github-repos/bilibili-API-collect

Fork 0

mirror of https://github.com/SocialSisterYi/bilibili-API-collect.git synced 2024-12-28 21:40:08 +08:00

Kataick f16cfe9793

规范bv转av(python)示例的类型 (#960 )

2024-02-01 14:13:56 +08:00

18 KiB

Raw Blame History

bvid说明

2020-03-23 B站推出了全新的稿件视频idbvid来接替之前的avid，其意义与之相同

详见：

概述

格式

“bvid”恒为长度为 12 的字符串，前两个固定为“BV1”，后 9 个为 base58 计算结果（不包含数字 0 和大写字母 I、 O 以及小写字母 l）

实质

“bvid"为“avid”的base58编码，可通过算法进行相互转化

avid发号方式的变化

从 2009-09-09 09:09:09 av2 的发布到 2020-03-28 19:45:02 av99999999 的发布B站结束了以投稿时间为顺序的avid发放，改为随机发放avid

~~暗示B站东方要完？泪目~~

算法概述

~~算法以及程序主要参考知乎@mcfx的回答~~
~~实际上该算法并不完整,新的算法参考自【揭秘】av号转bv号的过程~~
实际上上面的算法依然不完整，新的算法参考自 SocialSisterYi#740~~来自 B 站某个 JS 文件？~~

av->bv算法

说明

目前的 BV 格式为 BV1XXXXXXXXX，以 BV1 开头，后面包含 9 位有效数据。
AV 最大值为 2⁵¹。

算法

定义一个包含初始值为 ['B', 'V', '1', '0', '0', '0', '0', '0', '0', '0', '0', '0'] 的长度为 12 的数组bytes，用于存储转换后的字符。
- 定义变量 bv_idx 并初始化为数组 bytes 的最后一个索引。
- 将输入的 aid 与 avid 最大值（2⁵¹）进行按位或运算，其结果与常量 XOR_CODE（23442827791579）进行异或运算，得到变量 tmp。
- 当 tmp 大于0时，循环执行以下操作直到小于0：
  - 将 tmp 除以 58（码表的长度）的余数作为索引，从 FcwAPNKTMug3GV5Lj7EJnHpWsx4tb8haYeviqBz6rkCy12mUSDQX9RdoZf 码表中取出对应的字符，并将其赋值给 bytes[bv_idx]。
  - 将 tmp 与 58 求模赋值给 tmp。
  - 将 bv_idx 减1。
- 将 bytes 数组中索引为 3 和 9 的元素进行交换。
- 将 bytes 数组中索引为 4 和 7 的元素进行交换。
- 将 bytes 数组转换为字符串，并返回结果。

bv->av算法

是 #av->bv算法的逆向

将 bvid 中索引为 3 和 9 的字符进行交换。
将 bvid 中索引为 4 和 7 的字符进行交换。
删除 bvid 前3个字符（固定为 BV1）。
定义变量 tmp 并初始化为 0。
遍历 bvid 的每个字符，执行以下操作：
- 获取当前字符在 FcwAPNKTMug3GV5Lj7EJnHpWsx4tb8haYeviqBz6rkCy12mUSDQX9RdoZf 码表中的索引，并将其赋值给变量 idx。
- 将 tmp 乘以常量 58，并加上 idx，最后赋值给 tmp。
将 tmp 与常量 2⁵¹ - 1 进行按位与运算，其结果与常量 XOR_CODE（23442827791579）进行异或运算，得到最终结果。

编程实现

JavaScript/TypeScript

const XOR_CODE = 23442827791579n;
const MASK_CODE = 2251799813685247n;
const MAX_AID = 1n << 51n;
const BASE = 58n;

const data = 'FcwAPNKTMug3GV5Lj7EJnHpWsx4tb8haYeviqBz6rkCy12mUSDQX9RdoZf';

function av2bv(aid) {
  const bytes = ['B', 'V', '1', '0', '0', '0', '0', '0', '0', '0', '0', '0'];
  let bvIndex = bytes.length - 1;
  let tmp = (MAX_AID | BigInt(aid)) ^ XOR_CODE;
  while (tmp > 0) {
    bytes[bvIndex] = data[Number(tmp % BigInt(BASE))];
    tmp = tmp / BASE;
    bvIndex -= 1;
  }
  [bytes[3], bytes[9]] = [bytes[9], bytes[3]];
  [bytes[4], bytes[7]] = [bytes[7], bytes[4]];
  return bytes.join('');
}

function bv2av(bvid) {
  const bvidArr = Array.from(bvid);
  [bvidArr[3], bvidArr[9]] = [bvidArr[9], bvidArr[3]];
  [bvidArr[4], bvidArr[7]] = [bvidArr[7], bvidArr[4]];
  bvidArr.splice(0, 3);
  const tmp = bvidArr.reduce((pre, bvidChar) => pre * BASE + BigInt(data.indexOf(bvidChar)), 0n);
  return Number((tmp & MASK_CODE) ^ XOR_CODE);
}

console.log(av2bv(111298867365120));
console.log(bv2av('BV1L9Uoa9EUx'));

const XOR_CODE = 23442827791579n;
const MASK_CODE = 2251799813685247n;
const MAX_AID = 1n << 51n;
const BASE = 58n;

const data = 'FcwAPNKTMug3GV5Lj7EJnHpWsx4tb8haYeviqBz6rkCy12mUSDQX9RdoZf';

function av2bv(aid: number) {
  const bytes = ['B', 'V', '1', '0', '0', '0', '0', '0', '0', '0', '0', '0'];
  let bvIndex = bytes.length - 1;
  let tmp = (MAX_AID | BigInt(aid)) ^ XOR_CODE;
  while (tmp > 0) {
    bytes[bvIndex] = data[Number(tmp % BigInt(BASE))];
    tmp = tmp / BASE;
    bvIndex -= 1;
  }
  [bytes[3], bytes[9]] = [bytes[9], bytes[3]];
  [bytes[4], bytes[7]] = [bytes[7], bytes[4]];
  return bytes.join('') as `BV1${string}`;
}

function bv2av(bvid: `BV1${string}`) {
  const bvidArr = Array.from<string>(bvid);
  [bvidArr[3], bvidArr[9]] = [bvidArr[9], bvidArr[3]];
  [bvidArr[4], bvidArr[7]] = [bvidArr[7], bvidArr[4]];
  bvidArr.splice(0, 3);
  const tmp = bvidArr.reduce((pre, bvidChar) => pre * BASE + BigInt(data.indexOf(bvidChar)), 0n);
  return Number((tmp & MASK_CODE) ^ XOR_CODE);
}

console.log(av2bv(111298867365120));
console.log(bv2av('BV1L9Uoa9EUx'));

Python

来自：SocialSisterYi#847 (comment)

XOR_CODE = 23442827791579
MASK_CODE = 2251799813685247
MAX_AID = 1 << 51

data = [b'F', b'c', b'w', b'A', b'P', b'N', b'K', b'T', b'M', b'u', b'g', b'3', b'G', b'V', b'5', b'L', b'j', b'7', b'E', b'J', b'n', b'H', b'p', b'W', b's', b'x', b'4', b't', b'b', b'8', b'h', b'a', b'Y', b'e', b'v', b'i', b'q', b'B', b'z', b'6', b'r', b'k', b'C', b'y', b'1', b'2', b'm', b'U', b'S', b'D', b'Q', b'X', b'9', b'R', b'd', b'o', b'Z', b'f']

BASE = 58
BV_LEN = 12
PREFIX = "BV1"

def av2bv(aid: int) -> str:
    bytes = [b'B', b'V', b'1', b'0', b'0', b'0', b'0', b'0', b'0', b'0', b'0', b'0']
    bv_idx = BV_LEN - 1
    tmp = (MAX_AID | aid) ^ XOR_CODE
    while int(tmp) != 0:
        bytes[bv_idx] = data[int(tmp % BASE)]
        tmp /= BASE
        bv_idx -= 1
    bytes[3], bytes[9] = bytes[9], bytes[3]
    bytes[4], bytes[7] = bytes[7], bytes[4]
    return "".join([i.decode() for i in bytes])

def bv2av(bvid: str) -> int:
    bvid = list(bvid)
    bvid[3], bvid[9] = bvid[9], bvid[3]
    bvid[4], bvid[7] = bvid[7], bvid[4]
    bvid = bvid[3:]
    tmp = 0
    for i in bvid:
        idx = data.index(i.encode())
        tmp = tmp * BASE + idx
    return (tmp & MASK_CODE) ^ XOR_CODE

print(av2bv(111298867365120))
print(bv2av("BV1L9Uoa9EUx"))

Rust

参考 https://github.com/Colerar/abv/blob/main/src/lib.rs

Swift

fileprivate let XOR_CODE: UInt64 = 23442827791579
fileprivate let MASK_CODE: UInt64 = 2251799813685247
fileprivate let MAX_AID: UInt64 = 1 << 51

fileprivate let data: [UInt8] = [70, 99, 119, 65, 80, 78, 75, 84, 77, 117, 103, 51, 71, 86, 53, 76, 106, 55, 69, 74, 110, 72, 112, 87, 115, 120, 52, 116, 98, 56, 104, 97, 89, 101, 118, 105, 113, 66, 122, 54, 114, 107, 67, 121, 49, 50, 109, 85, 83, 68, 81, 88, 57, 82, 100, 111, 90, 102]

fileprivate let BASE: UInt64 = 58
fileprivate let BV_LEN: Int = 12
fileprivate let PREFIX: String = "BV1"

func av2bv(avid: UInt64) -> String {
    var bytes: [UInt8] = [66, 86, 49, 48, 48, 48, 48, 48, 48, 48, 48, 48]
    var bvIdx = BV_LEN - 1
    var tmp = (MAX_AID | avid) ^ XOR_CODE
    
    while tmp != 0 {
        bytes[bvIdx] = data[Int(tmp % BASE)]
        tmp /= BASE
        bvIdx -= 1
    }
    
    bytes.swapAt(3, 9)
    bytes.swapAt(4, 7)
    
    return String(decoding: bytes, as: UTF8.self)
}

func bv2av(bvid: String) -> UInt64 {
    let fixedBvid: String
    if bvid.hasPrefix("BV") {
        fixedBvid = bvid
    } else {
        fixedBvid = "BV" + bvid
    }
    var bvidArray = Array(fixedBvid.utf8)
    
    bvidArray.swapAt(3, 9)
    bvidArray.swapAt(4, 7)
    
    let trimmedBvid = String(decoding: bvidArray[3...], as: UTF8.self)
    
    var tmp: UInt64 = 0
    
    for char in trimmedBvid {
        if let idx = data.firstIndex(of: char.utf8.first!) {
            tmp = tmp * BASE + UInt64(idx)
        }
    }
    
    return (tmp & MASK_CODE) ^ XOR_CODE
}

print(av2bv(avid: 111298867365120))
print(bv2av(bvid: "BV1L9Uoa9EUx"))

老版算法存档

算法参考自【揭秘】av号转bv号的过程

av->bv算法

注：本算法及示例程序仅能编解码avid < 29460791296，且暂无法验证avid >= 29460791296的正确性再注：本人不清楚新算法能否编解码avid >= 29460791296

a = (avid ⊕ 177451812) + 100618342136696320
以 i 为循环变量循环 6 次 b[i] = (a / 58 ^ i) % 58
将 b[i] 中各个数字转换为以下码表中的字符

码表：

fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF

初始化字符串 b[i]=
按照以下字符顺序编码表编码并填充至 b[i]

字符顺序编码表：

0 -> 9

1 -> 8

2 -> 1

3 -> 6

4 -> 2

5 -> 4

6 -> 0

7 -> 7

8 -> 3

9 -> 5

bv->av算法

为以上算法的逆运算

编程实现

使用 Python C TypeScript Java Kotlin Golang Rust 等语言作为示例，欢迎社区提交更多例程

注: 新算法只提供了 Python 和 Rust 版本

Python

XOR = 177451812
ADD = 100618342136696320
TABLE = "fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF"
MAP = 9, 8, 1, 6, 2, 4, 0, 7, 3, 5


def av2bv(av: int) -> str:
    av = (av ^ XOR) + ADD
    bv = [""] * 10
    for i in range(10):
        bv[MAP[i]] = TABLE[(av // 58**i) % 58]
    return "".join(bv)


def bv2av(bv: int) -> int:
    av = [""] * 10
    s = 0
    for i in range(10):
        s += TABLE.find(bv[MAP[i]]) * 58**i
    av = (s - ADD) ^ XOR

    return av


def main():
    while 1:
        mode = input("1. AV to BV\n2. BV to AV\n3. Exit\n你的选择：")
        if mode == "1":
            print(f"BV号是：BV {av2bv(int(input('AV号是：')))}")
        elif mode == "2":
            print(f"AV号是：AV {bv2av(input('BV号是：'))}")
        elif mode == "3":
            break
        else:
            print("输入错误请重新输入")


if __name__ == "__main__":
    main()

C

#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <string.h>

const char table[] = "fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF"; // 码表
char tr[124]; // 反查码表
const unsigned long long XOR = 177451812; // 固定异或值
const unsigned long long ADD = 8728348608; // 固定加法值
const int s[] = {11, 10, 3, 8, 4, 6}; // 位置编码表

// 初始化反查码表
void tr_init() {
	for (int i = 0; i < 58; i++)
		tr[table[i]] = i;
}

unsigned long long bv2av(char bv[]) {
	unsigned long long r = 0;
	unsigned long long av;
	for (int i = 0; i < 6; i++)
		r += tr[bv[s[i]]] * (unsigned long long)pow(58, i);
	av = (r - ADD) ^ XOR;
	return av;
}

char *av2bv(unsigned long long av) {
	char *result = (char*)malloc(13);
	strcpy(result,"BV1  4 1 7  ");
	av = (av ^ XOR) + ADD;
	for (int i = 0; i < 6; i++)
		result[s[i]] = table[(unsigned long long)(av / (unsigned long long)pow(58, i)) % 58];
	return result;
}

int main() {
	tr_init();
	printf("%s\n", av2bv(170001));
	printf("%u\n", bv2av("BV17x411w7KC"));
	return 0;
}

输出为：

BV17x411w7KC
170001

TypeScript

感谢#417提供

export default class BvCode {
  private TABEL = 'fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF'; // 码表
  private TR: Record<string, number> = {}; // 反查码表
  private S = [11, 10, 3, 8, 4, 6]; // 位置编码表
  private XOR = 177451812; // 固定异或值
  private ADD = 8728348608; // 固定加法值
  constructor() {
    // 初始化反查码表
    const len = this.TABEL.length;
    for (let i = 0; i < len; i++) {
      this.TR[this.TABEL[i]] = i;
    }
  }
  av2bv(av: number): string {
    const x_ = (av ^ this.XOR) + this.ADD;
    const r = ['B', 'V', '1', , , '4', , '1', , '7'];
    for (let i = 0; i < 6; i++) {
      r[this.S[i]] = this.TABEL[Math.floor(x_ / 58 ** i) % 58];
    }
    return r.join('');
  }
  bv2av(bv: string): number {
    let r = 0;
    for (let i = 0; i < 6; i++) {
      r += this.TR[bv[this.S[i]]] * 58 ** i;
    }
    return (r - this.ADD) ^ this.XOR;
  }
}

const bvcode = new BvCode();

console.log(bvcode.av2bv(170001));
console.log(bvcode.bv2av('BV17x411w7KC'));

输出为：

BV17x411w7KC
170001

Java

/**
 * 算法来自：https://www.zhihu.com/question/381784377/answer/1099438784
 */
public class Util {
    private static final String TABLE = "fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF";
    private static final int[] S = new int[]{11, 10, 3, 8, 4, 6};
    private static final int XOR = 177451812;
    private static final long ADD = 8728348608L;
    private static final Map<Character, Integer> MAP = new HashMap<>();

    static {
        for (int i = 0; i < 58; i++) {
            MAP.put(TABLE.charAt(i), i);
        }
    }

    public static String aidToBvid(int aid) {
        long x = (aid ^ XOR) + ADD;
        char[] chars = new char[]{'B', 'V', '1', ' ', ' ', '4', ' ', '1', ' ', '7', ' ', ' '};
        for (int i = 0; i < 6; i++) {
            int pow = (int) Math.pow(58, i);
            long i1 = x / pow;
            int index = (int) (i1 % 58);
            chars[S[i]] = TABLE.charAt(index);
        }
        return String.valueOf(chars);
    }

    public static int bvidToAid(String bvid) {
        long r = 0;
        for (int i = 0; i < 6; i++) {
            r += MAP.get(bvid.charAt(S[i])) * Math.pow(58, i);
        }
        return (int) ((r - ADD) ^ XOR);
    }
}

Kotlin

/**
 * 此程序非完全原创，改编自GH站内某大佬的Java程序，修改了部分代码，且转换为Kotlin
 * 算法来源同上
 */
object VideoUtils {
    //这里是由知乎大佬不知道用什么方法得出的转换用数字
    var ss = intArrayOf(11, 10, 3, 8, 4, 6, 2, 9, 5, 7)
    var xor: Long = 177451812 //二进制时加减数1

    var add = 8728348608L //十进制时加减数2

    //变量初始化工作，加载哈希表
    private const val table = "fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF"
    private val mp = HashMap<String, Int>()
    private val mp2 = HashMap<Int, String>()

    //现在，定义av号和bv号互转的方法
//定义一个power乘方方法，这是转换进制必要的
    fun power(a: Int, b: Int): Long {
        var power: Long = 1
        for (c in 0 until b) power *= a.toLong()
        return power
    }

    //bv转av方法
    fun bv2av(s: String): String {
        var r: Long = 0
        //58进制转换
        for (i in 0..57) {
            val s1 = table.substring(i, i + 1)
            mp[s1] = i
        }
        for (i in 0..5) {
            r += mp[s.substring(ss[i], ss[i] + 1)]!! * power(58, i)
        }
        //转换完成后，需要处理，带上两个随机数
        return (r - add xor xor).toString()
    }

    //av转bv方法
    fun av2bv(st: String): String {
        try {
            var s = java.lang.Long.valueOf(st.split("av".toRegex()).dropLastWhile { it.isEmpty() }
                .toTypedArray()[1])
            val sb = StringBuffer("BV1  4 1 7  ")
            //逆向思路，先将随机数还原
            s = (s xor xor) + add
            //58进制转回
            for (i in 0..57) {
                val s1 = table.substring(i, i + 1)
                mp2[i] = s1
            }
            for (i in 0..5) {
                val r = mp2[(s / power(58, i) % 58).toInt()]
                sb.replace(ss[i], ss[i] + 1, r!!)
            }
            return sb.toString()
        } catch (e: ArrayIndexOutOfBoundsException) {
            return ""
        }
    }

}

Golang

package main

import "math"

const TABLE = "fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF"
var S = [11]uint{11, 10, 3, 8, 4, 6}
const XOR = 177451812
const ADD = 8728348608

var TR = map[string]int64{}

// 初始化 TR
func init() {
	for i := 0; i < 58; i++ {
		TR[TABLE[i:i+1]] = int64(i)
	}
}

func BV2AV(bv string) int64 {
	r := int64(0)
	for i := 0; i < 6; i++ {
		r += TR[bv[S[i]:S[i]+1]] * int64(math.Pow(58, float64(i)))
	}
	return (r - ADD) ^ XOR
}

func AV2BV(av int64) string {
	x := (av ^ XOR) + ADD
	r := []rune("BV1  4 1 7  ")
	for i := 0; i < 6; i++ {
		r[S[i]] = rune(TABLE[x/int64(math.Pow(58, float64(i)))%58])
	}
	return string(r)
}

func main() {
	println(AV2BV(170001))
	println(BV2AV("BV17x411w7KC"))
}

输出为：

BV17x411w7KC
170001

Rust

crate: https://github.com/stackinspector/bvid

// Copyright (c) 2023 stackinspector. MIT license.

const XORN: u64 = 177451812;
const ADDN: u64 = 100618342136696320;
const TABLE: [u8; 58] = *b"fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF";
const MAP: [usize; 10] = [9, 8, 1, 6, 2, 4, 0, 7, 3, 5];
const REV_TABLE: [u8; 74] = [
    13, 12, 46, 31, 43, 18, 40, 28,  5,  0,  0,  0,  0,  0,  0,  0, 54, 20, 15, 8,
    39, 57, 45, 36,  0, 38, 51, 42, 49, 52,  0, 53,  7,  4,  9, 50, 10, 44, 34, 6,
    25,  1,  0,  0,  0,  0,  0,  0, 26, 29, 56,  3, 24,  0, 47, 27, 22, 41, 16, 0,
    11, 37,  2, 35, 21, 17, 33, 30, 48, 23, 55, 32, 14, 19,
];
const POW58: [u64; 10] = [
    1, 58, 3364, 195112, 11316496, 656356768, 38068692544,
    2207984167552, 128063081718016, 7427658739644928,
];

fn av2bv(avid: u64) -> [u8; 10] {
    let a = (avid ^ XORN) + ADDN;
    let mut bvid = [0; 10];
    for i in 0..10 {
        bvid[MAP[i]] = TABLE[(a / POW58[i]) as usize % 58];
    }
    bvid
}

fn bv2av(bvid: [u8; 10]) -> u64 {
    let mut a = 0;
    for i in 0..10 {
        a += REV_TABLE[bvid[MAP[i]] as usize - 49] as u64 * POW58[i];
    }
    (a - ADDN) ^ XORN
}

// assert_eq!(*b"17x411w7KC", av2bv(170001));
// assert_eq!(170001, bv2av(*b"17x411w7KC"));

18 KiB Raw Blame History Unescape Escape

bvid说明

概述

格式

实质

avid发号方式的变化

算法概述

av->bv算法

bv->av算法

编程实现

JavaScript/TypeScript

Python

Rust

Swift

老版算法存档

av->bv算法

bv->av算法

编程实现

Python

C

TypeScript

Java

Kotlin

Golang

Rust

18 KiB

Raw Blame History