admin/mruby-mruby
1
0
Fork 0
mirror of https://github.com/mruby/mruby synced 2026-06-08 16:11:16 +00:00
Code Issues Packages Projects Releases Wiki Activity

mruby-pack: add README.md

Browse Source 
The document is written by Google Jules.
This commit is contained in:
Yukihiro "Matz" Matsumoto
2025-06-16 12:45:18 +09:00
parent 0fc6288ae2
commit 67dbaf856f
1 changed files with 97 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files
mrbgems/mruby-pack/README.md
+97 -39
View File
@@ -1,53 +1,111 @@
# mruby-pack (pack / unpack)
# mruby-pack
mruby-pack provides `Array#pack` and `String#unpack` for mruby.
The `mruby-pack` mrbgem enhances mruby by providing `Array#pack` and `String#unpack` methods. These methods facilitate the conversion between Ruby data types (specifically arrays and strings) and binary data representations, offering a powerful way to handle binary data.
## Installation
Add the line below into your build configuration:
```
conf.gem :core => 'mruby-pack'
```ruby
conf.gem :core => 'mruby-pack'
```
There is no dependency on other mrbgems.
## Supported template string
## `Array#pack`
- A : arbitrary binary string (space padded, count is width)
- a : arbitrary binary string (null padded, count is width)
- B : bit string (descending)
- b : bit string (ascending)
- C : 8-bit unsigned (unsigned char)
- c : 8-bit signed (signed char)
- D, d: 64-bit float, native format
- E : 64-bit float, little endian byte order
- e : 32-bit float, little endian byte order
- F, f: 32-bit float, native format
- G : 64-bit float, network (big-endian) byte order
- g : 32-bit float, network (big-endian) byte order
- H : hex string (high nibble first)
- h : hex string (low nibble first)
- I : unsigned integer, native endian (`unsigned int` in C)
- i : signed integer, native endian (`int` in C)
- J : unsigned integer, native endian (`uintptr_t` in C)
- j : signed integer, native endian (`intptr_t` in C)
- L : 32-bit unsigned, native endian (`uint32_t`)
- l : 32-bit signed, native endian (`int32_t`)
- m : base64 encoded string (see RFC 2045, count is width)
- N : 32-bit unsigned, network (big-endian) byte order
- n : 16-bit unsigned, network (big-endian) byte order
- Q : 64-bit unsigned, native endian (`uint64_t`)
- q : 64-bit signed, native endian (`int64_t`)
- S : 16-bit unsigned, native endian (`uint16_t`)
- s : 16-bit signed, native endian (`int16_t`)
- U : UTF-8 character
- V : 32-bit unsigned, VAX (little-endian) byte order
- v : 16-bit unsigned, VAX (little-endian) byte order
- x : null byte
- X : back up bytes
- Z : same as "a", except that null is added with \*
- @ : absolute position
The `Array#pack` method converts an array into a binary string. It takes a template string as an argument, which dictates how each element in the array should be converted and packed into the resulting binary string. The template string consists of directives, where each directive specifies the type and format of the data to be packed.
**Example:**
```ruby
arr = [65, 66, 67]
binary_string = arr.pack("ccc") # Packs three 8-bit signed integers
# binary_string will be "ABC"
arr2 = [72, 101, 108, 108, 111]
binary_string2 = arr2.pack("C*") # Packs all elements as 8-bit unsigned integers
# binary_string2 will be "Hello"
arr3 = [0x1234, 0x5678]
binary_string3 = arr3.pack("S>2") # Packs two 16-bit unsigned integers, big-endian
# binary_string3 will be "\x12\x34\x56\x78" (depending on native endianness if S is used without > or <)
# Using "n2" would be more explicit for network (big-endian) byte order:
# binary_string3 = arr3.pack("n2") # Packs two 16-bit unsigned, network (big-endian) byte order
# binary_string3 will be "\x12\x34\x56\x78"
```
## `String#unpack`
The `String#unpack` method performs the reverse operation of `Array#pack`. It takes a binary string and a template string as input. It extracts data from the binary string according to the directives in the template and converts it into an array of Ruby objects.
**Example:**
```ruby
binary_data = "test"
arr = binary_data.unpack("aaaa")
# arr will be ["t", "e", "s", "t"]
binary_data2 = "\x01\x02\x03"
arr2 = binary_data2.unpack("C3") # Unpacks three 8-bit unsigned integers
# arr2 will be [1, 2, 3]
binary_data3 = "Hello\x00World"
arr3 = binary_data3.unpack("Z*Z*") # Unpacks two null-terminated strings
# arr3 will be ["Hello", "World"]
```
## Supported Template Directives
The template string is a sequence of characters that specify the type and format of the data to be packed or unpacked. Each directive can be followed by a count (e.g., "C4" for four 8-bit unsigned integers) or `*` to consume all remaining items/bytes.
Here is a list of supported template characters and their meanings:
| Directive | Description |
| --------- | ------------------------------------------------------------------------------------------------------- |
| `A` | Arbitrary binary string (space padded, count is width) |
| `a` | Arbitrary binary string (null padded, count is width) |
| `B` | Bit string (descending bit order) |
| `b` | Bit string (ascending bit order) |
| `C` | 8-bit unsigned integer (unsigned char) |
| `c` | 8-bit signed integer (signed char) |
| `D`, `d` | Double-precision float, native format (64-bit) |
| `E` | Double-precision float, little-endian (64-bit) |
| `e` | Single-precision float, little-endian (32-bit) |
| `F`, `f` | Single-precision float, native format (32-bit) |
| `G` | Double-precision float, network (big-endian) byte order (64-bit) |
| `g` | Single-precision float, network (big-endian) byte order (32-bit) |
| `H` | Hex string (high nibble first) |
| `h` | Hex string (low nibble first) |
| `I` | Unsigned integer, native endian (`unsigned int` in C) |
| `i` | Signed integer, native endian (`int` in C) |
| `J` | Unsigned integer, native endian (`uintptr_t` in C) |
| `j` | Signed integer, native endian (`intptr_t` in C) |
| `L` | 32-bit unsigned integer, native endian (`uint32_t`) |
| `l` | 32-bit signed integer, native endian (`int32_t`) |
| `m` | Base64 encoded string (see RFC 2045, count is input bytes for pack, output chars for unpack) |
| `N` | 32-bit unsigned integer, network (big-endian) byte order |
| `n` | 16-bit unsigned integer, network (big-endian) byte order |
| `Q` | 64-bit unsigned integer, native endian (`uint64_t`) |
| `q` | 64-bit signed integer, native endian (`int64_t`) |
| `S` | 16-bit unsigned integer, native endian (`uint16_t`) (Use `S>` for big-endian, `S<` for little-endian) |
| `s` | 16-bit signed integer, native endian (`int16_t`) (Use `s>` for big-endian, `s<` for little-endian) |
| `U` | UTF-8 character |
| `V` | 32-bit unsigned integer, VAX (little-endian) byte order |
| `v` | 16-bit unsigned integer, VAX (little-endian) byte order |
| `x` | Null byte (skip forward one byte) |
| `X` | Back up one byte |
| `Z` | Null-terminated string (when unpacking, reads until NULL; when packing, appends a NULL if count is `*`) |
| `@` | Moves to absolute position (offset from the beginning of the string) |
**Modifiers:**
Some directives can be followed by modifiers that affect their behavior:
- `>`: Big-endian (for `S`, `s`, `L`, `l`, `Q`, `q`, etc., when native order is not desired).
- `<`: Little-endian (for `S`, `s`, `L`, `l`, `Q`, `q`, etc., when native order is not desired).
These modifiers are typically used with integer and float types to specify byte order explicitly. For network byte order (big-endian), directives like `n` (16-bit) and `N` (32-bit) are commonly used.
## License