What could be causing my program to hang on cin.getline() in C++ code?

[yungman]

I have no idea why the program got hung on cin.getline() in my program. cin >> works just find. Here is the program. It is the really scaled down version of my long program. I just deleted all the irrelevant things. The structure is defined in the header file.

I use debug to step through the program. It just got stuck on any of the cin.getline(). I don't even know where the program go because it just got stuck on that line. It's going nowhere. debug doesn't even help because it just stuck on the cin.getline().

This is the Source.cpp:

#include <iostream>
#include <cstring>
#include <vector>
#include <iomanip>
#include "fileVector.h"
using namespace std;
fileVector fileV;

int main()
{
  int nameLen = 25, phoneLen = 16, addLen = 41, eAddLen = 35;
  fileVector::Directory Temp2;
  cout << " Enter last name:  ";
  //cin.getline(Temp2.lastName, nameLen);
  cin >> Temp2.lastName;
  cin.ignore();
  cout << " Enter first name:  ";
  cin.getline(Temp2.firstName, nameLen); cin.ignore();
  cout << " Enter street number and street: ";
  cin.getline(Temp2.address1, addLen);cin.ignore();
  cout << " Enter city, state and zip: ";
  cin.getline(Temp2.address2, addLen); cin.ignore(256);
  cout << " Enter phone number: ";
  cin.getline(Temp2.phone, phoneLen); cin.ignore(256);
  cout << " Enter email address: ";
  cin.getline(Temp2.emailAdd, eAddLen); cin.ignore();
  return 0;
}

Here is the fileVector.h:

#ifndef fileVector_H
#define fileVector_H
#include <vector>
#include <fstream>
#include <iomanip>
class fileVector
{
private:
    const static int nameLen = 25, phoneLen = 16, addLen = 41, eAddLen = 35;
public:
    struct Directory
    {
        char lastName[nameLen], firstName[nameLen];
        char address1[addLen], address2[addLen];
        char emailAdd[eAddLen];
        char phone[phoneLen]; int element;
    };
};
#endif

Please help. Thanks

 

[jtbell]

What are all those 'cin.ignore()'s for? What happens if they're not there?

 

[yungman]

Thanks you so much.

I thought I have to put cin.ignore() before any cin.getline()! Or else it will read the "Enter" and skip the next read from reading the last "Enter". So I put cin.ignore() before EVERY cin.getline()!
Why the cin.ignore() make the next cin.getline() hung? I have no way to debug myself as it just got hung on that line, that's the end of the debug for me!

Do I just put the cin.ignore() ONLY if the last cin is streaming eg. cin >> variable?

Thanks for the help.

 

[Mark44]

I thought I have to put cin.ignore() before any cin.getline()!

You really need to look at the documentation of the ignore function.
This function takes two arguments, both with default values. If you call ignore with no arguments, as in cin.ignore();, what character, if encountered, causes ignore to return?

 

[yungman]

You really need to look at the documentation of the ignore function.
This function takes two arguments, both with default values. If you call ignore with no arguments, as in cin.ignore();, what character, if encountered, causes ignore to return?

My understanding is it will ignore the next character in the buffer. Everytime you enter a word, you hit enter, that '\n' is stored in the buffer. The next getline() will read the '\n' and be done and ignore the rest.

That's the reason I put cin.ignore BEFORE the next getline() to clear the buffer. Why does that cause the next getline() to get stuck?

I read on line, they all talked about how many character to ignore and what to ignore. I tried cin.ignore(256, '\n') and cin.ignore(1, '\n'). Still got stuck.

Point is that it should NOT cause the NEXT getline() to get stuck by clearing the buffer before the next input.

Thanks

 

[Mark44]

My understanding is it will ignore the next character in the buffer.

No, that isn't what it does, and is why I said you should read the documentation. See https://docs.microsoft.com/en-us/cpp/standard-library/basic-istream-class?view=msvc-160#ignore.

 

[yungman]

No, that isn't what it does, and is why I said you should read the documentation. See https://docs.microsoft.com/en-us/cpp/standard-library/basic-istream-class?view=msvc-160#ignore.

It's a long document that ONLY have few lines on ignore(). Nothing I learn more than from other online site.

Just the first parameter give how many to ignore, I tried 256 and I tried 1, both hung on the immediate ignore().

I have to correct myself, it hung on the cin.ignore(). But still why?

Bottom line, it should NOT be stuck on that. AND it sure none of the document say anything about getting stuck if not using it right. If you see anything on ignore() getting stuck if it is not used correctly, please give me the link. I spent quite a bit of time online already, NOT that I am too lazy to read and search.

Thanks

 

[Dr Transport]

https://www.dreamincode.net/forums/topic/238918-what-will-cinignore;-do-without-any-arguments/

do some investigation on your own, it took me exactly 3 seconds to find out what happens when you pass no arguments into cin.ignore. You are not using it correctly.

 

[yungman]

https://www.dreamincode.net/forums/topic/238918-what-will-cinignore;-do-without-any-arguments/

do some investigation on your own, it took me exactly 3 seconds to find out what happens when you pass no arguments into cin.ignore. You are not using it correctly.

I know that, it just dump the next character, and that's exactly what I want. I know and I read many articles like this on line already. If there is no character, then there's nothing to dump. It's just an insurance to put it there.

My question is why does it get stuck. If ignore() will get stuck if the buffer is empty, shouldn't anyone mention about this in all the documents online? If you see any warning on this, please let me know.

Thanks

 

[Vanadium 50]

I tried 256 and I tried 1, both hung on the immediate ignore().

And why do you think the problem can be solved by guessing what number to put in?

People are telling you that you are not using it correctly. You are claiming that you are. And yet your code doesn't work. So who is right?

 

[yungman]

And why do you think the problem can be solved by guessing what number to put in?

People are telling you that you are not using it correctly. You are claiming that you are. And yet your code doesn't work. So who is right?

No, why is it not correct? Is there any document saying that if you put cin.ignore() after cin.getline(), it will get stuck and why?

I know how to fix it, just the question why and why I have not found anything about it will get stuck. Each and every single article talked about the arguments, how many characters to ignore. I even remember reading something when in doubt, make the number larger to guaranty emptying the buffer.

 

[yungman]

Guys, it's not that I don't take your words. Obviously don't follow cin.getline() with cin.ignore(). I already put into my notes. The question is how come I cannot find anything about it get stuck on line? What is the reason?

There is no way to find out using debug, Stepped through that already, it just stepped into cin.ignore() and that's it, not coming back, nothing to debug, just stuck. That's dangerous! But every article I read is about what is the meaning of the first and second argument only.

 

[Vanadium 50]

Is there any document saying that if you put cin.ignore() after cin.getline(), it will get stuck and why?

Is there any document that will tell you that the exact code y = x/(i-2) will fail if i = 2?

Your first mistake was ignoring what people said when your first discovered cin.ignore.

Your second mistake was sprinkling cin.ignore in random places hoping that it would make things work better.

Your third mistake was assuming the problem was with getline.

Your fourth mistake was to post your entire code, rather than just the code needed to reproduce the problem. This is lazy, inconsiderate, inexcusable, and had you done that you might have actually solved the problem yourself.

Here is the code you should have posted.

#include <iostream>
int main() {
  char name[26];
  cout << " Enter name:  ";
  cin.getline(name,25);
  cin.ignore();
}

Note that it demonstrates the problem in 7 lines and not 47.

Furthermore, if you put a "cout << name" in between lines 5 and 6 you would have realized that the problem was not with getline. That would have saved you from making mistake three.

Final mistake: you failed to ask yourselves two questions:

• What do I intend cin.ignore to do? (And "fix my code" is not an answer.)
• What does it actually do?

This is not programming. This is Harry Potter - hoping the right incantation will produce the right result.

 

[yungman]

Is there any document that will tell you that the exact code y = x/(i-2) will fail if i = 2?

Your first mistake was ignoring what people said when your first discovered cin.ignore.

Your second mistake was sprinkling con.ignore in random places hoping that it would make things work better.

Your third mistake was assuming the problem was with getline.

Your fourth mistake was to post your entire code, rather than just the code needed to reproduce the problem. This is lazy, inconsiderate, inexcusable, and had you done that you might have actually solved the problem yourself.

Here is the code you should have posted.

#include <iostream>
int main() {
  char name[26];
  cout << " Enter name:  ";
  cin.getline(name,25);
  cin.ignore();
}

Note that it demonstrates the problem in 7 lines and not 47.

Furthermore, if you put a "cout << name" in between lines 5 and 6 you would have realized that the problem was not with getline. That would have saved you from making mistake three.

Final mistake: you failed to ask yourselves two questions:

• What do I intend cin.ignore to do? (And "fix my code" is not an answer.)
• What does it actually do?

This is not programming. This is Harry Potter - hoping the right incantation will produce the right result.

You manage to make all the insults without answering the question. You feel better? Physics Forum condone this?

 

[Jarvis323]

I have no idea why the program got hung on cin.getline() in my program. cin >> works just find. Here is the program. It is the really scaled down version of my long program. I just deleted all the irrelevant things. The structure is defined in the header file.

I use debug to step through the program. It just got stuck on any of the cin.getline(). I don't even know where the program go because it just got stuck on that line. It's going nowhere. debug doesn't even help because it just stuck on the cin.getline().

This is the Source.cpp:

#include <iostream>
#include <cstring>
#include <vector>
#include <iomanip>
#include "fileVector.h"
using namespace std;
fileVector fileV;

int main()
{
  int nameLen = 25, phoneLen = 16, addLen = 41, eAddLen = 35;
  fileVector::Directory Temp2;
  cout << " Enter last name:  ";
  //cin.getline(Temp2.lastName, nameLen);
  cin >> Temp2.lastName;
  cin.ignore();
  cout << " Enter first name:  ";
  cin.getline(Temp2.firstName, nameLen); cin.ignore();
  cout << " Enter street number and street: ";
  cin.getline(Temp2.address1, addLen);cin.ignore();
  cout << " Enter city, state and zip: ";
  cin.getline(Temp2.address2, addLen); cin.ignore(256);
  cout << " Enter phone number: ";
  cin.getline(Temp2.phone, phoneLen); cin.ignore(256);
  cout << " Enter email address: ";
  cin.getline(Temp2.emailAdd, eAddLen); cin.ignore();
  return 0;
}

It didn't get stuck when I ran it, but this is the output.

Enter last name:  Lisa
Enter first name:  Mona
// had to press enter a second time here
Enter street number and street: NA             
// had to press enter again here
Enter city, state and zip: NA
<256 newlines here> //pressed enter 256 times
Enter phone number: NA
<256 newlines here> // pressed enter 256 times
Enter email address: NA

Here is a stack overflow answer that might help.
https://stackoverflow.com/questions/9349575/cin-ignore-is-not-working

 

[Vanadium 50]

Telling you that you are making mistakes is not the same as being insulting.

 

[yungman]

1)That was a way scaled downed program to the last few absolutely relevant lines.

2)Of cause I know I need cin.ignore() after cin >> name. That's common sense.

3)I am not sprinkling cin.ignore() everywhere, just after each cin.getline().

4) I read about when in doubt, put cin.ignore(256, '\n') to make sure to empty everything from the buffer. I know cin >> name only leave a single '\n' in the buffer, what's wrong putting cin.ignore() to get rid of one character.

5) Lastly, OF CAUSE

It didn't hang when I ran it, but this is the output.

Enter last name:  Lisa
Enter first name:  Mona
// had to press enter a second time here
Enter street number and street: NA              
// had to press enter again here
Enter city, state and zip: NA
<256 newlines here> // had to press enter 256 times here
Enter phone number: NA
<256 newlines here> // had to press enter 256 times here
Enter email address: NA

Thank you so much. You are the first one that actually answer the question. I did not try hitting enter again to make it unstuck.

After I see you post, I actually went back to play with my program, if I put cin.ignore(3), I have to hit Enter 3 times to get out of it. I guess it has to ignore 3 times to get pass that.

It is funny, I read an article that when in doubt, put cin.ignore(256, '\n') to make sure I get pass everything in the buffer. I know cin.ignore() without anything will ignore once. I used cin.ignore(256) before in my program and I never ran into problem. I wonder why.

But I am surprised none of the article mention this. Believe me, I went through quite a few articles today, every single one talked about the argument tells how many to ignore, nothing about if you put too many, you have to hit so many times to get pass the ignore().

That's the essence of why I am asking in this thread. You know I got pass this one two nights ago and finished the program! I am not that stupid to insist on using this. I even put it in my notes DO NOT DO THIS unless you have a cin>>name before that.

Thanks so much, you come to the rescue again. That's all I am looking for, a reason why and you show me. Also, thanks so much in the other thread helping me to finish the big program.

Alan

 

[Jarvis323]

You could also input 255 characters and then press enter once instead of enter 256 times.

 

[Jarvis323]

You are the first one that actually answer the question.

If that is the case, are you still sure it's to your benefit. Others have tried to steer you in the right direction, and give tips on how to figure it out yourself. Teach a person to fish, and they won't go hungry type of thing.

 

[Vanadium 50]

Of cause I know I need cin.ignore() after cin >> name. That's common sense.

And it's wrong. And there was a truly excellent post by @pbuk explaining why and another excellent post by @jtbell explaining the proper way to do this.

 

[yungman]

If that is the case, are you still sure it's to your benefit. Others have tried to steer you in the right direction, and give tips on how to figure it out yourself. Teach a person to fish, and they won't go hungry type of thing.

I used cin.ignore(256) before and did not give me problem, I don't know where the program is anymore. That's the reason it never cross my mind that I have to count how many times to hit enter. I swear I read somewhere they put 256 for the hell of it and it worked for me before.

Now I know I have to count how many character I have to ignore. Like after cin>>, I need to ignore one character. cin.getline() or cin.get() do not leave anything behind, so don't use it.

Thanks

 

[yungman]

And it's wrong. And there was a truly excellent post by @pbuk explaining why and another excellent post by @jtbell explaining the proper way to do this.

How is it wrong? I read about using cin.ignore() is a must after a cin >>.

jtbell asked me about why put all those in in post 2, that confirmed to me what I did yesterday removing them and got my program working and posted it yesterday in my other thread.

I don't know where you find the post from pbuk, it's not in this thread. I don't have time to search other threads on the forum that is not related to me, I have a life and I don't go through other posts in other threads that are not addressed to me.

 

[Jarvis323]

I used cin.ignore(256) before and did not give me problem, I don't know where the program is anymore. That's the reason it never cross my mind that I have to count how many times to hit enter. I swear I read somewhere they put 256 for the hell of it and it worked for me before.

The documentation says that it will keep ignoring characters until it's either ignored n of them, or it sees the delimiter. The default delimiter is EOF (which is an int, probably -1). So this is exactly what it did.

I think you probably thought you were doing cin.ignore( 256, '\n' ); which will stop when it sees '\n', the newline character (when you press enter).

Anyways, the answer to your question is right there in the first part of the documentation. Mark hinted where specifically in the docs the answer lies.

You really need to look at the documentation of the ignore function.
This function takes two arguments, both with default values. If you call ignore with no arguments, as in cin.ignore();, what character, if encountered, causes ignore to return?

If the educator is asking you a question like this, it's probably a sign that answering it is the key to solving the problem. I think people are annoyed a little, because you let those hints go ignored, and in doing so you burden yourself quite a lot when it could have been avoided.

 

[yungman]

The documentation says that it will keep ignoring characters until it's either ignored n of them, or it sees the delimiter. The default delimiter is EOF (which is an int, probably -1). So this is exactly what it did.

I think you probably thought you were doing cin.ignore( 256, '\n' ); which will stop when it sees '\n', the newline character (when you press enter).

Anyways, the answer to your question is right there in the first part of the documentation. Mark hinted where specifically in the docs the answer lies.
If the educator is asking you a question like this, it's probably a sign that answering it is the key to solving the problem. I think people are annoyed a little, because you let those hints go ignored, and in doing so you burden yourself quite a lot when it could have been avoided.

That got to be it. I used that after cin >>. No wonder it works.

That was long time ago, I avoid using cin>> for cstring until lately. I even have a post on that because I want a keyboard that is more forgiving, that's why I go back to cin>>. Here is the post:

https://www.physicsforums.com/threa...ble-to-hitting-enter-key-accidentally.995802/

I should have left good alone. That started all the trouble when I try to switch back to cin>>.

Thanks

 

[phinds]

That started all the trouble when I try to switch back to cin>>.

And what started all the trouble in this thread and others is your infuriating habit of ignoring what people are telling you.

Instead of getting huffy and accusing me of being insulting for saying that, how about you give some thought as to WHY I would say what I just said.

 

[Mark44]

But every article I read is about what is the meaning of the first and second argument only.

Apparently you didn't understand any of these articles, particularly what they were saying about the second argument.
From the VS docs for ignore, for which I provided a link many posts back:

 basic_istream<Char_T, Tr>& ignore(    streamsize count = 1,    int_type delimiter = traits_type::eof());

count
The number of elements to skip from the current read position.
delimiter
The element that, if encountered before count, causes ignore to return and allowing all elements after delimiter to be read.

What the above means is that if you write this code -- cin.ignore(); -- you get the default arguments, the same as if you had written cin.ignore(1, EOF);

what's wrong putting cin.ignore() to get rid of one character.

What's wrong is that didn't realize, after several hints, which character you told your code to get rid of -- IOW, which character would cause ignore() (with no parameters) to return.

And what started all the trouble in this thread and others is your infuriating habit of ignoring what people are telling you.

Amen to that.

 

[yungman]

And what started all the trouble in this thread and others is your infuriating habit of ignoring what people are telling you.

Instead of getting huffy and accusing me of being insulting for saying that, how about you give some thought as to WHY I would say what I just said.

I did not accusing you at all, my problem is with Vanadium ONLY, I am grateful you and everyone else helping me. If there is any miss understanding, I apologize. I cannot say THANKS enough to everyone else for all the help.

I tried to send private message to explain a little better privately, but it was blocked, so I have no choice to say it out in the open. This has been an on going dispute with Vanadium, I tried to ignore his sarcasm and remarks for a while, I even pleaded with him to put me on ignore two week ago, it did not work. If you look back to my threads how he respond, he NEVER help, he always mock. Does Physics forum condone this? I was in your shoe 10 years ago on EE part of the forum, if I don't like someone, I just not respond, I don't go out of my way to insult them. I was there to help, I have a life, I don't go there to start a fight. It sure doesn't make me a better person insulting others that look for help.

Please excuse me if I offended you or anyone else. I have no choice but to say it out and I know it's unprofessional on this forum. Like I said, I really tried to send as private message instead of open post like this.

If there is a way to send pm in this forum, let me know, this is really bad of me to say it out like this out in the open. But I have no other choice. I know the difference between someone being blunt to me to help or just mocking and insulting.

Sorry.

 

[phinds]

PM is done by simply clicking on someone's avatar and selecting "start conversation"

As for the rest, I can see that you are a polite person and do not mean to cause offense but it is STILL true that in many threads people tell you the same thing over and over and you don't seem to listen.

@Vanadium 50 is one of our more gruff members, BUT ... he only tells the truth, it's just that he gets offended when someone does not listen. It is frustrating for people on this forum, not just him and not just me, to spend time answering your questions only to have their efforts turn out to be a waste of time or even to see that happening when someone else answers and you seem to ignore them.

Again, I can see that you do not mean to cause offense. I advise you to make more of an effort to understand each answer that you get.

Also, it often seems that you are trying to advance too quickly rather than taking things slowly and methodically, and you sometimes seem to have not made much of an effort to understand things before jumping in here with questions. That may be a misconception on our part but I just point it out so you will know how it appears.

And one more thing, I haven't noticed myself but others have commented that you have made apparently contradictory statements about (1) where you live exactly and (2) your level of education. You can clear those up easily by putting the correct information in the "ABOUT" section of your profile.

If I can be of further help, just let me know.

 

[yungman]

PM is done by simply clicking on someone's avatar and selecting "start conversation"

As for the rest, I can see that you are a polite person and do not mean to cause offense but it is STILL true that in many threads people tell you the same thing over and over and you don't seem to listen.

@Vanadium 50 is one of our more gruff members, BUT ... he only tells the truth, it's just that he gets offended when someone does not listen. It is frustrating for people on this forum, not just him and not just me, to spend time answering your questions only to have their efforts turn out to be a waste of time or even to see that happening when someone else answers and you seem to ignore them.

Again, I can see that you do not mean to cause offense. I advise you to make more of an effort to understand each answer that you get.

Also, it often seems that you are trying to advance too quickly rather than taking things slowly and methodically, and you sometimes seem to have not made much of an effort to understand things before jumping in here with questions. That may be a misconception on our part but I just point it out so you will know how it appears.

And one more thing, I haven't noticed myself but others have commented that you have made apparently contradictory statements about (1) where you live exactly and (2) your level of education. You can clear those up easily by putting the correct information in the "ABOUT" section of your profile.

If I can be of further help, just let me know.

Thanks for you advice.

I did click the person's name and went into starting a conversation to try writing to individual people last night, I started with Jarvis323, I wrote half a page and when I tried to send, it said I am not able to send to him! That's the reason I wrote the post here.

Regarding to this thread, I really did listen, I went on line to search and read quite a few including what Mark44 linked to me. Everyone of them talked about how many to ignore and explained all that. My problem was I used ignore(256, '\n') before and it works, that made me thing I can just put it in like insurance if the buffer is empty, it doesn't matter. I did NOT know or read about if I put the ignore(x) in, I have to hit Enter x times before it will exit out. That's the part I did not get.

I do try to listen and do my search, but just not necessary lucky all the time.

I have a BS in Chemistry, but I never use it, never work a day in Chemistry. My career is EE and manager of EE. I only had an AA degree in the late 70s mainly on microprocessor and assembly language, I completely self study in analog, RF, electromagnetics and math. I live in Sunnyvale Ca. in the heart of Silicon Valley close to San Jose. I studied very hard in my life, I published two papers in America Institute of Physics Review of Scientific Instruments and have two US patents solely under my name. All from self studying, nothing to put in the description in the Forum here. That's why I don't put anything there. Here are the links of those. all in electronics.

https://aip.scitation.org/doi/10.1063/1.1144464

https://aip.scitation.org/doi/abs/10.1063/1.1310339?class=pdf&journalCode=rsi

https://patents.google.com/patent/US7561438B1/en?inventor=yungman+liu&oq=yungman+liu

https://patents.google.com/patent/U...yungman+alan+liu&oq=Inventor+yungman+alan+liuThanks

 

[Mark44]

Again, I can see that you do not mean to cause offense. I advise you to make more of an effort to understand each answer that you get.

Strongly agree to both statements.

Also, it often seems that you are trying to advance too quickly rather than taking things slowly and methodically, and you sometimes seem to have not made much of an effort to understand things before jumping in here with questions. That may be a misconception on our part but I just point it out so you will know how it appears.

I don't think this is a misconception. The point about taking things more slowly and methodically has been made before -- several times, I believe.
A slower approach would help to mitigate incorrect conclusions about what is causing a particular problem; for example, concluding incorrectly that a problem was caused by getline(), when the actual culprit was a misunderstanding of how ignore() works.

 

[phinds]

I don't think this is a misconception.

I know. I was practicing being very polite and tactful