Question

好奇这是否可以简化......

internal static IEnumerable<string> Split(string str, char sep = ',')
{
    int lastIndex = 0;
    bool quoted = false;
    bool escaped = false;
    bool bracketed = false;
    char lastQuote = '\0';

    for (int i = 0; i < str.Length; ++i)
    {
        if (str[i] == '[')
        {
            if (!quoted && !escaped)
                bracketed = true;
            escaped = false;
        }
        else if (str[i] == ']')
        {
            if (!quoted && !escaped)
                bracketed = false;
            escaped = false;
        }
        else if (str[i] == '\\')
        {
            escaped = !escaped;
        }
        else if (str[i] == '"' || str[i] == '\'')
        {
            if (!escaped)
            {
                if (quoted)
                {
                    if (lastQuote == str[i])
                        quoted = false;
                }
                else
                {
                    quoted = true;
                    lastQuote = str[i];
                }
            }
            escaped = false;
        }
        else if (str[i] == sep)
        {
            if (!quoted && !escaped && !bracketed)
            {
                yield return str.Substring(lastIndex, i - lastIndex);
                lastIndex = i + 1;
            }
            escaped = false;
        }
        else
        {
            escaped = false;
        }
    }

    yield return str.Substring(lastIndex);
}

写这个方法来分割不在[]内的逗号，没有引用，也不会被转义。这本质上是一个棘手的问题，还是我采取了愚蠢的方法？

输入：

foreach(var sel in SharpQuery.SplitCommas("\"comma, in quotes\", comma[in,brackets], comma[in \"quotes, and brackets\"], \"woah, 'nelly,' \\\"now you,re [talking, crazy\\\"\"")) {
    Console.WriteLine(sel);
}

预期产出：

"comma, in quotes"
 comma[in,brackets]
 comma[in "quotes, and brackets"]
 "woah, 'nelly,' \"now you,re [talking, crazy\""

Answer 1

保持自动机状态的一些尴尬选择。在这种情况下，我会使用单个变量或堆栈。因此，您当前的状态始终为stateStack.Peek()。易于阅读。易于处理嵌套状态。

编辑：这是一个快速的样本。我相信你可以扩展它以添加错误处理和规则细节。

    enum ParserState
    {
        Text,
        Bracketed,
        Quoted,
        EscapChar,
    }

    internal static IEnumerable<string> Split(string str, char sep)
    {
        int lastIdx = 0;
        char c;
        ParserState s;
        Stack<ParserState> state = new Stack<ParserState>();
        state.Push(ParserState.Text);

        for (int i = 0; i < str.Length; i++)
        {
            c = str[i];
            s = state.Peek();

            if (s == ParserState.EscapChar
                || (s == ParserState.Bracketed && c == ']')
                || (s == ParserState.Quoted && c == '"'))
            {
                state.Pop();
            }
            else if (c == '[')
                state.Push(ParserState.Bracketed);
            else if (c == '"')
                state.Push(ParserState.Quoted);
            else if (c == '\\')
                state.Push(ParserState.EscapChar);
            else if (s == ParserState.Text && c == sep)
            {
                yield return str.Substring(lastIdx, i - lastIdx);
                lastIdx = i + 1;
            }
        }
        yield return str.Substring(lastIdx);
    }

如何简化这种方法（拆分不带引号，不带括号，未转义的逗号）？

1 个答案: