytes

When you declare char str[n], you don't know what the alignment of str is. That is, (&str[0])%4 might be 0, 1, 2, or 3. If (&str[0])%4 is not zero, then coercing an int pointer from &str[0], and trying to dereference it would cause a fault. That's why I said at least on of the two casts in the original post would cause a fault.

k5054 18-Dec-23 13:12pm View

For x86, x86-64, arm-32, and arm-64, it is not a requirement that an int be on a 4 byte boundary. But arm7 (eg. raspberry pi 32 bit OS), does have alignment requirements for floating point. So if you were try the same thing on a PI w. 32 bit os, but with float rather than int you will get a bus error when you try to access/modify an un-aligned floating point variable.

k5054 18-Dec-23 12:05pm View

As far as I can recall &(str[n]) and (&str[n]) are equivalent.

You didn't so much get a failure for the case of str[3], but invoked Undefined Behavior. Assuming an int is 4 bytes long, then an int starting at str[3] occupies str[3], str[4], str[5], str[6], and str[7]. You defined str to be only 5 bytes long, so the values of str[5] .. str[7] are unknown. If you have other variables on the stack, you might be looking into variables declared either before or after them. Or you might be looking at padding bytes, or data outside the current function's memory area, or you might even be bumping up against the boundary of a memory page, in which case you might get a segfault. I'm sure there are other things that might be going on. The C standard doesn't say what happens when you try to access elements beyond the end of an array.

k5054 3-Dec-23 13:33pm View

my 5. For having the fortitude to examine the code dump, if nothing else :)

k5054 23-Nov-23 11:57am View

We could even delete the avg() function and use a lambda:

 averages.push_back( [v](){ return (v[0] + v[1])/2.0; }() )

k5054 11-Nov-23 16:28pm View

"it doesn't work" isn't very helpful. What doesn't work? Does it crash? Are the imported values garbage? Does the program seem to hang? Have you tried the debugger?

k5054 5-Nov-23 11:42am View

How are you assigning values to pSelected? The documentation for SendMessage indicates that all paramaters are IN parameters. If you're not assigning a value to pSelected[n] then the default seems to be to initialize the array elements to zero.

k5054 20-Oct-23 13:27pm View

I already did.

k5054 20-Oct-23 12:33pm View

Too bad we can't up-vote comments ...

k5054 20-Oct-23 12:32pm View

You don't even need to buy it: https://colorcomputerarchive.com/repo/Documents/Books/The%20C%20Programming%20Language%20(Kernighan%20Ritchie).pdf
This book is THE reference for learning C, IMO. I'ts what I used to get started, back in the pre-ANSI days. I'm sure its the go-to book for just about everyone, except perhaps Kernighan and Ritchie themselves

k5054 19-Oct-23 12:02pm View

See this SO response to a related question : https://stackoverflow.com/a/22745579
The author makes some interesting points that relate to using the LARGEADDRESSAWARE flag, that you might want to consider before using it. I agree with Dave and jeron1 : unless you know that you're going to need 2GB of data, you should start by doing some memory analysis and see where all your memory is going.

k5054 17-Oct-23 16:10pm View

int matrix[3][3] defines a block of memory to hold 9 ints. The declaration int **ptr; defines a pointer that points to a pointer to int. These are not the same thing! E.G.

int mat[3][3] = { {0,1,2},{3,4,5},{6,7,8} };
int **ptr = (int **)&mat;
printf("%d\n", mat[1][1]); /* prints "4" */
printf("%d\n", ptr[1][1]); /* seg faults! */

To get a pointer to a row of the mat, you need to use int (*p)[3] = mat; Now you can access the elements of mat through the pointer ptr as you would expect: printf("%d\n", ptr[1][1]);

I think this is a fairly common mistake. I know I've made the mistake of thinking a double-pointer e.g. **ptr is equivalent to a pointer into a 2D array, but it is not.

k5054 12-Oct-23 12:26pm View

Oops. Typo on my part. Should be
A *b = new A(a);

Updated solution to reflect this.

k5054 8-Sep-23 13:19pm View

If hunspell_add expects a non-const char *, then passing it a const char * could be the issue. ie hunspell_add may modify the the memory contents

Just because you don't explicitly change the value of a variable, doesn't mean that it doesn't get changed - you might have a buffer overflow, or an unitialized or badly initialized pointer, or ...

The debugger can help with the exit code. But I think you should get a result code if you launch from within you IDE. I'm not a windows guy, so others might be able to help you more with that.

Note that the debug and release versions might work differently. This is because of optimizations the compiler can make which might change things e.g. variable lay-out might be altered for efficiency, so a buffer overrun in debug might not cause a problem, but will be found in release. Also, debug versions may initialize variables for you (eg int to zero, pointers to NULL), but this may not get done for release versions.

k5054 8-Sep-23 12:37pm View

My apologies. I'll try to remember next time.

k5054 8-Sep-23 11:57am View

No, you do not: cf https://www.geeksforgeeks.org/function-pointer-in-c/

k5054 21-Aug-23 13:43pm View

could it be something to do with the type of m_current? If its not a uint8_t or equivalent, then m_current += size may not be doing what you expect. But that's all I got, and it's probably wrong :(

k5054 20-Aug-23 19:07pm View

What is read_data() doing? Is it somehow causing the file offset pointer to be somewhere other than current + s when it returns?

k5054 20-Aug-23 14:46pm View

"That will give you a different random sequence of numbers every time you run your program"
Yes, but only if you don't start the program multiple times in the same second. That's probably not an issue 99% of the time, but something the OP should be aware of. The "randomness" of the system rand() function is also of dubious quality. Again, for 99% of tasks, particularly homework and/or personal projects that don't need good quality randomness, then its probably sufficient. If to OP does need better random data, then the facilities provided in the <random> header may be more useful.

k5054 5-Aug-23 13:33pm View

This is probably not a good fit for bash or any other shell interpreter. You might find this much easier in python or similar scripting language.

k5054 28-Jun-23 11:42am View

I don't know what you mean. But see Richard's comment about the code that's generates the error. He and I are basically saying the same thing. What you have posted here seems to be correct. So there's something else that's causing the error. Are you including the declaration for class node in the file for main? Have you inadvertently redefined what a node is? Maybe something else....

k5054 23-Jun-23 15:23pm View

This applies to pre-increment and pre/post decrement of loop control variables as well. In general modifying a loop control variable inside the loop is not the right thing. There are times when you do want to modify the loop control variable, but not usually, especially when using the loop control to iterate over a data collection of some sort.

k5054 28-May-23 11:56am View

Would std::any do the job? https://en.cppreference.com/w/cpp/utility/any

k5054 22-May-23 17:29pm View

In addition to what OriginalGriff has said in solution 1, consider that the web is not a homogeneous sea of Windows X86 systems. What about people on Apple OSX or Linux? Or any of the other multitude of OS's out there. There's even Windows on ARM these days, which may or may not be able to execute a windows x86 .exe.

k5054 20-May-23 11:33am View

capacity is only available for std::vector. Furthermore, using swap does not release process allocated memory, at least not for std::list. Nor does the object going out of scope release memory allocated to the process. For example, suppose we put the body of OP's main function in a separate function, and call that from main. ps will report the same data resident size at the end of the function, and just before main returns. If we call the allocating function twice, the first time allocating a large number of list elements, and the second time with a small number of elements, the drs does not change between the first and second calls.

k5054 5-May-23 11:45am View

Turbo C++ is interesting as an archaeological artifact, but otherwise should not be used. It does not support any modern (e.g. C++11 or later) C++ standards, so almost any tutorial you might google upon will not be helpful. Since Visual Studio has a free download option, I recommend that you look into that, rather than sticking with something that is obsolete.

k5054 19-Apr-23 11:14am View

This compiles cleanly on compiler explorer, where I tested g++ 8.5, g++ 12 and MSVC v19.30. So its one of: you're using a broken compiler, you've posted the wrong code or your compiler flags are wrong - i.e. trying to compile this with C++ 98. It compiles cleanly with gcc C++11, C++ 14, C++17 and C++20.

k5054 17-Apr-23 15:37pm View

The usleep() man page states that POSIX.1-2001 declares this function obsolete; use nanosleep(2) instead..

k5054 8-Apr-23 20:15pm View

Google translate: This script is scheduling some tasks, however, it is not scheduling. Can you help me?

k5054 6-Mar-23 12:46pm View

You should probably also include the signature to BTUL->ProcessCommand_Index(), or at least a link to it's documentation (assuming its not a method of your own devising). Passing an index of 1 seems like a bit of a red flag. But not knowing what verify is or what the function does, or arguments expected makes this a bit of a black hole for anyone not familiar with the objects in question.
The only thing I would note is that you probably want const char *command = "blah". If you actually need a non-const string then you should probably define it as char command [] = "blah".
Other than that, the only thing I can suggest is fire up the debugger, and see what ProcessCommand_Index is receiving. Hopefully, wherever that comes from, there's a debug version that you can use to work out whats going on.

k5054 6-Mar-23 12:34pm View

In the given example you're passing something called command to function, not command_TEST. Is that an mistake when you typed in the provided snippet, or is that the bug in your code?

k5054 25-Feb-23 15:26pm View

should pushq call pushd, or is that a typo?

k5054 13-Feb-23 10:49am View

You said in your original question that you already have a program that prints some labels. Does it use an SDK, or just send ZPL commands to the printer? If it uses an SDK, you'll have to track that down and get the docs for it. Otherwise, see the programmers manual noted above, and make the changes you need.

k5054 10-Feb-23 10:52am View

No idea. I did all my ZPL programming on a Linux box. Testing and learning were all done with text files and curl. I'm old school I guess.

k5054 4-Feb-23 13:29pm View

You do know that you can do
systemctl enable <service>
to get a service to start at boot, right?

k5054 3-Feb-23 13:21pm View

There's a few options for the return value. It could be void, if there is no possibility for set_rightneigbor() to fail. If there is a possible failure mode for set_rightneigbor() you might return an int with 0 meaning success and other values indicating the failure type. Alternatively, you could return a gstate, which might be the previous value of the rightneighbor. This might be useful if you need to know what that value was before it was changed.

k5054 3-Feb-23 11:20am View

Even 64 bit numbers aren't large enough. If I understand this stackexchange answer correctly :
https://math.stackexchange.com/a/624296
you can approximate the sum of primes using
sum = (x^2)/(2 * log(x))
For a value of 100^6 (10^12) that gives a value of magnitude 10^22 or 2^75

k5054 2-Feb-23 11:32am View

Note 2: mykernel.bin and mykernel.Bin are different targets.

k5054 2-Feb-23 10:34am View

Also, the rules for a target need to be indented using 1 TAB character. You may already have that and the cut/paste to here removed them.

k5054 31-Jan-23 14:28pm View

Yes, and we've all done it, or something similar.

k5054 24-Jan-23 17:21pm View

My 5. Your solution was more correct than mine.

k5054 24-Jan-23 16:32pm View

It looks like localDevices.allDevices() returns a container of some sort. So, correcting myself, maybe you want

auto allDevices = localDevices.allDevices(); // gets a list of devices
for( auto& device : allDevices ) {
    // device has the same type as the contained object of allDevices
    auto& name = device.name();
    std::cout << name << '\n';
}

If you don't need to modify the device object within the for loop, then you should declare device as const auto& device

k5054 10-Jan-23 18:15pm View

Any path that does not start with a / is a relative path to the current directory. Every directory has two special entries, . and .., which are this directory and the parent directory. I'm not aware that the "../" symbol actually has a name, its just referred to as "parent", if at all.
Maybe what you're struggling with is how to get the compiler to search other directories for include files. From the command line you would say cc -I project/includes prog.c -o prog, where -I tells the compiler to add project/includes to the search path. Then when you say #include "project.h" the compiler will also search projects/includes for the "project.h" file. A compile command can have several -I directives, only limited by the maximum length of a command line. If you are using a IDE, say QtCreator, you'll have to look into the documentation on how to add the include paths to the compile command.

k5054 13-Dec-22 14:05pm View

It's not a requirement that a column that is used as a foreign key reference use the same name as the column that it references, they just need to be of the same data type. In this case, your DBS is trying to tell you've already created an object called "employee_pk", which is the name of a constraint on your employee table. When you define your salesperson table, the name of the constrain needs to be unique. The actual name isn't important. You could just as easily do

create table employee(
  emp_id int,
  name varchar(20),
  constraint "jim_bob" primary key (empl_id) );

create table salesperson(
   empl_id int,
   shift varchar(20),
   constraint "billy_bob"  primary key (empl_id),
   constraint "daisy_duke" foreign key (empl_id)  );

As long as the constraint names are unique, and don't conflict with other database objects (tables, indexes, etc), then you're good to go. It's rare to need to refer to a constraint by its name, so the actual name is usually unimportant. Having said which, its a good idea to use a naming convention that makes it clear to a human looking at the database schema what an object might be.

k5054 10-Dec-22 22:23pm View

Deleted

And what is your question?

k5054 7-Dec-22 12:52pm View

If I had a $ for every time I'd done something similar ...

k5054 25-Nov-22 9:59am View

Correct.
Or you could do it in one line

   return choices.get( rand.nextInt( choices.size() ) );

k5054 23-Nov-22 14:47pm View

when using std:string you can just call getline(std::cin, mystring). See the example at cppreference.com: https://en.cppreference.com/w/cpp/string/basic_string/getline#:~:text=%27)%3B-,Example,-The%20following%20example

k5054 23-Nov-22 11:12am View

As I understand this: https://en.cppreference.com/w/cpp/io/basic_istream/getline#:~:text=count%2D1,is%20executed). you do not need to subtract 1 for the length of the input buffer. Still, better safe that sorry. But better yet, why not use the std::string version of getline() and not worry about buffer overflows?

k5054 22-Nov-22 15:27pm View

You might get some suggestions here https://stackoverflow.com/a/115735
As to deleting a question, I have no idea. It might be that you can't delete a question that already has solutions posted. And I'm not sure where you would go to ask ...

k5054 21-Nov-22 18:14pm View

Why not? It's a perfectly valid way of ending a loop.

k5054 10-Nov-22 10:26am View

This is probably not the best place to ask about this. Maybe the good folks over as CMU can help? https://sourceforge.net/p/cmusphinx/discussion/help/

k5054 8-Nov-22 10:09am View

If you're trying to connect your client to netcat, then netcat needs to be in listen mode nc -l localhost portnum

k5054 8-Nov-22 9:04am View

How are you invoking netcat? The issue might be on that side rather than the code presented here.

k5054 2-Nov-22 14:29pm View

I added mine after yours, but thought the OP should know the difference between #include <file.h> and #include "file.h", etc

k5054 31-Oct-22 17:37pm View

In this case the reverse() function would be the function you supplied a def for. Suppose instead of def reverse(text): you had declared def myreverse(text):, then you would call print( myreverse("Hello world!") to get the output you want.

k5054 31-Oct-22 12:27pm View

Show us your work! We'll be glad to help you with the parts you are stuck on, but you need to show us how far you've got and what problems you're facing.

k5054 24-Oct-22 12:57pm View

It's unlikely that the good people here will just give you a homework answer. That's not the way we do things here. But it sounds like you've got started, but you're now stuck. If you can give some indication of what's not clear to you, you may get some helpful response. Use the "improve question" link and add some more detail.

k5054 18-Oct-22 17:12pm View

SQLite is not a "multi-user" database. The entire database is locked for update. There are many use cases where that is not acceptable. The OP needs to either go with a DB that can do the heavy lifting for them, or will need to "roll your own" data access, which, of course, is far more difficult than plugging in to SQLServer or Postgresql or MySql.

k5054 1-Oct-22 13:47pm View

"It doesn't work at all".
Does it crash?
Are there assembly errors?
Does it produce the wrong output?
Does it hang?
Does it never stop?

Please be more specific about what "doesn't work at all" means.

k5054 13-Sep-22 16:52pm View

This is a fairly straight forward problem, with a fairly clear description of what you need to do. Where are you having problems? What don't you understand?

k5054 6-Sep-22 13:09pm View

I seem to have missed the trailing semicolon in the line: for(j = 0; ...);. You also have a stray space at the end of the "is all good\n "; line, which may not be what you want.
Now would be a good time to get acquainted with the Debugger... It may have helped you narrow down this issue, right away...

k5054 10-Aug-22 10:57am View

You'll have to discuss the particulars with someone who knows, most likely a lawyer that is familiar with the legal requirements of your country.

k5054 31-Jul-22 16:10pm View

Just bracket the lefthand side of the trinary op e.g.

int sz = (m_buffer_size - m_buffer_index) < size ? m_buffer_size - m_buffer_index : size

k5054 28-Jul-22 9:09am View

Beat me to it by about a minute.

k5054 26-Jul-22 21:31pm View

Better yet, rather than manually counting the number of chars in the string, just use the length() member of string.

pass += data[ rand() % data.length() ];

k5054 23-Jul-22 6:57am View

Nothing. That $ before Regular is a typo too. Of course, it works fine if you you don't have use strict; in effect.

k5054 14-Jul-22 18:29pm View

Lets take a quick look at your requirements:
60 + clients
100 + files @ 1 GB each
I make that over 6 TB of data to transfer. With a 1 GB NIC, That's going to take about 90 minutes to transfer, assuming there's nothing else using bandwidth at the same time.
First question: Can the data be compressed for transfer?
Second question: Would some sort of network filesystem (e.g NFS or Samba) be an acceptable solution? That would mean that rather than uploading the files to the server, the files would always be available to the server. So, rather than updating the server once per day, updates would be live, as and when they occur on the client.

k5054 13-Jul-22 13:30pm View

"I like a solution using Linunx"
Well then the vim solution should work for you.
And VsCode has a linux version.
If you're using KDE, then maybe take a look at kate, which also folds code on #ifdefs ... but it folds from an #ifdef to the matching #endif, where VsCode folds just to the matching #else, when there is one. geany is also a possibility.

k5054 7-Jul-22 17:16pm View

Why no YouTube? Jason Turner's C++ weekly column did a short intro to C++ regex some time ago. Its only 8:30 minutes long, so a quick watch, and quite clear and understandable. It does assume that you know a little bit of regex's already, so if you're struggling with how to build a regex, it won't be too helpful. If you're just trying to figure out the match portion, this might be enough to get you going. https://www.youtube.com/watch?v=IOxKjqC1Ozo&list=PLs3KjaCtOwSZ2tbuV1hx8Xz-rFZTan2J1&index=65
If you're learning C++, some of his columns might help clear up some of the mysteries.

k5054 24-Jun-22 23:40pm View

"i[s] doesn't even compile ... it's meaningless garbage."
Not true. In C and C++ 5[x] is perfectly valid, as is explained here:
https://stackoverflow.com/a/381549

I should add that except for lecture notes, trying it for yourself, obfuscated programming contests, and just someone being a right git, you should never see this usage in real life.

k5054 24-Jun-22 13:49pm View

If your PC and IPad are both connected to the same WiFi network, then any client/server TCP solution would work. Likewise, if the PC has bluetooth, then some sort of bluetooth transfer should be possible. Then there's USB, in which case you might have to treat the IPad as a storage device on the PC.

k5054 15-Jun-22 12:05pm View

I just tried this within VS, rather than compiling directly on the ubuntu WSL instance, and get the same behavior as you do in the "Linux Console Window" that VS gives you. But running the VS compiled executable on the Linux host (in this case a Pi, but assume its the same on an Ubuntu WSL instance), the executable performs as expected, I.E. a bare return works as expected. So it's something that VS is doing in the Console Window. Bleah. Presumably whatever that is is eating the CTL-C too. All I can suggest at the moment is to try running the program from an Ubuntu console, rather than the VS Linux Console Window.

k5054 2-Jun-22 12:38pm View

"Maybe lowering the priority needs permission too". Any user can reduce a processes nice value. The intent is that if you know you have a resource hog, you can make it "nice" and run at a lower scheduling priority. Important back in the days of single processors and limited RAM, etc. But you have to have root privileges to escalate a processes priority above the base of zero.

k5054 1-Jun-22 9:00am View

It looks like the remote scp command is trying to use password authentication. Are you sure you have the correct path for your /private/key/file, and that its the correct identity file for the remote host?
You could try turning off password authentication by adding -o PasswordAuthentication=false to your command line, You might also try adding -vv (yes, 2 v's) to get debug output to try to see why the remote scp is not working.

k5054 25-May-22 9:23am View

You've tagged this as C++, but it's clearly Java. Can you fix the tags? That might help getting help with your issue.

k5054 19-May-22 12:20pm View

Show us the parts that are causing you issues, and we'll try to help. We're not going to give you a full solution, thought.

k5054 12-May-22 11:14am View

Ah. Ok, so you're trying to do floating point math, whereas the shell only supports integer math. What you'll need to do is to use something like perl to do this, eg elapsed=$(perl -e "print $end_time - $start_time")

k5054 12-May-22 11:04am View

when using let, do not add spaces to the the statement. Use let var=a$-$b not let var=$a - $b

k5054 21-Apr-22 12:57pm View

I didn't look too closely at your code .. the problem is here

int main(){    new Admin[5];

While that does create an array of 5 new Admin objects on the stack, it does not associate it with any variable try instead

Admin *admins = new Admin[5];
// ...
     admins[1].setvalues(/* ... */);

k5054 8-Apr-22 11:10am View

What is it you're trying to achieve? The packages to for SSL for PHP are probably available as standard RHEL rpms available through dnf. If you've got those installed and now need to know how to create an SSL certificate, there are two options available to you. If you do not need to have global access, you can google how to generate a "self signed certificate". These are fine for internal development, and/or internal services that only you or your company is going to use - you should be able to add security exceptions in your browser to accept the cert, even though it does not come from a recognized authority (e.g. Twate). If you do need to be accessible from the Web, then you need to get a cert from a recognized authority. You'll probably want to research Certificate Authorities, what the annual cost is, etc. Each authority will probably have a submission form for you to fill out to get your Cert.

k5054 2-Apr-22 10:52am View

I thought it worth while to explain why the compiler accepted the code. I'm sure we've all bumped into this from time to time. And will probably do so again. Fortunately, compiler warnings and debuggers make finding the brain-fades much easier these days!

k5054 1-Apr-22 13:06pm View

"the expression (pw,password) == 0 is invalid"
I think maybe you meant to say the expression is incorrect, not invalid, as it compiles just fine. If you turn on /Wall, the compiler will warn you that the expression before the comma has no effect. (Although that turns on a lot of other things you probably don't want, so use sparingly...)

The OP has inadvertently invoked the comma operator, see https://en.cppreference.com/w/cpp/language/operator_other#:~:text=function%20called%0A7.000000-,Built%2Din%20comma%20operator,-The%20comma%20operator
In brief (pw,password) == 0 evaluates pw (the address of the pw variable), throws that away, then evaluates password (the address of the passwword variable), and then uses that value as the left-hand-side of the comparison operation. Since password in on the stack, its address will never be zero (NULL), so the second comparison will always fail.

k5054 26-Mar-22 21:40pm View

Would it be correct to assume that you work with a buffer then tell the display that you have some data for it? In which case, if the [8x8]xn array for the display is contiguous, can't you just work with the 8x8 cells as a 8x8xn array and then inform the display that you've got data starting at location X of length n (or whatever it requires)?
Just trying to understand the problem ... since an int X[a][b][c] covers a*b*c*sizeof(int) bytes, its directly analogous to int X[a*b*c] If that works for you, you could venture into the dangerous world of type-punning or abuse of unions.

k5054 24-Mar-22 11:36am View

The OP has marked this question as C, not C++, so struct node is required. It is also the case that C does not require that assignments from void * be cast to the correct type, but it is still a good idea.

k5054 23-Mar-22 13:57pm View

dnf downloads and installs packages for you. All you need to do is as noted above
sudo dnf install libmcrypt-devel
That will download and install the latest libmcrypt-devel package (headers and probably a pkg-config profile). If you do not already have the latest libmcrypt package installed, that will be downloaded and installed or updated too. If libmcrypt or libmcyrpt-devel requres another package, for example say libImNeedtedToo, then it will be downloaded and installed for you too, as will anything libImNeededToo requires, and so on.

k5054 21-Mar-22 11:10am View

That sounds like a good way to allow a malicious user to grab root on your Pi, probably not something you want to do. In general anything that is user controlled and needs to have sudo root privileges to execute does not pass the "smell test", and would probably be banned by any competently run IT department. What you probably want to do is to re-think your startup procedure. Maybe do something with the pi users .bash_profile or .bashrc? Or modify the startup html document for chromium browser to use a value stored in a file (or database!) that a regular user does have access to without having to resort to sudo tomfoolery.

k5054 20-Mar-22 13:49pm View

Heh. Old Farts Brain Fart. Happens all the time :)

k5054 20-Mar-22 13:48pm View

See my reply to Richard, RE sudo nano. Another option would be to change the file owner to root andset the suid execute bit on the python script- if python scripts obey the suid permission settings. It's not clear why you'd want to do this from a script, unless you're writing a PI setup/configure script to have an automated way of producing PI's that perform some task, e.g. an info Kiosk, that you need to have several instances of.

k5054 20-Mar-22 13:43pm View

I think you meant just enter "sudo myscript.py" at the terminal. sudo nano will start the nano editor in sudo mode, at which point you'd need to do CTL-T to get a "command to execute" prompt, all of which seems a bit overly complicated.

k5054 17-Mar-22 22:23pm View

The compiler explorer (https://godbolt.org/) has no issues compiling your code. But I did need to add the /std:c++latest flag. I tested with X64 MSVC v19.29 VS 16.10 through X64 MSVC v19.latest. Maybe there's an issue with the 2022 Preview tools?

k5054 13-Mar-22 11:05am View

sudo su is a bit redundant. sudo runs the command line argument as another user - by default that is root, but you can specify other users with the -u flag. su does exactly the same thing. The difference being that sudo has finer grained controls over what users can execute which commands. If you need to get a shell as another user, you can use the -i or -s flags. So for example you could do sudo -u otheruser -i to get the equivalent of a login shell as otheruser.

k5054 11-Mar-22 15:21pm View

The ignore should happen just after reading in the single char for the confirmation response, not after the while statement. This is because if you do not answer si, you still need to remove the trailing newline before you call getline() at the top of the loop.

k5054 11-Mar-22 11:35am View

I would add a default: case to your switch block. Serial.parseInt() may not be returning what you think it is. In any case, it would be a good idea to let the user know when the input is invalid.

k5054 11-Mar-22 10:50am View

"how did you get the right solution?"
Experience, I suppose - although I'm quite capable of making this same error myself, even now - although I'd be much more likely to do another getline() rather than an ignore in this case, and then examine the first character to see if it matched, probably using tolower() so that case didn't matter.

k5054 6-Mar-22 16:50pm View

I assume you've googled for "C++ friend template class" and found this, which does not achieve what you want?
https://stackoverflow.com/a/8967610

k5054 23-Feb-22 18:11pm View

Two things:
1) You probably want to use float or double for your variables rather than int. Gravity near the earth is about 9.8 m/s^2, so a should be a floating point, at least.
2) The formula given gives you distance over time, To find the time the ball takes to drop from one floor to the next, you need to solve for t, which involves some algebra - I wish you luck.

k5054 22-Feb-22 13:59pm View

You have an erroneous value at the end of your list - eg 870780922 You'll have to take a look at your data file and see what's going on with it - otherwise your code works fine.

k5054 7-Feb-22 13:01pm View

Please add the actual error message that you are getting to your question. Otherwise, its impossible to know. Is the mysql server running? Is there a firewall issue? Are you trying to connect to the right port? Is your device and the PC on different network segments? Is networking on your device turned on? And probably many more possible issues.

k5054 6-Feb-22 17:39pm View

What compiler/OS are you using? Neither recent versions of GCC or MSVC give the error message you present here.
There is a possible typo in your Employee class, where the second constructor is spelled as Epmloyee

k5054 2-Feb-22 11:04am View

5'd. What a great mini-tutorial on how to go about finding and fixing syntax errors!

k5054 28-Jan-22 12:39pm View

But, but, but ...
you asked "How to select 2 values separated by a comma"! You need to give a exact example of your data! jsc42 has given you the right answer to the question you asked. So now you need to reframe the question. What, exactly, does your column data look like?

k5054 15-Jan-22 22:54pm View

I see no attempt at a linear search here. But unless you're required to write a linear search as part of your homework, then you should probably use the library standard lsearch() (or _lsearch() or _lsearch_s() depending on your compiler environment) routine.

k5054 9-Jan-22 11:15am View

Without some code to show us, we can't help. Are you calling the function correctly? Is the return value being used correctly or ignored? Does the function return a value for error and you're not checking?

k5054 5-Jan-22 13:11pm View

What compiler and options are you using to get that? I've tried with GCC and MSVC 19.30, both return an error about unrecognized literal suffix.

In any case int/int results in an int. In this case, 4/10 = 0, so multiplying by a double doesn't change things.

k5054 3-Jan-22 11:52am View

Every time you open a file with ios::out, the file is truncated. If you use the mode ios::app the file is opened in append mode - which appends written data to the end of the file. You may want to take a look at your Save_File function - it may not do what you think it does.

k5054 2-Jan-22 13:07pm View

Why? Assuming that you're using the free Turbo C++ 3.0, that goes back to 1992. There's better options available, unless you're stuck with a 1990's era computer and/or OS.

k5054 2-Jan-22 10:33am View

What compiler are you using? using namespace std; is absolutely correct for any implementation that supports C++-98 or later. Unless you're stuck on a DOS box, for some reason, then you should be using a new compiler. Visual studio is available as a free download (community edition), so you should consider that. There's also options for other compilers that could be used with eclipse or code-blocs or other free IDE's for windows.

k5054 30-Dec-21 23:07pm View

Presumably if you're using QString that means you're using QT. In which case, maybe you should use QProcess https://doc.qt.io/qt-5/qprocess.html
I'm not familiar with QT, but the web page says "The QProcess class is used to start external programs and to communicate with them", which sounds like its what you want.

k5054 30-Dec-21 11:30am View

While its not likely that a single input line is going to be over 10,000 chars long, you can guarantee that you won't end up with in the middle of some ginormously stupid input by using std::numeric_limits<std::streamsize>::max()

k5054 28-Dec-21 12:26pm View

+5 for inventiveness. Must be a slow day at work.

k5054 28-Dec-21 9:41am View

Thank you for your critique. Made my day.

k5054 28-Dec-21 5:22am View

"but it's not working" doesn't help much. What goes wrong? Does the program compile? If it does, what happens when you run it? Does it crash, or do you get unexpected results?

k5054 27-Dec-21 16:03pm View

And your question is?

k5054 26-Dec-21 17:17pm View

Without a base class to work from, how do you expect anyone to do anything?

Anything provided on the basis of this request is highly unlikely to integrate with any work that you've done. In addition, we have no knowledge of what topics you've covered in class. If we were to supply a template class using lambda expressions, for example, and those are topics you have not yet covered, your instructor might think that either you're a very advance student, or far more likely, guilty of plagiarism.

Maybe what you're trying to ask is how to derive a class from a base class? If so, see here: https://en.cppreference.com/w/cpp/language/derived_class If that page doesn't make much sense to you, then either that's way ahead of where you've got to in class or, more likely, you really need to speak with your Professor or Student Services and work out a way to get caught up.

Whatever you decide to do, it seems unlikely that you're going to be able to submit your assignment on time.

k5054 26-Dec-21 16:28pm View

Help we can do. Complete assignments we will not. I'm sorry you were in hospital. You should talk to your Professor and or student services. You'll not be alone in this. The school/uni you are attending probably has policies regarding prolonged absences due to health reasons - particularly at the moment with the pandemic.
But as to help, you'd need to show us your work, and give some details about what's happening (or not happening), with your code. But don't go dumping hundreds of lines of code on us. Try to break down the problems you're having into something small and concise.
If you can do Python, and understand Python objects, you've got a start. Start with creating a class for an employee, then derive an Hourly and Salaried employee from that. But you're going to have to do the heavy lifting. We'll help when you get stuck. But none of the regulars here will do your assignment for you

k5054 25-Dec-21 10:01am View

I've 5'd you, but I have to point out that the OP seems to have been given some function signatures to work with. Given that the requirement seems to be void EPF(double) and void CalculateBonus(double);, I'm not sure how to solve the problem without using global variables. Additionally the signature double DisplayTotalPay(double,double,double); seems particularly ridiculous, given that the purpose, extrapolated from the name of the function is to display something, there's no point in returning a value, and that other functions that should return a value don't. I wonder if the OP misstyped the assignment? If not, then I wonder about the qualifications of the instructor!

k5054 25-Dec-21 9:51am View

The algorithm library reduces much of this to single lines of code. But since you haven't started on it in class, then its probably best not to delve into it at this point.

If you look at other people's questions here about homework, the standard answer is "We don't do homework for you! You have to do it yourself. But, we will try to help you understand why your code doesn't work".

You've got some code here. That puts you ahead of many who post for homework help. But all we know at this point is that you have some sort of issue with it. You haven't said what that issue is. Does it fail to compile? Does it produce the wrong output? If so what's wrong with the output? Are you having an issue grasping the concept, or in this case a stage of the algorithm provided, and can't quite see how to proceed? What, exactly do you want assistance with?

k5054 25-Dec-21 9:27am View

And your question is?
Also, have you been introduced to the algorithm library yet? Much of the work for this exercise can be achieved with functions found there.

k5054 23-Dec-21 7:24am View

Yeah, you will. They sneak up on you!

k5054 23-Dec-21 7:05am View

I think the OP must some have other problem that he should examine. C++ will add a return 0 if no other return value from main is specified. If the OP's code is returning some other value, and main() does not have a return statement, then something is wrong. Unfortunately, the OP's code does not compile, even allowing for missing #includes. There are various variables that are accessed but not defined, and there's no closing brace for main(), so there's no knowing what the code might be doing in the missing sections.

k5054 22-Dec-21 21:16pm View

Your expected results got lost somewhere. Please update your question.

k5054 22-Dec-21 18:48pm View

You should probably investigate how to use the debugger. It will help you understand why either your compress or decompress routines are failing. I have not looked into your code too deeply, but an "off by one" error is a common mistake, even for experienced programmers!

In the previous reply, I should have pointed out that you might need to add 1 to the amount of memory you malloc() to allow for a NUL ('\0') byte at the end of the string. Not allowing for the fact that C strings have a NUL terminating char is one of the many sources of the 'off by one' bug.

k5054 22-Dec-21 18:42pm View

malloc and free are essential tools for the C programmer. You should google for some malloc tutorials if you don't know about them, yet.
Basically malloc returns a block of memory, suitably aligned for the storeage of any data. so a call to malloc(1024), returns a pointer to a block of memory of 1024 bytes. In your case you would do


char *Message = (char *)alloc(count);

and then later call free(Message)
I'm not sure why you think you need to multiply count by sizeof(long). count is the size of the file, so you don't need to multiply anything here.
In C, you don't need to cast the return value of malloc to the type of the object you are assigning it to. Whether you do or not is a matter of style, and/or coding standards. For C++ a cast is required, but for C++ you should use one of the C++ cast operators.

k5054 14-Dec-21 12:43pm View

Quote:My PC is almost too fast for this Some possible work arounds would be to get a pi-zero and peg its max CPU freq to 700MHZ ( = min). That's still probably way faster than your IoT device, but it might be slow enough to evaluate code speed. Alternatively, maybe a Qemu, or other emulator, would suffice. Emulating is generally slow, so might be able to assist. With an emulator, even if you're emulating a Pi, you can probably limit the amount of RAM so you can get some idea how things are affected in that area too.
As to a hash tables, have you looked at hcreate/hsearch/hdestory? The big downside to those is that there can only be one hash table at a time, but there is a GNU extension hcreate_r() etc. that allows you to specify a search table. You could, of course, look a the code in gnu-libc, but then you get LGPL issues, which might be a deal breaker.

k5054 10-Dec-21 23:26pm View

Its late and I goofed on typing the text. Ten lashes with a wet noodle for me, but it was enough to get you over the hump, I hope.

k5054 10-Dec-21 18:03pm View

Unless you've got a good reason not to, you should probably stick with strlen(), rather than writing your own - especially if you're dealing with long strings. Your system supplied strlen will probably make use of assembler instructions that will greatly improve performance. For example, on my home PC calling your get_string_length_from_pointer() on a 5MB string takes about 60 times as long as calling the system provided strlen(). That's without any optimization, but even turning on full optimization, strlen() still outperforms by a factor of about 9:1. The same will be true of other string functions in the string library.

k5054 8-Dec-21 11:33am View

Have you read (and understood) this: Template Instantiation (Using the GNU Compiler Collection (GCC))[^] I'm wondering if extern templates do what you need. If you are using G++-8 or earlier, there's also a -frepo option that might help. You could drill down from here GCC online documentation- GNU Project[^] to get to the closest G++ version to what you're using. All this supposes that the g++ provided by the device manufacturer has not omitted this from their toolchain.

k5054 7-Dec-21 17:37pm View

Oops, you're right (there probably should be, though hint hint). I guess I meant Voltage::get_Vout()

k5054 7-Dec-21 17:18pm View

I'll give you a clue to help with the debuging: If you put a break point at Voltage::voltage_divider_Calc(), what are the values of the voltage member variabes _R1, _R2, and _Vs? Why?

k5054 7-Dec-21 7:31am View

5ed. I wish I could give you more ;).

k5054 6-Dec-21 17:32pm View

Further reading the Common Function Attributes (Using the GNU Compiler Collection (GCC))[^] suggests that maybe the attribue "flatten" might get you closer to what you want. From the docs: Quote:flatten
Generally, inlining into a function is limited. For a function marked with this attribute, every call inside this function is inlined, if possible. Functions declared with attribute noinline and similar are not inlined. Whether the function itself is considered for inlining depends on its size and the current inlining parameters.

I read that as the flatten attribute applies to all function calls within the annotated function, not the function itself. So perhaps a combination of flatten for the caller and always_inlinefor the called functions gets you there. Or at least closer.

k5054 6-Dec-21 10:31am View

Because String hey = "Hello" is NOT an assignment. The compiler treats this as construction with an initial value. I.E. the compiler treats this as String hey{"Hello"}. Perhaps if you were to construct, then assign a non-array String using your original code, you would also get an error.

k5054 5-Dec-21 12:10pm View

What happens when you try to run the program? Does it crash? Does it print the wrong value? Does it print anything? Please let us know what issue you have and we might be able to help you resolve them.

k5054 5-Dec-21 11:41am View

You will need to provide the code for the String class, otherwise we cannot know. There might be anything in there that's messing up the array assignment. Does you assignment operator work for non array instances? As you know from your earlier question the code String a = "hello" does not call the operator=() method, but a constructor. So prove to yourself, at least, that assignment works for your String class. If basic assignment works, then come back here and post the code for your String class and maybe we can help you.

k5054 18-Nov-21 11:11am View

I was surprised at how many web resources make the same mistake. It took checking several links to find a correct one. That being said, I guess it should be pointed out that given modern hardware, the likelihood of realloc failing is quite small, particularly in a small program in 64-bit mode where your virtual memory is on the order of 256TB. But that doesn't excuse sloppy programming!

k5054 18-Nov-21 9:41am View

Except don't use that example, as it has a bug. In particular, if realloc() can't allocate new memory, it returns NULL, and does not change the contents of what was pointed to. So in the expression ptr = realloc(ptr, newsize), if no memory is available, then ptr gets set to NULL, and you've lost the handle to the memory you have allocated. Take a look at https://www.techonthenet.com/c_language/standard_library_functions/stdlib_h/realloc.php instead

k5054 25-Oct-21 22:46pm View

Addendum: Sometimes I think the Universe likes synchronicity. Posted earlier in Insider News : https://www.codeproject.com/Messages/5839439/Cplusplus-Smart-pointers-and-arrays

k5054 25-Oct-21 16:04pm View

Unless, of course the OP is using plain C rather than C++. In that case there's a whole lot of understanding to make sure that the non tread-safe function behaves correctly. It probably involves mutexes and/or condition variables to control access to the non-thread safe sections of the g() --- assuming the OP has access to g() and permission to modify the source.

k5054 21-Oct-21 21:05pm View

Have you considered std::any as your key value? That might work, but you might then have to box/unbox the values. If you do need to box/unbox, it might be better to create an ADT with a hash() member function - or perhaps even a template class.

k5054 19-Oct-21 18:15pm View

Checking gcc-11 and clang at the compiler explorer (https://godbolt.org/), the small string optimizations don't come into play - even at -O3 both compilers make a call to std::string.compare(). Just FYI.

k5054 14-Oct-21 11:28am View

What are the specs on the PC? Specifically what motherboard? Ubutuntu 16.04 was released in 2016, and went EOL in April of last year, so almost all PC's manufactured in that era should be supported. How are you installing? From DVD or USB stick? Can you boot a live DVD and see the SATA controller? e.g lspci | grep SATA should show list the controller. Once you know the SATA controller model, maybe a google search for that Ubuntu 16.04 might give useful results.

k5054 8-Oct-21 17:09pm View

You should probably use the "Have a question or comment" button below a solution to, rather than starting a new solution, if you have further questions about it. When you do that the person proposing the solution should get notification about it and know to respond. Otherwise they might not know that further information is needed.

If you only compile main.cpp, then you'll never get the code from human_function.cpp added to the program. As both Rick York and I have noted in Solution 3, it's not the right thing to do to #include "human_function.cpp" in human.h. If you want to compile in a single step, then use
g++ -Wall -Wextra -g main.cpp human_function.cpp -o main
this compiles and links in a single command line.
Does that answer your question?

k5054 8-Oct-21 16:48pm View

Further to what Rick has said, lets imagine you've implemented a stack, with the class header stack.h, and the implementation details in stack_impl.cpp. Now, if you have a program made up of 3 separate files, say main.cpp, automobile_stacks.cpp and aircraft_stacks.cpp, all of which include stack.h. With your solution, each compilation module will include a copy of the source code from stacks.cpp, and compile a copy of the code. This means that each object file (.o) will contain a definition of stack::push(), stack::pop() etc. Now when you try to link the 3 object files together the linker won't know which version of stack::push to link with, and will produce error listing complaining of multiple definitions of stack::push.

k5054 24-Sep-21 15:19pm View

What system are you getting the message "No child processes" on? The RPi or your PC? Perhaps add a call to strerror() to add to the error message so you know why your open() is failing.

k5054 20-Sep-21 10:52am View

Please post the linker error.

k5054 17-Sep-21 15:37pm View

It looks like std::uniform_int_distribution produces different results on windows and linux. Windows and linux both produce the same output for the following program:

#include<iostream>
#include<random>

int main()
{
    std::mt19937 rnd(1);

    for(int i = 0; i < 10; ++i)
        std::cout << rnd() << '\n';

}

I get the same results, namely:

Modifying the program to

#include<iostream>
#include<random>
#include <cstdint>

int main()
{
    std::mt19937 gen(1);
    std::uniform_int_distribution<int64_t> rnd(0, 100);

    for(int i = 0; i < 10; ++i)
        std::cout << "rand = " << rnd(gen) << '\n';

}

I get the following:

Linux   Windows
  42      23
  100     25
  72      37
  94      100
  0       100
  12      95
  30      72
  100     14
  14      49
  23      84

Unless I've done something wrong, here.

k5054 29-Jul-21 13:51pm View

Error code C2065 refers to an undeclared identifier. Have you declared and initialized s?

k5054 26-Jul-21 9:33am View

Your Listclass compiles fine. Your main function is incomplete. This suggests that the error you are seeing is within main<code>. As Richard has said, please fill in the details.

k5054 2-Jul-21 14:39pm View

Playing around with compiler explorer https://godbolt.org/ it turns out gcc, clang and MSVC are all capable of optimizing away the strlen call. Even better, as near as I can tell, MSVC manages to inline the strlen call.

k5054 2-Jul-21 14:20pm View

One probably shouldn't call strlen() for each character placed in the result. If one or two or both are long, think several MB, then the performance of this will probably be very poor --- assuming the compiler isn't able to optimize away the strlen calls. Better to create a const int and assign the return value of strlen to it for len_one and len_two, then use those tokens in place of the strlen() calls. You could also do a little math ahead of time and only use the for loop to the min(len_one, len_two), and then use strcat to copy the rest of the other string. I'll leave the actual implementation to the OP, if they're inclined

k5054 27-Jun-21 13:46pm View

Quote:it is always a bad idea to use void pointers because the compiler wont help you anymore. True enough, but sometimes with C, there's no other option. C does not have any concept of generics, so resorting to a void* is the best a C programmer can do. Even the standard library needs to rely on this for functions like qsort() and bsearc().

k5054 27-Jun-21 13:37pm View

Is some of your code missing? You say "it does not mark any error either", which I take to mean your code compiles cleanly, but the given code does not. In particular there is no declaration of variables i, x, N in scope for functions Columna and Renglon.

k5054 19-Jun-21 14:10pm View

Your script does not "go" anywhere. Maybe you mean you need to do the equivalent of

echo `echo system > /sys/firmware/zynqmp/shutdown_scope`

. If so, there a 2 options. Either use the system() call, or open,write,close the file programatically e.g.

// use system
system("echo system > /sys/firmware/zynqmp/shutdown_scope");

// open, write, close FILE
FILE *scope = fopen("/sys/firmware/zynqmp/shutdown_scope", "w");
fprintf(scope, "system\n");
fclose(scope);

Note that I have not done any error checking, so that should probably be something you should consider. If the system() call might produce an error or other feedback, take a look a pope().

k5054 15-Jun-21 15:16pm View

the statement std::list<unsigned short>* myList; declares the variable myList but it does not initialize it. Assume, for the purposes of discussion, that the compiler gives myList the value NULL when declared, unless assigned at declaration time. What you have then is the equivalent of

 (NULL)->push_back(val)

. You need to assign a value to myList, either through new (or a smart pointer), or by assigning the pointer to an already existing list of a conforming type. Just declaring a pointer variable, and then trying to use it without pointing at something first is a programming error.

k5054 24-May-21 18:20pm View

If you're not seeing anything in wireshark, that suggest that the problem might be on the other end. Do you know its completed the read successfully on the opposite side, and has send back the confirmation?

k5054 19-May-21 15:09pm View

When you have

char *s = "Hello"

the string compiler effectively uses

char *s = (char []){ 'H', 'e', 'l', 'l', 'o', '\0'}

Writing the string in quotation marks is really just a shorthand for the fully decorated array of characters. Knowing that, you can, if you so desire, declare a int pointer (or double, float, etc) exactly the same way, e.g.

int *x = (int []){ 0, 1, 2, 3 };

k5054 15-May-21 17:46pm View

That's a pointer declaration.

k5054 15-May-21 10:00am View

Take a look at functionThreee. The parameter passed in as delta is a double. The for loop uses a integer index i, but then adds the double delta to i. Since i is an integer, it can only have the values 0, ±1, ±2, ±3 etc, then adding an delta < 1.0 does not change the value of i.

k5054 12-May-21 16:59pm View

I guess it's all what you're used to. I've been doing K&R style since before the days of ANSI C standard. Like many, I learned C programming from their book. While I use the "hanging brace" style for inner control blocks, I have a strong aversion to hanging braces on the same line as the function definition. That's just wrong. There's a wikipedia page about indentation styles, for the interested: Indentation style - Wikipedia[^]

k5054 12-May-21 14:30pm View

To be fair, K&R also puts braces on the same line as an if, while, for, etc. but on a separate line for functions. And so does Stroustrup, so I think the OP is in good company with his brace style

k5054 2-May-21 12:24pm View

Also, you have a ORDER BY CLEARING_CYCLE that appears in the query, but not in either of the tables, nor as an alias to a column in the query. Please clarify.

k5054 2-May-21 12:01pm View

It would appear that IMG_PDC_OUTWARDCLEARING is "Table 1" in your data example, but I don't see where "Table 2" makes an appearance in your Query. Perhaps you could add actual table names to your data sample to help clear things up a bit?

k5054 1-May-21 13:10pm View

Perhaps this is a translation issue from OP or Consultants native language. Maybe a callback function is what the OP is after: https://en.wikipedia.org/wiki/Callback_(computer_programming)

k5054 1-May-21 12:06pm View

Or better yet, give a (short!) example.

k5054 1-May-21 12:03pm View

Lets try this again: I think this has to do with how scanf handles whitespace, in that given a format string "%d %d", scanf will consume any and all whitespace between integer values. So if your format string ends with whitespace, (tab, space, newline, etc), then scanf will try to consume all trailing newlines to satisfy the format string. The only way to proceed, then, is to enter a non whitespace character.

k5054 1-May-21 11:56am View

Deleted

I think it has to do with how scanf() handles white space. e.g.

you get the values 7 and 24 printed out, regardless of how much white-space (including new lines) you have in the buffer. So if your format string for scanf ends in any white space, then what scanf is doing is continually trying to consume the ending whitespace to satisfy the format string, which means that you can only proceed to the next input value by entering a non-whitespace character.  So using my example code above, if you enter "10
a" (or even "10a"), then scanf can proceed (or as in this case, terminate) when it finds the ''

k5054 1-May-21 11:11am View

Interesting. I've hit enter about 10 times after entering a number, and gdb says I'm still in scanf. eg:

[k5054@localhost]$ cat example.c
#include <stdio.h>
#include <stdlib.h>

int main()
{
    int n;
    printf("enter a number: ");
    scanf("%d\n", &n);
    printf("n = %d\n", n);

    return 0;
}
[k5054@localhost]$ gcc example.c -o example
[k5054@localhost]$ ./example.c
bash: ./example.c: Permission denied
[k5054@localhost]$ ./example
enter a number: 10





^C
[k5054@localhost]$ gdb example
GNU gdb (GDB) Fedora 10.1-4.fc33
Copyright (C) 2020 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from example...
(No debugging symbols found in example)
(gdb) r
Starting program: /home/ebacon/tmp/example 
enter a number: 10



^C
Program received signal SIGINT, Interrupt.
0x00007ffff7ec4442 in __GI___libc_read (fd=0, buf=0x4056b0, nbytes=1024) at ../sysdeps/unix/sysv/linux/read.c:26
26        return SYSCALL_CANCEL (read, fd, buf, nbytes);
(gdb) bt
#0  0x00007ffff7ec4442 in __GI___libc_read (fd=0, buf=0x4056b0, nbytes=1024) at ../sysdeps/unix/sysv/linux/read.c:26
#1  0x00007ffff7e55122 in _IO_new_file_underflow (fp=0x7ffff7f96800 <_IO_2_1_stdin_>) at libioP.h:948
#2  0x00007ffff7e563a6 in __GI__IO_default_uflow (fp=0x7ffff7f96800 <_IO_2_1_stdin_>) at libioP.h:948
#3  0x00007ffff7e2ed5b in __vfscanf_internal (s=<optimized out>, format=<optimized out>, argptr=argptr@entry=0x7fffffffd930, 
    mode_flags=mode_flags@entry=2) at vfscanf-internal.c:3023
#4  0x00007ffff7e2c712 in __isoc99_scanf (format=<optimized out>) at isoc99_scanf.c:30
#5  0x0000000000401163 in main ()
(gdb) quit
A debugging session is active.

        Inferior 1 [process 195134] will be killed.

Quit anyway? (y or n) y
[k5054@localhost]$ 
[k5054@localhost]$ lsb_release -ir
Distributor ID: Fedora
Release:        33
[k5054@localhost]$

k5054 1-May-21 10:21am View

Did you try this? For me, this just hangs in scanf.

k5054 29-Apr-21 13:01pm View

You need to call initscr() to initialize ncurses. Once you've done that getch() should work. But then you really should use the curses output routines to draw your playing surface, move the ball and move the paddles. Don't forget that cout is buffered, so it doesn't send output to the screen until either you write a newline ('\n') or use std::endl. In either case, a newline is sent to the screen, which could result in the screen rolling up. If you insist on using stdout, then you should probably use std::flush, instead, which sends output to the screen without adding a newline.

k5054 27-Apr-21 22:21pm View

I'm not sure why you're getting the error in songs.ccp at line 60, since it compiles fine for me, but the warning type qualifiers ignored on function return type is because you have const boolshowAll(). The compiler is telling you that the const qualifier for those functions is being ignored. An integer type is returned as a value, so is in a sense always const as you can't reach into the function and alter the value. The caller of the function gets a copy of the value, and is free to do what it will -- either compare against another value, or store it for later use. So if I had

const int f();
/* ... */
int x = f();

I have assigned the copy of the value returned from f() into variable x, and can now modify x as I please, and it can't affect the return value from f(). If I need x to be unmodified, I have to specify that on the caller side e.g. const int x = f().

k5054 27-Apr-21 17:45pm View

Hmm.. Maybe we need to define terms. I've always called () brackets, {} braces and [] square brackets. With that in mind, lets rewrite the code, as pasted, with a slightly different format:

ball->changeDirection(
                         (
                              edir
                         )
                         (
                             (
                                 rand() %3
                             ) +4
                         );

Clearly, there's a missing closing bracket ) for the call to changeDirection.

k5054 10-Apr-21 14:10pm View

C++ is a superset of C (mostly), so if you're happiest writing C style code, but compiling with C++, then do so - especially if its just for your own amusement/consumption. In a work related situation, you would probably be expected to follow company guidelines and coding conventions, but that's a different kettle of fish.
I think that there's boilerplate around here that says anything posted here is considered in the Public Domain. Even if its not, the code I posted here is pretty generic. I would expect any more-or-less experienced programmer to come up with something almost identical. And, in any case, this site is supposed to be about sharing ideas, so copy away!

k5054 10-Apr-21 12:39pm View

Seeing as you've tagged the question with C++, and used cout in your program, I hope I can be forgiven for assuming you were looking for C++ code :)

I don't actually do a lot of C++ myself - I learned C back when K&R was about the only reference there was. This was in the days of pre ANSI C so we didn't even have function prototypes, or void*. So most of my C++ tends to be C + STL, which is just fine, I think. ISTR that Stroustrup has said that C++ is a multi-paradigm language, and any way you want to use it, that's fine by him.

That being said, if you are going to be using the C++ compiler, you might as well add some of C++ goodness, like vectors and maps, ranged for loops, auto type deduction, etc. If nothing else, I think it produces somewhat easier to read code, and by pulling in the STL you've got a huge bag of tricks, written by really smart people, to rely on.

k5054 2-Mar-21 17:51pm View

What's in #include "TDU attamp_t.h"? If I comment that out, remove using namepace std and properly specify cin, cout, pair, etc as being in namespace std, it compiles fine.

k5054 22-Feb-21 17:33pm View

True, but what's the reverse of 1000? Using integer math you'd get back "1", which isn't really correct. If only there was a way to convert a number to a string, and reverse the string.

k5054 17-Feb-21 12:17pm View

Maybe you could explain a bit more about what you are trying to get at the end of this? For example given the following directory structure:

/home/TestC1/
|-- able/
|-- baker/
|   |-- alpha/
|   |-- beta/
|   `-- gamma/
`-- charlie/

What should the output be? Did you want just the last path element, the whole thing, anything after the base (/home/TestC1 in this case), or ???

k5054 16-Feb-21 13:00pm View

There is also a copy of the Window object created in the call to Room::AddWindow(). So by the object that vector::push_back() adds to the vector is a copy of the copy.

k5054 12-Feb-21 12:18pm View

You're welcome! Glad to be of assistance

k5054 10-Feb-21 11:04am View

Have you looked at the packet contents? Wireshark https://www.wireshark.org/ is my tool of choice. It should be available as an APT or DNF/YUM package if you are using linux, and is available for download for windows from their site. At least then you would know if your ISP is injecting something (naughty, naughty!), or if its a buffer overflow/reuse issue.

k5054 9-Feb-21 13:18pm View

Did you read everything I wrote? I suggested using

#pragma once

rather than include guards. I don't see where in your issue description you say you have a problem with that suggestion.

k5054 20-Jan-21 13:19pm View

You'll need to give us some more information, please. What OS and what SQL database engine are you using? Take a look at your database documentation, many databases now can produce JSON output, rather than a text table or CSV values.

k5054 17-Jan-21 16:02pm View

You only need to set SO_REUSEADDR on the listening socket. Even without SO_REUSEADDR, the port will become available again, after some amount of time. I think that's usually 3-5 minutes. If you're on Linux you can use

netstat -t

to show network connection status. A connection with status TIME_WAIT is a connection that has closed and is in the "timeout" section of the connection tear-down. Unless SO_REUSEADDR has been used on the connection, no new connections can be established on that port until the TIME_WAIT status has elapsed. See the man page for netstat for more useful information about network connections available from the command line

k5054 14-Jan-21 20:40pm View

Have you considered using rsyslog to send the logs to a central server. That way you would not have to visit various hosts to get the logs. Googling for "linux syslog remote servier" should get you various hits.

k5054 11-Jan-21 15:57pm View

In what way does this not work?
Why "Do not modify below this line" ??
I think it would be helpful if you did something like:

int maxIndex = lastMax(n, x);
printf("Max value: x[%d] = %d\n", maxIndex, x[maxIndex]);

But you should probably test that maxIndex is not -1 first.

k5054 10-Jan-21 13:41pm View

I think that should be

#include "../include/Calculator.hpp"

, for linux.

k5054 7-Jan-21 10:36am View

I could, but I think you would get more out of it if you tried to solve it yourself. Take a look at documentation for string ops (strcpy, strcmp, etc), and give it a try! If you still have problems, post another question.

k5054 19-Nov-20 18:02pm View

No, but it looks a lot simpler than mucking about with gmp. But there's a boost multi-precision class that might be a better option. One of the types wraps gmp bigints so you get a cleaner interface and trusted goodness of gmp and boost.

k5054 19-Nov-20 17:49pm View

Maybe try ttmath.org / Bignum C++ library[^]

k5054 13-Nov-20 9:32am View

works for me too ... Maybe the op actually has '\' followed by 't' to replace, rather than tabs?

Comments by k5054 (Top 200 by date)