Comments by merano99 (Top 200 by date)

+5, I come to the same conclusion here.

I had written it exactly the same way, right? " ...already done with the merge() function."

nothing to add, 5+

The alternative to a vector would be to wrap even a concatenated list in a class. But that would be unnecessary in C++ and not recommended. Rick's suggestion to use an array and reallocate memory each time would mean that everything would have to be copied multiple times. If you don't use a C++ container you have to take care of memory yourself and release it at the end.

+5

First of all, you can do the math: 40 x 40 x 4 bytes = 6.25 kByte
As long as the know-how for dynamic memory management is not enough you can (if necessary) leave it like this. As soon as the grids get bigger, you need more than one or recursion is desired it will not work well anymore.

The grid is defined by the decalation and also initialized by the assignment with empty. The code that assigns all elements with consecutive numbers should probably just be deleted. From my point of view there is no additional problem here.
If you want to keep the grid variable and also use very large grids it would be recommended to get memory from the heap with calloc(). However, this would be much more complex and requires some experience.

I had overlooked that the number of calls of the "operation" are counted, not the number of necessary swap operations.
If the "operation" is really only allowed to always replace each existing letter a in str1 with a (freely selectable) letter ß, one would have to call the operation several times and do additional work. A lookup table would also be helpful here. More runs would then be necessary when the problem occurs and additional characters would have to be managed in it.

I had noted errors that probably change the flow completely. It would be necessary to take these remarks into account in the code and adjust it. Where at the run tested by me still an error occurs I did not experience.
Instead of writing another new output here as a comment it would surely be much better to write an example output at the top of the question and explain exactly where something goes wrong, what the calculation base should be and what output is actually expected.
Since the source code is long and confusing, also because of the gotos, a further limitation in the source code would be helpful. Usually one writes subroutines and avoids gotos.

thanks!

str1="bacd"; str2="abcd"; // 3 changes ??
There are only 2 changes here. Introducing an additional letter does not seem to make sense and would significantly complicate the task. Would that really be part of the task? It would mean that no letters may be exchanged directly, which already occur. One would then have to introduce new letters, remember which places would still have to be changed in further runs.

The question is answered completely. How to write comparisons and assignments in C or C++ should also be easy to read in tutorials and books.

He didn't address the aspect that the application should be able to continuously output during execution and also receive on commands, did he?

As I said before, cross-process synchronization is not (easily) possible with pure C++. Since neither the operating system nor the compiler are known, several variants would have to be discussed here. It would be good if you could narrow down the framework.

+5. The messages disappear as soon as the functions are declared with "dllexport".

Thanks for the tip, it was right at the top - someone must have resurrected the part. I better delete the entry again.

Did you have it compiled as x64?

How does the program to be modified in b) differ from the first?

For which input does it not work and what is the result? Are you perhaps calling the subroutine incorrectly?

I had tested it, of course. What exactly is wrong?

The task is to add several lines up to the next blank line and assign the sum of the calories to one elf each. However, as Richard has already correctly pointed out, none of this happens in this program and is therefore not a valid solution.

Often there are several ways to solve tasks. A simple easy to understand solution would be Solution 4.

I would agree with you if the solution was complete. Without own contribution at the places where TODO is written it will not work. To complete the solution, you will have to understand the rest of the program. The multiplication is also completely missing here.

Thx

Why do you use Min/MaxInt here at all? You can simply set the start value to the first value and then omit it in the for loop. If you already use it, then from <limits.h>. You can never tell for sure how big an Int is or will be at some time.

The 8 should set the highest bit and that would give the largest negative value. If one assumes an int with 4 bytes, there should be a total of 8 hexadecimal digits. So here a 0 would be missing on the right side. If one already assumes that the numbers could be negative I would not output these also with %u.
The reason why the program of OriginalGriff had worked at all was that "accidentally" had used a very large number.

I'm sure you don't have to explain to CPallini what a cast does.

I cannot find the function RegCreateAddSetString(). Google is also at a loss.

You can use CreateProcess(). See update.

+5

Unfortunately, you have ignored all the solutions suggested here and come up with a solution that is even worse that the original code. Also, it is not desired that you answer your own question here. It would be better if you marked the new code as an update and attached it to the question above.
The code is not only wrong, but makes complete nonsense. Moreover, it cannot even be compiled.
After improving the incorrect brackets you would get the answer 1 if you enter e.g. 100.

OK. I guess I had forgotten about this version. Makes sense, of course. Have improved it.

You're right, according to the documentary, count is fine without -1. With std::string you would have to ensure the capacity and write it directly into the buffer. And then you would have to specify this length. Methods that could automatically increase the string are not used with getline(), in my opinion. I would not want to read multiple lines here.

Please do not post a solution to your question here yourself. Better would be to add to the question above already found solutions with appropriate reference.
Maybe it would be better to mark the helpful solutions here as a solution and give an appropriate rating. Then ask a follow-up question with new questions

I don't know what problem is meant by "the problem exists". I have extended my solution again.

+5

Please do not post questions as answers, but directly under the solution to which are questions. Alternatively, you could expand your original text above. See also Solution 3.

You never will see 21 at the end.

+5

my 5

my 5

Thx

+5

You propose yourself a solution? If yes, you should explain it, if no it belongs rather above under "What I have tried".

The function mygetline() reads individual characters until either the end of the line is reached or the end of the buffer. For this the loop for (int k = 0; (k <= len); ) with the index k is run through. If the line is finished and not empty it is terminated with break. A line that has not been read completely is then deleted up to the end of line character. If the line was read completely, 0 is returned otherwise 1. It makes sense to look at the whole thing line by line with a debugger.

I needed almost no special knowledge for this and I wonder how to comment "learned all this". For C, the book by Kerninghan & Ritchie (second edition) is enough to start with. However, it doesn't deal with the brand new features of C, but you can learn them later. In addition, you get experience by programming yourself and reading in forums ... I don't know the book "Absolute Beginner's Guide to C by Greg Perry", but if you can't do anything with loops and termination conditions after 2/3 of the book, I would find it questionable.

I can't use it with Visual-Studio and the doc says it's gone.
https://learn.microsoft.com/en-us/cpp/c-runtime-library/gets-getws
With problem I meant the not limitable length.

See Solution 4 for my version.

Unfortunately gets() has been removed in the meantime for a good reason. It has the same problems as scanf() with %s. there are alternatives like fgets() or better C++ and getline().

+5

You should not fill the buffer completely with fgets() and hope that strlen() finds an end of string. I would recommend to keep the last byte free for a 0.

Never use scanf with %s, because buffer overflows are not preventable.
My suggestion would be fgets, even if it is not a 100% replacement.
NOTE: The function gets() is no longer available in C or C++ (as of C11 & C++14)
Alternatives are discussed e.g. here
https://stackoverflow.com/questions/43174201/alternative-to-gets

For me the tags indicate what it should be or what solution is desired. If something completely different appears under "What I have tried:" I assume that the questioner has problems to implement it in the desired environment or is not even aware that his approach does not fit. With C and C++ there is however as well known a large intersection.
I am not sure if it would be acceptable to simply delete or ignore the possibly wrong tags. I would only do that if it is obviously wrong or incorrect.

+5, really inefficient way - exactly that

Firstly, the answer has nothing to do with question "Convert Python to C code", and secondly, the question is already 2 years old.

Visual Studio analyzes the string argument of printf very precisely, both the number and the variable type are checked. I am rather surprised that this did not always work with every compiler. At runtime, on the other hand, nothing is checked, then the bytes are displayed that are on the stack, or otherwise in the expected places. The output is then either random, or with %s it can also crash.

When calculating the days and derived quantities, it would probably be useful to consider leap years as well.

According to the table above, it should be 1/16. Other entries in the table are probably nonsense.

yes should be printed if it is NOT a prime ? With you it is currently exactly the other way around (which would also be usual). A recursion without return is also rather unusual.

yes, +5 (some code, like IPC commands like mutex; see my suggestion)

yes, +5

"3. When the value of the Analog signal increases above 1.6 volts on all the LEDs" ???
With incomplete sentences we would have to guess how to proceed.
So there is an unknown hardware here with a two-digit 7-segment display on which an average value of a data series is to be displayed?

If there is already program code, what is wrong with it? Unfortunately, you have neither described if there are problems nor asked any questions. So what do you need exactly?

+5

+5; iostream not needet, stdlib missing

Of course, an instruction like n[i]++ would work and increment the contents of the array at location i.

Why was the question tagged C? C# and C are two completely different languages and cannot be mixed.
If no solution with C (or C++) comes into question it would be good to remove the flag.
If you want to program a word search game instead of a Scrabble game, also change the title and add the text of the question with your answers of the questions to improve the chance of getting suitable answers.

Thanks for the advice. So sure, various limitations should be considered.
Multiple users would work - but just not well in parallel and I didn't assume "heavy work".
At least SQLite uses reader/writer locks and seems to contain synchronization for multiple processes and is (allegedly) threadsafe. For Windows more restrictions are documented, but usually multiple users don't really work in parallel on a Windows desktop system.
Restrictions are mentioned here:
https://www.sqlite.org/faq.html#q5

The questioner does not tell us whether multiple users actually work in parallel, whether multithreading is necessary, and whether multiple users should be able to write at all. I interpret in the given information that the database does not have to be networkable, at least it is not explicitly required.

If one actually has very high requirements, the effort will be accordingly high.

+5

It would be necessary that you write your own code under "What I have tried:". Feel free to write where you get stuck or if you can't find an error. Just not doing anything does not fulfill the purpose of your exercise. Especially the analysis and the search for a solution would help you in the future. Please click on "Improve question" and also add the previous code and your questions. The request to post a complete solution including explanation is not a question.

+5

https://en.wikipedia.org/wiki/Memory_management_unit

There is a source code approach, but no question?
You can already see if the number is even or odd, only the two functions that do the conversion are missing. It would also be nice if you could make the source code easier to read with the appropriate code tags.

Yes, no problem. See my solution 3

With which compiler and C++ standard should the shorthandversion work?
I get the error message that a function parameter must not have a type that contains "auto". But I would generally assume that you could declare the parameter as a class, which can then contain several data types and can also output it again via corresponding ostream overloads.
The template version works with

template<class T1, class T2> void printValues(T1 val1, T2 val2) ...

Have my 5!

I guess the question should be, if you could store a square as a variable to build a triangle out of it? In C++ you could do something like that with a class. You could then add the instances in a line and then output a complete line. Within the class you could manage the three lines in a std::string or std::vector. See my Solution 3 to implement it.

Based on Greg's response, I formatted the text to get a better view of the desired output.
The code shown does not match the desired result and an question ist still missing.

Thanks!

My 5. If you had taken the solution with the vector as you suggested yourself, neither check nor release would have been necessary.

Thank you too. Although it is not quite clear what the desired function of the OP's program is, I would guess that spaces are to be skipped instead of "abbreviating" the input, since he used continue instead of break. Unfortunately, the OP did not explain the meaning of the previously uninitialized variables p and q either. The original program contains parts that will probably never be used in this way. However, the headline then does not fit the program.

my 5

The basepointer contains the pointer to the only area requested with malloc(). There is only one malloc(), so a free() is also sufficient. The element pointer "item" contains a copy of basepointer and is then only offset. Therefore you must not free the element pointer separately as well.

My answer was the reaction to your "You also need to do the same for item". Here it would be just not good to do the same, but to use pointers directly instead of malloc.

Swapping the strings by copying the letters is not a good idea. It is better to swap only the pointers, then you don't need additional memory for item. See my proposed solution.

"you might want to make it bigger." That wouldn't be a good solution either. Much better would be to limit the input to the length, e.g. with fgets(). But otherwise good explanation 5+ for it.

First, this is not an answer to the question above, and second, the question was already from 2020.

+5

How did you measure the times and what are the results?

What should "if __name__=='__main__':" mean? Here only the pair of integers a, b is mentioned, what about b and c? The "following operations with a, b" are announced in the heading, but are completely missing in the question. The question would need to be revised as it makes no sense as it is. In addition, we expect a minimum amount of our own efforts here.

Oops, thanks for the hint. I guess it was too late to recognize the corpse. I'll leave it here anyway.

nice, +5

It may be that the questioner meant consonants, he asked for constants.

From my point of view, first of all the question should be formulated more clearly and there are some things to clarify.
What is meant by check operations on strings?
Should vowels and consonants be counted or captured?
What is considered a constant in a string ?
Where is the own attempt or exactly where the conversion fails?

Thank you

An alternative result output would be nice.

if (items > 0)
  {
  ...    
  }
else
  puts("no input");

The requirement states: "The data is output in the order of the first occurrence of the measurement result.. ". Additional effort in data management and output would probably be necessary for a set.

Wouldn't a std::set sort what was explicitly not wanted here?

Thanks!

Improving the imprecision of float simply by replacing it with double is often not enough. Mathematical transformations of trigonometric functions, as well as the "correct" order when calculating often bring improvement. It is also better to multiply first and only then to divide.

If you don't pay attention to the details when learning, you probably won't later. I would hope that the questioner would not use a while loop to copy a string in the real world, but perhaps rather strcpy_s(). But sure, if it's only enough to get some code running with effort, you have to compromise unfortunately .

It would be good to query the limit of a[] as well.

Are you serious? i < 13; It would make more sense to check if pin points to 0 and index i is inside sizeof(out), right?

This solution is still missing the detail that the same digits should be sorted.
Note: The solution is relatively limited and works only if both numbers have exactly 2 digits and are separated by exactly one space. At least this would be just enough for the given example. Also additional characters like letters instead of numbers would be processed unnoticed, was that intentional?
Since C11, gets() is replaced by gets_s(). The gets() function does not perform bounds checking. Some newer Compiler dont support it. Here if your input is to long your Program will crash.

Besides: A question that has been here since 2018 should hardly interest the questioner.

Loadimage is based on GDI+ and that supports many formats such as: BMP, GIF, JPG, PNG, TIFF, WMF, EMF, ICON.
The description "simple 32 bit bitmap" had me wondering a bit. All the usual "simple" 24 bit data formats work. The requirement that transparency is also required was not in the description.
Otherwise LoadResource() was also the solution. my 5!

You write "environment is windows", but you tagged Linux an DOS?

ups. was not saved. now done.

I have updated my solution to give you a clue. However, you should really take care of solving it yourself. Usually we only give hints and not a ready solution. Maybe you lack the necessary knowledge to understand the data structure "Linked list". In this case do your own research e.g. here: https://en.wikipedia.org/wiki/Linked_list It is on you to solve your task.

+5

+5

Thank you for pointing this out. I will use a commented out TODO in the future. Of course, this will not eliminate the problem of the expected own contribution by itself.

You can use Rick's suggestion to write subroutines like AddSection() and AddCage(). And of course you call the functions multiple times, for each section and for each cage.

What exactly cannot be compiled?
As usual, the associated includes have to be added and also the AddSection() and AddCage() functions have to be added by yourself, of course. The proposal as a concept should be an essential part of the solution to the problem. The completed source code was compiled without errors and warnings.

My 5

If one assumes that there is a sorted code table and no umlauts or the like are to be counted, this works.
But there are also functions in C to check letters. To use gets() and str safely with strlen(), I would initialize them with zeros like arr.
Also, it would be highly recommended to use gets_s() or fgets() instead of gets() to make sure there is no overflow. (The current version of Visual Studio doesn't seem to know gets() at all ?). Because of int main(), a return would also be formally expected.
My 5.

Wouldn't it be much better to use the fgets() function instead of scanf("%s",..)?

You write "thankyou! got it" but so far you have neither given stars nor marked posts as a solution. It would be nice if you still do that. Or what would you still be missing?

correct, +5

Thanks for the tip. The code was also somehow familiar to me.
There are often several ways to get a valid solution. Here there seem to be difficulties with the implementation. As usual, OriginalGriff has already posted the reference to "Testing and Debugging".

Yesterday you uploaded source code as a solution. I have already given you some comments on this. The suggestion was that you put the source text, preferably already revised, above under the heading "What I have tried:".
Is the source code currently being revised, or has the question been resolved? After all, the question was marked as solved, which doesn't seem to be the case.

Posting your faulty code attempt here as a solution makes no sense at all. Please delete this and copy it to the question above.
<iostream.h> is not supported by current C++ compilers. The program also has many bracket problems and can never work like this. Instead of using a Stack class, it would make much more sense not to write the function in the class and call it recursively instead. That would also be a sensible solution for several bracket levels.

Wrong Output? In this case, there are still very many errors in it, which make the process almost impossible. But it would also be good if you provide data documenting the errors and also write what you would have expected as outputs.

Thanks!

Someone woke the corpse. Thanks for the hint.
I'll delete it again because it makes so little sense.

The problem seems to be that he does not understand the code he copied from my answer. It really would have been better if I had left out the mistakes and just commented on what didn't fit. Unfortunately, he didn't correct anything himself, not even the typos.

He copied my answer to a question and the comments mark where it was wrong before. Apparently he copied without understanding it.

First you would have to decide which programming language it should be. Please remove the tags that do not apply.

The question might be, "How can I sort a date of the form DD-MMM-YYYY?" In fact, other questions would also be possible.

Ok. As always, there are many ways to solve the task. Your suggestion should help. You have my 5 for it.

If the plan is to recode the date into a string of the form "YYYYMMDD" then what is the point of the structure?

The number of days in the month could be used for checking and would therefore be useful. The number for the month, on the other hand, would be superfluous, since it can be derived from the index.
But with the sorting suggestion above, the structure is not needed anyway.

"And what are you even talking about with the rest of the commentary?"
Here seems to be the problem. We make suggestions that don't seem to be understood and may therefore not be used.
"Nobody ever said anything about debuggers in the previous questions."
Yes, that's right. I usually assume that it is known that there are debuggers. I don't feel good about pointing this out, but if it helped in your case, that's good.
I suggest writing functions for reading in the password, the random-password and the username.
This makes troubleshooting easier and you can only ask for this function. At the moment the code doesn't seem to be very good in so many places. When we get such a big morsel in front of us, the list goes on.
I had already suggested a function for your password problem as an example. If you would take over that and would also take over the know-how for other parts, we would get further. Unfortunately, you're stuck in places like the arr[] , which wouldn't be a good solution to your problem.
My suggestion was to read a complete line with getline(), why don't you just do that?
Don't get frustrated, but please also make sure that we don't get frustrated when the old problem comes up again as a new question.

Exactly that, and the suggestion to use switch-case instead of the countless if queries. +5

oh, wrong place, thanks for hint

Thanks!

You can check each letter in a string individually. Consider using find_first_not_of() if you have a whitelist of letters you allow. You can convert characters with toupper() instead of nagging them.

It's valid, but nonsensical. It's never going to go that flat. Data type string with checking of characters in a loop would be one of the few possibilities.

With the data type char and cin as input there is no way because the space as a separator is not read.

Nothing for the compiler, it's just text that was missed when translating, it means 'or'. I replace it.

If you process the text as a char, then what is "the first"? It would be better to read in a string and then check the individual characters in it.

I suggest you check if our suggestions improve your solution. If there are new questions, it is best to open a new question.

Something else. Thanks for marking my text as solution. It would be nice if you also fill out the rating with stars for our posts. A solution without stars looks strange.

Neither the rules were explained nor is it clear how to call solve(), because the function main() is missing. It is not reasonable that we should find out the missing parts ourselves for testing.

Checked the Code with VS. Each calculation has the expected result. Which development environment (compiler, operating system incl. version) produce the result?

It's not an infinite loop, since already after 100000 runs the job is done ;-)
I also find it strange that names always have to be 10 characters long and spaces are completely ignored.

I'm glad that my comment finally led to the solution. As suggested I create a solution.

In fact, there are many ways to read a byte from a file under Windows with C or C++. Actually, they should all work if the file is readable and not empty. Please list what did not work and what the error was.

A table-based approach for a signal that consists of two straight lines, i.e. is rather easy to calculate, would probably be a waste of space. The author probably forgot to indicate that he is working on a microcontroller. In general, however, this is really a high-performance approach for complex signals.

I already thought that you are on a microcontroller. In hardware the problem would be easier to solve. An operational amplifier and a capacitor give a clean (analog) solution ...

"I have a sine wave and a square wave at that frequency." Do you know the frequency?
If we assume that it is easier to generate a triangle signal from a symmetrical square wave signal, you would have to sample the signal with a fairly exact timing. You need a lot of information about the signal to be sampled. Latencies and other effects occur as a result of the sampling and the calculation.

The two dimensions are clear, but the vector's values ​​are probably copied every time, aren't they?
Just tested with VS. The end result is the same if you don't keep updating the vector rowvec in the loop. In your version, however, many new classes are created and destroyed.

My condolences that you have to use an outdated compiler. It may be that my suggestions will also work with your C++-like compiler, but some things may also not work.

Why is the creation of rowvec in the loop?

Either he can't or he doesn't understand what we're talking about. The fact that we have already received several almost identical, almost unchanged problem programs is not a good sign.

thanks!

thanks ;-)

added some more error places

Yes, unfortunately, but the misery must come to an end at some point.

For passing a single int the best solution. +5

It could keep crashing if value1 and value4 should be on the stack. Maybe the simple cast would be safer here.

You forgot an l at the end of the first link (https://www.emmet-gray.com/Articles/ProjectConverter.html)

That's exactly what I meant when I suggested classes. Other classes in a hierarchy would make sense, such as deck, player, and game.The original question what's wrong with this code, see my solution below. (+5 for you)

In C, VLAs have been allowed for a long time. However, they are not without controversy, since one has no control over what happens when memory is not available. The questioner has not yet said whether he uses Visual-Studio, nor has he given a C standard. Whether VLAs are a possible cause of errors here is not clear.

Microsoft states that Visual-Studio supports C99 with a few exceptions, but this is not true, for example, in this point. see: docs.microsoft.com

GCC had VLA as an extension before C99, one that also extends into its C++ dialect.
In C++ it is expected to use a vector in these cases anyway.

The topic is also discussed here: Link

The XListCtrl seems not to be compatible to newer VS-Versions. Even after the usual adjustments, I couldn't get it to work. It comes with a lot of dependencies that looks like a lot of effort.

Glad I could help. You are welcome.

You could improve the program if you used a std::array and called std::sort(). In terms of C++, variables and functions could also be packed in a class.

+5

Since the formulation and translation of my answer took some time, you were faster. (Congratulation)
My suggestion (see below) would be to move the variables into a class instead of having to constantly pass them in. This would also help for other functions affected by the same problem, e.g. calcMember().
(+5)

nice, my 5!

As an example of image processing, the loaded image is converted to gray and saved to disk.The example shows how to iterate over image data and how to access and change the pixel values.
https://solarianprogrammer.com/2019/06/10/c-programming-reading-writing-images-stb_image-libraries/#basic_usage

Since the author has revised his question, he seems to be still looking for an answer.

Since this is your first answer here, I would like to ask you if you think someone else is interested in a question from Nov 2020.

please vote and try again

Nice, +5!
With \x31 there will be a lot lost, Special characters that you do not want in texts end with \x1F (or \037)

You said it. Unfortunately noticed several times, and also resistant to advice. He doesn't get any more answers from me. Maybe answers about uninitialized pointers are too hard.

Yes, partially correct. Since I don't use Code::Blocks I can't say exactly what should be there. solution updated.

Note: I would advise against using multibyte strings.

Yes, possibly. I don't know why anyone would have to change more than 100 functions either:-D

Thanks

Yes, auto should always match since the compiler sticks to the declaration.

You're right when you say that error handling is still completely missing. Even system calls should check by checking the return values. For the sake of clarity, the test in forums is often initially omitted. When debugging, however, they often do a good job. (5+)

For system calls like Write File(), I recommend using exactly the variable types that are intended for here. A C++ data type does not fit here!

Comments cannot be rated here. My solution is currently the only one. If the error with the ampersand wasn't a careless mistake, I would recommend improving C or C++ a bit.

No, it could have been coded like that in the early 80s. I gave Richard's dynamic solution a 5, but I also wanted to offer a simple alternative here with only one malloc.

The Task was to store names in one dimensional array. See Title.

I always find it a pity when no solution is found for questions. Basler cameras, for example, are not so rare that someone will not need the solution at some point.
That a valid solution could lead to a ban would surprise me.

If the Answer does solve the problem vote and accept as solution.

You cant do this easily, i already told you:

L"\xef\xbb\xbf - hello - J - こんにちは - abcdefghijklmnopqrstuvwxyz",

Unfortunately, the result at i != 0 is wrong and thus does not have the same effect at all.
If you look at the generated code, there is no question of saving, on the contrary.

A look at the generated assembler code shows that the creative coding rather means more effort and is also more difficult to understand for humans.

for (int i = S.size() - 1; i >= 0; i--) {
 
00A17B45 83 E8 01             sub         eax,1  
00A17B48 89 45 D4             mov         dword ptr [ebp-2Ch],eax  
00A17B4B 83 7D D4 00          cmp         dword ptr [ebp-2Ch],0  
00A17B4F 0F 8C A2 00 00 00    jl          $LN13+2Bh (0A17BF7h)

for (int i = S.size() - 1; ~i; i--) {

00A07B45 83 E8 01             sub         eax,1  
00A07B48 89 45 D4             mov         dword ptr [ebp-2Ch],eax  
00A07B4B 8B 45 D4             mov         eax,dword ptr [ebp-2Ch]  
00A07B4E F7 D0                not         eax  
00A07B50 85 C0                test        eax,eax  
00A07B52 0F 84 A2 00 00 00    je          $LN13+2Bh (0A07BFAh)

(not optimized code, as you can see)

I Think converting BOM48 to wstring will not work at all.

   HANDLE hDFile = CreateFile(flename, GENERIC_READ | GENERIC_WRITE, 0, NULL, CREATE_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL);
    unsigned char BOM48[3]{ 0xef, 0xbb, 0xbf };
    bErrGHJ = WriteFile(hDFile, (LPCVOID)BOM48, (DWORD)sizeof(BOM48), &dwBytesWrittenXPP, NULL);

    do {
        wstring write_stuff77 = L"XYZ";

        // Convert to two LPCVOID.
        // LPCVOID Nf2 = (LPCVOID)&write_stuff77[0];  NO!!
        LPCVOID LpAll = (LPCVOID)write_stuff77.c_str();

        // wstring wstrFirst(static_cast<const wchar_t*>(Nf1)); !!JUST DONT DO THIS!

        Bytes_that_need_To_Write = write_stuff77.size() * sizeof(write_stuff77[0]);

        bErrGHJ = WriteFile(hDFile, LpAll, Bytes_that_need_To_Write, &dwBytesWrittenXPP, NULL);

        bytes_which_have_written_already += dwBytesWrittenXPP;

    } while (bytes_which_have_written_already < Bytes_that_need_To_Write);

    bErrGHJ = CloseHandle(hDFile);

Thanks for the hint. With some programs you can switch on the writing of the BOM under options. In fact, e.g. Chinese characters are sometimes still not displayed correctly. Notepad++ and Firefox understand it and even my Visual Studio has no problem with the UTF-8 BOM.

I'm glad that there is now a solution as requested. Congratulations on the success!

Be careful when using C++ Datatypes with System-Calls.
//BOOL bErr = WriteFile(hFile, (LPCVOID)&write_stuff[0], sizeof(wchar_t)*write_stuff.size(), &dwBytesWritten, nullptr);
BOOL bErr = WriteFile(hFile, (LPCVOID)write_stuff.c_str(), write_stuff.size()*sizeof(TCHAR), &dwBytesWritten, NULL);


I didn't take care of the file format myself. When I got the special character pasted by copy and paste, the Visual Studio code editor changed the file format to include special characters can be saved.
It's not a good idea to edit code with a text editor like Wordpad, though the code that initially looks correct can contain invisible special characters that subsequently disrupt the compiler.
I edit source texts preferably with the associated development environment (IDE) or with Textpad or Notepad++. Textpad notices that special characters are included and offers to "convert" them, the special characters are then either gone or changed. Notepad++ seems to interpret and preserve the special characters correctly.
Windows tries to associate the file format with the extension. If you have an unfavorable or wrong link here, there will be problems. The problem arises here because a source file does not have a header in which the format is clearly defined. In contrast to graphics files, an editor then has to guess the format or rely on the extension. I've been wondering if there's anything wrong with that. It just grew that way historically and is no longer easy to change.
I took a closer look at the file format chosen by my IDE for the demo.cpp. The source text with special characters was encoded in UTF-8.

Please rate the (both?) solutions mark as solution. For further questions better open a new topic.

Thanks, I had overlooked this time.

updated my solution

In a wstring object, the string is stored as wchar_t*, so you can assign it, or specify it directly at creation.
wchar_t* test1 = TEXT("Test");
wstring test2(test1);
Which compiler and version do you use exactly?