Problem with strcat() function

**sureshkumarct** · January 1st, 2007, 10:59 PM

I want to create an html file using C and as the first step , I wrote a simple program as follows:

#include<stdio.h>
#include<conio.h>
#include<string.h>
void main()
{
char *name="My Namel",*str="";
clrscr();
strcat(str,"<html><head>");
strcat(str,name);
strcat(str,"</html></head>");
puts(str);
getch();
}

But the puts(str) statement is giving unexpected result.What is getting printed is "<html><head>My Name Name ad>". Why doesnt it print "<html><head>My Name </html></head>" as is expected? .Ultimately I want to write the string str to the html file using file operation. I tried adding a null character at the end of str, but that throws some illegal operations exceptions and Turbo C is terminated ubruptly.Thanks in advance

**Paul McKenzie** · January 1st, 2007, 11:45 PM

Read the documentation on strcat(). The strcat() requires that there is room for all of those characters that you are adding.

Code:

str="";

This does not define a character buffer big enough to hold all of those characters:

Code:

char str[200];

That is what it should be.

Also, main() returns int, not void.

Regards,

Paul McKenzie

**TheCPUWizard** · January 2nd, 2007, 04:10 AM

Paul is right. Get away from char pointers NOW and use std::string

1) Many Companies are outlaying the use of such structures.
2) Many companies are not hiring if you are not fluent in STL

If you are learning from a book that is over three years old, then get a ney one.

**humptydumpty** · January 2nd, 2007, 04:32 AM

Originally Posted by TheCPUWizard

Paul is right. Get away from char pointers NOW and use std::string

1) Many Companies are outlaying the use of such structures.
2) Many companies are not hiring if you are not fluent in STL

If you are learning from a book that is over three years old, then get a ney one.

TheCPUWizard .i think you Didn't get OP question .he is writing a code in C .then why C++ Suggestion Comes in Picture.and if still OP want to use Pointers in Code he can use malloc and Realloc Function for this.with the help of those above Function he can write his code in a efficient manner.

Thanx

**NMTop40** · January 2nd, 2007, 10:11 AM

To build the string in C it is probably preferable to use sprintf rather than strcat.

thus:

Code:

sprintf( str, "<html><head>%s</head></html>", name );

You have to ensure str points to an adequately allocated amount of memory. Remember to include one for the null terminator. In this example I make that 27 + strlen( name ). Preferable to avoid actually putting magic numbers like 27 into the code.

**exterminator** · January 2nd, 2007, 11:42 AM

Originally Posted by TheCPUWizard

1) Many Companies are outlaying the use of such structures.
2) Many companies are not hiring if you are not fluent in STL

I am not saying against std::string but I haven't heard of this. There is no such outlaying or not hiring rules that I know of. There are huge systems already in C and older C++ days where std::string did not exist. Any documentation of such companies that you know of to support your words?

Originally Posted by humptydumpty

TheCPUWizard .i think you Didn't get OP question .he is writing a code in C .then why C++ Suggestion Comes in Picture.and if still OP want to use Pointers in Code he can use malloc and Realloc Function for this.with the help of those above Function he can write his code in a efficient manner.

How can repeated calls to malloc and/or realloc be an efficient manner? Better is to allocate once (after estimating the max storage) and work with sprintf as suggested by NMTop40.

**SuperKoko** · January 2nd, 2007, 03:37 PM

Originally Posted by exterminator

How can repeated calls to malloc and/or realloc be an efficient manner? Better is to allocate once (after estimating the max storage) and work with sprintf as suggested by NMTop40.

And, if possible, check that the string is not too long for the buffer and fail with a pretty error message in that case, instead of getting a buffer overrun and undefined behavior.

**humptydumpty** · January 3rd, 2007, 12:24 AM

Originally Posted by exterminator

I am not saying against std::string but I haven't heard of this. There is no such outlaying or not hiring rules that I know of. There are huge systems already in C and older C++ days where std::string did not exist. Any documentation of such companies that you know of to support your words? How can repeated calls to malloc and/or realloc be an efficient manner? Better is to allocate once (after estimating the max storage) and work with sprintf as suggested by NMTop40.

i think you get your question ans.

Thanx

**exterminator** · January 3rd, 2007, 12:38 AM

Originally Posted by humptydumpty

i think you get your question ans.

No. I didn't.

**Vanaj** · January 3rd, 2007, 01:06 AM

Originally Posted by exterminator

How can repeated calls to malloc and/or realloc be an efficient manner? Better is to allocate once (after estimating the max storage) and work with sprintf as suggested by NMTop40.

What if max storage can not be estimated... I.E. a word processor...

**exterminator** · January 3rd, 2007, 01:34 AM

Originally Posted by Vanaj

What if max storage can not be estimated... I.E. a word processor...

That is a different problem altogether, isn't it? Looking at the OP, I think he can do with what I suggested. Otherwise, there needs to be an encapsulation provided. As SuperKoko suggests, check for overflow before appending - and then double the size of the buffer and copy over the previous written bytes. The logic would be similar to how a vector and probably a string object grows. Still that would not need repeated calls to malloc/realloc every time anything needs to be appended. At least, that is the idea that I had in my mind regarding the problem.

**Vanaj** · January 3rd, 2007, 01:44 AM

Originally Posted by exterminator

That is a different problem altogether, isn't it? Looking at the OP, I think he can do with what I suggested. Otherwise, there needs to be an encapsulation provided. As SuperKoko suggests, check for overflow before appending - and then double the size of the buffer and copy over the previous written bytes. The logic would be similar to how a vector and probably a string object grows. Still that would not need repeated calls to malloc/realloc every time anything needs to be appended. At least, that is the idea that I had in my mind regarding the problem.

Take what NMTop40 and SuperKoko posted and combine the suggestions and all will work just fine...

**exterminator** · January 3rd, 2007, 01:48 AM

Originally Posted by Vanaj

Take what NMTop40 and SuperKoko posted and combine the suggestions and all will work just fine...

Ok...

Thread: Problem with strcat() function

Thread Tools

Display

Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Re: Problem with strcat() function

Posting Permissions