ASCII table

ASCII (American Standard Code for Information Interchange) was first published in the US in 1963 and last revised as ANSI X3.4-1986. It is a 7-bit code: 128 values, each stored in a byte whose top bit is 0. RFC 20 (1969) adopted it for data sent over the ARPANET, and protocols like HTTP and email headers still use it.

Unicode starts with the same 128 characters, and UTF-8 stores them as the same single bytes, so plain ASCII text is valid UTF-8.

Most of the 33 control codes came from teleprinters and are unused today. These are the ones that still matter:

Caret notation (^C, ^[) names a control code by the key you press with Ctrl. Ctrl keeps only the low 5 bits of the key's code, so C (67) becomes 3 and [ (91) becomes 27, which is why Ctrl+[ works as Escape in Vim. ^? is the exception: it means DEL, 127.

• Case is one bit. A lowercase letter is its uppercase letter plus 32 (bit 5, 0x20). c ^ 0x20 flips the case of an ASCII letter and c | 0x20 makes it lowercase.
• Digits are 48-57. c - 48 (or c - '0' in C) turns a digit character into its value.
• Uppercase sorts first. Plain byte comparison puts "Zebra" before "apple", because Z (90) is less than a (97). Use a locale-aware compare for human sorting.
• Hex digits are not contiguous. 0-9 are 48-57 but A-F are 65-70, so hex parsing needs two ranges.
• Not every symbol has a named HTML entity. ˜ is the small tilde U+02DC, not ~ (126), and the hyphen-minus (45) has none. Use ~ or the character itself.

// JavaScript
'A'.charCodeAt(0)          // 65
'😀'.codePointAt(0)        // 128512 (use codePointAt beyond ASCII)
String.fromCharCode(72, 105)  // "Hi"
[...'Hi'].map(c => c.charCodeAt(0).toString(16))  // ['48', '69']

# Python
ord('A')                  # 65
chr(97)                   # 'a'
'Hi'.encode('ascii')      # b'Hi'  (raises UnicodeEncodeError for é)
[hex(b) for b in b'Hi']   # ['0x48', '0x69']

/* C */
char c = 'A';
printf("%d %x %c\n", c, c, c + 32);   /* 65 41 a */

# bash
printf '%d\n' "'A"        # 65
printf '\x41\n'           # A
printf 'Hi' | xxd         # 00000000: 4869  Hi
man ascii                 # the full table in your terminal (Linux)

Bytes 128-255 are outside ASCII, and different encodings put different characters there. Windows-1252 has € at 0x80 and curly quotes at 0x91-0x94, where ISO-8859-1 has invisible C1 control codes. From 0xA0 up, both match Unicode U+00A0-U+00FF. In UTF-8, these characters take two or more bytes instead of one.

That mismatch causes mojibake. The UTF-8 bytes for é are C3 A9; read as Windows-1252 they show as é. If you see that, the text is UTF-8 and the reader picked the wrong encoding. Declare charset=utf-8 in the Content-Type header or <meta charset="utf-8">.

For characters beyond ASCII, use the Unicode converter. To turn text into bits, try text to binary or the binary converter, and to escape strings for JSON, HTML or a shell, the string escape tool. For editors, linters and other code tools, browse code tools on DevHunt.