Question

I have been struggling with this issue for one week, and still far away from the solution.

I am reading a plain text from a file by strings. It reads everything successfully until a character with a certain large number.
This number is specific for each file, and is different for different files.

After reaching the character with this large number, it reads “\x01\0\0\0” sequences (consist of 4 chars) for one or more times instead of the original characters, and then reads everything correctly (until the next large number).

In a couple of words – instead of reading this:

... many characters ...
First read failure!
... many characters ...
Second read failure, second read failure!
... many characters ...
etc.

it reads this:

... many characters ...
First read f\x01\0\0\0re
... many characters ...
Second read failu\x01\0\0\0\x01\0\0\0\x01\0\0\0ead failure!
... many characters ...
etc.

Do you have any ideas about the cause of this problem?

Additional information:

 1) The "ferror" condition is not true.
 2) I am reading from file using streams (fopen, fread, fclose)
 3) Have tested different read methods: "fread cycle" and "fgets".
    The results are the same.
 4) The binary optimization is disabled in the compiler's settings.
    It seams that the problem is not connected with a compiler.
    Neither GCC, nor Apple LLVM gives me the desired result.
 5) Attempting to solve the issue, I converted the whole project from C++ to C,
    but the problem doesn't disappear.

Answer 1

As I suggested in comments:

• you should put your source code in the question (which you can always edit to improve it), not in comments
• you should use POSIX 2008 getline function on your FILE*.
• you should free the buffer implicitly malloc-ed by getline later, e.g. when fclose-ing your FILE* handle.
• you should learn to use a debugger like gdb
• you should learn to use a memory leakage detector like valgrind