如何拼接两个字符串并省略首尾重复的连续语句段？

问题场景 #

开发语音转文本模块时，需要合并两段文本：从第一段文本的末尾向前查找，找到与第二段文本开头重复的语句段，最终合并为无重复的完整语句。例如：

const str1 = "It was great, I finally had some time to"
const str2 = "had some time to relax and catch up on my reading."
// 期望结果："It was great, I finally had some time to relax and catch up on my reading."

原代码的问题 #

你提供的formPhrase函数通过去重所有单词来实现合并，这种逻辑只适用于重复内容仅出现在str1末尾和str2开头的场景。当重复内容同时存在于文本的其他位置时就会失效，比如：

const str3 = "this code transforms this";
const str4 = "transforms this string";
// 原函数输出："this code transforms string"
// 正确期望输出："this code transforms this string"

原函数会错误地移除str1中中间出现的重复单词，不符合“仅合并末尾-开头重复”的需求。

正确实现方案 #

我们需要找到str1后缀与str2前缀的最长匹配片段，再进行合并，确保只去除末尾和开头的重复部分：

function mergeWithOverlap(str1, str2) {
    const words1 = str1.split(' ');
    const words2 = str2.split(' ');
    const maxPossibleOverlap = Math.min(words1.length, words2.length);

    // 从最长的可能重叠长度开始检查，确保找到的是最长匹配
    for (let overlapLength = maxPossibleOverlap; overlapLength > 0; overlapLength--) {
        const str1Suffix = words1.slice(-overlapLength).join(' ');
        const str2Prefix = words2.slice(0, overlapLength).join(' ');
        
        if (str1Suffix === str2Prefix) {
            // 合并：str1完整内容 + str2去掉重叠前缀的部分
            return [...words1, ...words2.slice(overlapLength)].join(' ');
        }
    }

    // 无重叠时直接拼接两段文本
    return `${str1} ${str2}`;
}

// 测试示例1
const str1 = "It was great, I finally had some time to";
const str2 = "had some time to relax and catch up on my reading.";
console.log(mergeWithOverlap(str1, str2));
// 输出：It was great, I finally had some time to relax and catch up on my reading.

// 测试示例2
const str3 = "this code transforms this";
const str4 = "transforms this string";
console.log(mergeWithOverlap(str3, str4));
// 输出：this code transforms this string

逻辑说明 #

1. 将两段文本拆分为单词数组，便于按片段检查；
2. 从最长的可能重叠长度（两段文本单词数的较小值）开始遍历，确保找到的是最长的匹配重叠段；
3. 找到匹配的后缀和前缀后，合并两段文本（str1完整保留，str2去掉重叠的前缀部分）；
4. 若没有任何重叠，直接拼接两段文本。

内容的提问来源于stack exchange，提问作者Maurício Giordano