I need to read a large file of either text, binary, or combination, such as a JPEG file, encrypt it, and write it to a file. At some later time I will need to read the encrypted data, and decrypt it.
The end goal is to verify that the decrypted data matches the original data.
My problem is that with large files greater than 1Meg, I don't want to read and write character by character. I am targeting this code for a phone and I/O will cause too long a delay for the user.
With a pure text file, using fread() and fwrite() convert the data to binary, and the result is different than the original. With a jpeg image, it appears that there is some textual content mixed in with the binary data.
Is there a way to efficiently read in an arbitrary type of file and write it back in the original format?
Or is character by character the only option?
Or am I still out of luck?
After debugging it turned out that the decrypt function had the plain text and cipher text buffers assigned backwards. After swapping the buffer assignments, the decrypted results matched the original data. I originally thought that maybe reading the text as binary and then rewriting as binary would not appear as text, but I was wrong.
Reading the entire file as binary works just fine.
Related
I have a dataset, a ".inp" format file, and I need to read this file in c++. However, the fopen() fread() method seemed to fail and read the wrong data(e.g. the first integer should be 262144, the fread yields an integer much larger than this nevertheless).
To be more specific, my ".inp" file contains a few integers and float points, how can I read them successfully in c++?
enter image description here
This is the screenshot of the "*.inp" file from Notepad++. Basically this is a text file.
I solved it by coping the data into a txt. However, I am still not aware how to read "*.inp"
I found some info about INP file extension. It seems like there are multiple variances of it, each meant to be used for different purpose. Where is your file coming from? As for soultion, if you can't open the file using fopen/fstream normally, you could treat it as binary and read each value in the way you specify. Other than that, I could think of calling system functions to get file contents (like cat in linux for example), then if there are some random characters, you could parse your string to ommit them.
Here is example of how to call cat in C++:
Simple way to call 'cat' from c++?
In a game I'm making I need to read a map from a file. Assuming some of the data in the beginning is written in characters, but the tile map is written in binary, I would open the file in text mode then switch it to binary mode once it reaches the tile data.
Is there an easy, or standard, way of changing an ifstream from text mode to binary mode while keeping the same position in the file?
This also applies to the writting part, I will need to start writting into the file using characters, then change to binary mode.
EDIT: I'm using text mode to make this readable and to read strings of unknown size. For example, this line:
map-name=TestMap
I'd read this with
getline( mapFile, attribute, '=' );
getline( mapFile, mapName, '\n' );
How would I read this in binary mode if there won't be newline characters?
The mode is established when the file is opened, and cannot be
changed later. If there is any binary data in the file, you
must use binary mode. But where is the problem? You can read
text in binary mode; line endings might appear a bit strange
(but not if you also wrote it in binary mode), but otherwise,
there should be no problem as long as the binary data actually
is text.
If you are responsible for writing the files as well, the simplest (and perhaps sanest) solution might be to write two files.
One in text, for text you wish to be human readable.
And the second as binary, for things like maps. In fact that way you could have one binary map file for each map.
I am working on a project where we need to convert WAV file audio input into plain ASCII characters. The input WAV file will contain a single short alphanumeric code e.g. asdrty543 and each character will be pronounced one by one when you play the WAV file. Our requirement is that when a single character code is pronounced we need to convert it into it's equivalent ASCII code. The implementation will be done in C/C++ as un-managed Win32 DLL. We are open to use third party libraries. I am already googling for directions. However, I will really appreciate it if I can get directions/pointers from an experienced programmer who has already worked on similar requirement. Thank you in advance for your help.
ASCII characters like Az09 are only a portion of the ASCII Table. WAV files like any other file is stored and accessed in bytes.
1 byte has 256 different values. Therefore one can't simply convert bytes into Az09 since there are not enough Az09 characters.
You'll have to find a library which opens WAV files and creates the wave format for you. In relation to the wave's intensity and length, a chain of Az or Az09 characters can be produced.
I believe you're trying to convert the wave to a series of notes. That's possible too, using the same approach.
I wrote a program in C++ that compresses a file.
Now I want to see the contents of the compressed file.
I used hexdump but I dont know what the hex numbers mean.
For example I have:
0000000 00f8
0000001
How can I convert that back to something that I can compare with the original file contents?
If you implemented a well-known compression algorithm you should be able to find a tool that performs the same kind of compression and compare its results with yours. Otherwise you need to implement an uncompressor for your format and check that the result of compressing and then uncompressing is identical to your original data.
That looks like a file containing the single byte 0xf8. I say that since it appears to have the same behaviour as od under UNIX-like operating systems, with the last line containing the length and the contents padded to a word boundary (you can use od -t x1 to get rid of the padding, assuming your od is advanced enough).
As to how to recreate it, you need to run it through a decryption process that matches the encryption used.
Given that the encrypted file is that short, you either started with a very small file, your encryption process is broken, or it's incredibly efficient.
I am learning C++ and decided to train me by making a little program that extract files from zip, like text files, images, or even other zip files (but I don't want to extract them directly, one thing a time) with the libzip library.
So I made my program, but now I have a problem.
It extracts well text files, but not files like images or zip. It detects them, gives me exact names and sizes, but once extracted, they are just a few bytes. (but they are located where they should).
Here is my code: http://pastie.org/6221955
So if someone could help me to extract files that aren't texts from zip, it would be great! Thank you!
You're reading and writing binary data as a textual string. The problem is that strings use the presence of a NULL character (0-byte) to indicate end-of-string. Binary data can (and definitely does) contain zeros all over the place, not just at the end.
You need to use ofstream's .write (buffer, <size in bytes>) to write to the disk; by manually specifying the size in bytes, you force it read that many bytes instead of stopping at the first instance of a NULL character.
The issue is with the << operator. You output a character array / string. Strings in C are null terminated. Thus the first binary 0 will terminate your output.