Real binary write PHP

Question

How do I do something as simple as (in PHP) this code in C:

char buffer[5] = "testing";
FILE* file2 = fopen("data2.bin", "wb");
fwrite(buffer, sizeof buffer, 1, file2);
fclose(file2);

Whenever I try to write a binary file in PHP, it doesn't write in real binary.

Example:

$ptr = fopen("data2.bin", 'wb');

fwrite($ptr, "testing");

fclose($ptr);

I found on internet that I need to use pack() to do this...

What I expected:

testing\9C\00\00
or
7465 7374 696e 679c 0100 00

What I got:

testing412

Thanks

What makes you think it's not writing in "real binary"? What does that mean? — Barmar, Jul 04 '13 at 04:13
What do you mean "it doesn't write in real binary" ? What were you expecting the result to be, and what did you get? — Jonathon Reinhart, Jul 04 '13 at 04:14
I'm expecting to get something like 4d61 7572 6963 696f 0063 ecbf 855a 61b7 00d6 7bb7 1200 0000 00ba 7f47 but I only get plaintext like a regular text file. — Maurício Giordano, Jul 04 '13 at 04:15
Whether a file is binary or text is just an interpretation by the program using it. Everything is binary internally, that's all that computers know about. — Barmar, Jul 04 '13 at 04:20
Note that neither of your example code matches your output! Please don't show the code from one thing, and the output from another... it is *ridiculously* confusing (to you and us!) — Jonathon Reinhart, Jul 04 '13 at 04:37
same as my question, same problem https://stackoverflow.com/questions/62144562/crazy-php-fopen-fwrite-file-put-contents-bug-with-binary-mode-writing-nothin — SSpoke, Jun 02 '20 at 04:30

Jonathon Reinhart · Accepted Answer · 2013-07-04T04:30:55.897

5

You're making the classic mistake of confusing data with the representation of that data.

Let's say you have a text file. If you open it in Notepad, you'll see the following:

hello
world

This is because Notepad assumes the data is ASCII text. So it takes every byte of raw data, interprets it as an ASCII character, and renders that text to your screen.

Now if you go and open that file with a hex editor, you'll see something entirely different¹:

68 65 6c 6c 6f 0d 0a 77 6f 72 6c 64          hello..world

That is because the hex editor instead takes every byte of the raw data, and displays it as a two-character hexadecimal number.

_{1 - Assuming Windows \r\n line endings and ASCII encoding.}

So if you're expecting hexadecimal ASCII output, you need to convert your string to its hexadecimal encoding before writing it (as ASCII text!) to the file.

In PHP, what you're looking for is the bin2hex function which "Returns an ASCII string containing the hexadecimal representation of str." For example:

$str = "Hello world!";
echo bin2hex($str);      // output:  48656c6c6f20776f726c6421

Note that the "wb" mode argument doesn't cause any special behavior. It guarantees binary output, not hexadecimal output. I cannot stress enough that there is a difference. The only thing the b really does, is guarantee that line endings will not be converted by the library when reading/writing data.

edited Jul 04 '13 at 04:30

answered Jul 04 '13 at 04:24

Jonathon Reinhart

132,704
33
254
328

1

Thanks! But why PHP doesn't do that already? In C, if I put "wb" as argument on fopen, it saves everything in binary... But in PHP it seems that it is just ignored. – Maurício Giordano Jul 04 '13 at 04:27
1

@MaurícioGiordano No, C absolutely does not do what you just described. Please compile and run your C code from the question, and open the result in a *text editor*. You'll see ASCII. The `"wb"` mode does **not** do what you think it does. – Jonathon Reinhart Jul 04 '13 at 04:28
But it doesn't save in ASCII, because if I open with Notepad, it shows strange characters... – Maurício Giordano Jul 04 '13 at 04:29
1

Sample: char buffer[7] = "testing"; int number = 412; FILE* file2 = fopen("data2", "wb"); fwrite(buffer, 1, sizeof(buffer), file2); fwrite(&number, 1, sizeof(number), file2); fclose(file2); – Maurício Giordano Jul 04 '13 at 04:30
1

Right, you're writing an **integer** `412` to the file. So if you look at just that part in a hex editor, you're going to see `9c 01 00 00`, which is a 4-byte little-endian integer shown as four 1-byte hexadecimal numbers. If you wrote the **string** `"`char str[] = "412"`, you would see something totally different. – Jonathon Reinhart Jul 04 '13 at 04:34
3

You should really play around with this stuff until it makes *perfect sense*, as it is very fundamental to how computers and binary data work. – Jonathon Reinhart Jul 04 '13 at 04:36

Real binary write PHP

1 Answers1