C++ Boost Spirit解析器可选表达式求值

C++ Boost Spirit解析器可选表达式求值,c++,parsing,boost-spirit,boost-spirit-qi,C++,Parsing,Boost Spirit,Boost Spirit Qi,我正在尝试解析文本文件中的一行,其形式如下: [int_:] [int_/int_] [(int_, string)] string [string:int_]... 其中,[]是可选参数,但将包含(”:“,”(“,”),“/”)等标记。 最后一种格式是重复格式“key:value”组合。e、 g: 10: 0x1/2 (8, INC) rd API:2 SI:100 当所有参数都可用时,我能够解析整行。 但如果缺少任何起始可选参数,则解析器将失败 如何忽略Boost Spirit库中的可选

我正在尝试解析文本文件中的一行,其形式如下:

[int_:] [int_/int_] [(int_, string)] string [string:int_]...
其中,
[]
是可选参数,但将包含(
”:“
”(“
”),
“/”
)等标记。 最后一种格式是重复格式
“key:value”
组合。e、 g:

10: 0x1/2 (8, INC) rd API:2 SI:100
当所有参数都可用时,我能够解析整行。 但如果缺少任何起始可选参数,则解析器将失败

如何忽略Boost Spirit库中的可选参数?(即跳过将可选变量指定为默认值。)

以下是qi语法规则:

quoted_string = lexeme[+(char_ -(lit(' ') | lit(')')))];
hex_num = ((lit("0x") | lit("0X")) >> hex) | uint_;

start = (hex_num >> lit(":"))
    >> (hex_num >> lit("/") >> hex_num )
    >> lit("(") >> hex_num >> lit(",") >> quoted_string >> lit(")")
    >> quoted_string
    >> quoted_string;

qi::rule<Iterator, std::string(), ascii::space_type> quoted_string;
qi::rule<Iterator, uint32_t(), ascii::space_type> hex_num;
qi::rule<Iterator, employee(), ascii::space_type> start;
quoted_string=lexeme[+(char-(lit(“”)| lit('))];
十六进制数=((点亮(“0x”)|点亮(“0x”)>>十六进制);
开始=(十六进制数>>点亮(“:”)
>>(十六进制数>>亮(“/”>>十六进制数)
>>亮(“”>>十六进制数>>亮(“”,“”>>带引号的字符串>>亮(“”)
>>带引号的字符串
>>带引号的字符串;
qi::规则引用的字符串;
qi::规则十六进制数;
qi::规则开始;

为AST节点建模以反映解析器树:

struct ratio_t { uint32_t a,b; };
struct opcode_t { uint32_t id; std::string name; };

struct Node {
    uint32_t label; // prefix:

    boost::optional<ratio_t> ratio; // a/b
    boost::optional<opcode_t> opcode; // (id, name)

    std::string extra;
    std::multimap<std::string, uint32_t> params;
};
您可以使用类似的解析树简单地解析到其中:

    // lexemes
    unquoted_string = +(graph - ')');
    num = (no_case[ "0x" ] >> hex) | uint_;
    param = +(graph - ':') >> ':' >> num;

    // skipping productions
    opcode = '(' >> num >> ',' >> unquoted_string >> ')';
    ratio  = num >> '/' >> num;
    prefix = (num >> ':') | attr(0);                      // defaults to 0
    start  = prefix
            >> -ratio
            >> -opcode
            >> unquoted_string
            >> *param;
现在,当您分析这些测试用例时:

for (std::string const input : {
        "10: 0x1/2 (8, INC) rd API:2 SI:100",
        "10: 0x1/2 (8, INC) rd API:2",
        "10: 0x1/2 (8, INC) rd",
        "10: 0x1/2 rd API:2 SI:100",
        "10: rd API:2 SI:100",
        "0x1/2 rd API:2 SI:100",
        "rd API:2 SI:100",
    })
{
    It f = input.begin(), l = input.end();
    AST::Node data;
    bool ok = qi::phrase_parse(f, l, p, qi::ascii::space, data);
    if (ok) {
        std::cout << "Parse success: " << data << "\n";
    }
    else {
        std::cout << "Parse failure ('" <<  input << "')\n";
    }

    if (f!=l) {
        std::cout << "Remaining unparsed input: '" << std::string(f,l) << "'\n";
    }
}
完整演示

#include <boost/spirit/include/qi.hpp>
#include <boost/fusion/adapted.hpp>
#include <boost/optional/optional_io.hpp>
#include <map>

namespace qi = boost::spirit::qi;

namespace AST {
    struct ratio_t  { uint32_t a,b; };
    struct opcode_t { uint32_t id; std::string name; };

    struct Node {
        uint32_t label; // prefix:

        boost::optional<ratio_t> ratio; // a/b
        boost::optional<opcode_t> opcode; // (id, name)

        std::string extra;
        std::multimap<std::string, uint32_t> params;
    };

    std::ostream& operator<<(std::ostream& os, ratio_t const& v) {
        return os << v.a << "/" << v.b;
    }
    std::ostream& operator<<(std::ostream& os, opcode_t const& v) {
        return os << "(" << v.id << ", '" << v.name << "')";
    }
    std::ostream& operator<<(std::ostream& os, Node const& v) {
        os << v.label << ": " << v.ratio << " " << v.opcode << " " << v.extra;
        for (auto& p : v.params) os << " " << p.first << ":" << p.second;
        return os;
    }
}

BOOST_FUSION_ADAPT_STRUCT(AST::ratio_t, a, b)
BOOST_FUSION_ADAPT_STRUCT(AST::opcode_t, id, name)
BOOST_FUSION_ADAPT_STRUCT(AST::Node, label, ratio, opcode, extra, params)

template <typename It, typename Skipper = qi::ascii::space_type>
struct P : qi::grammar<It, AST::Node(), Skipper> {
    P() : P::base_type(start) 
    {
        using namespace qi;

        // lexemes
        unquoted_string = +(graph - ')');
        num = (no_case[ "0x" ] >> hex) | uint_;
        param = +(graph - ':') >> ':' >> num;

        // skipping productions
        opcode = '(' >> num >> ',' >> unquoted_string >> ')';
        ratio  = num >> '/' >> num;
        prefix = (num >> ':') | attr(0);                      // defaults to 0
        start  = prefix
                >> -ratio
                >> -opcode
                >> unquoted_string
                >> *param;

        BOOST_SPIRIT_DEBUG_NODES((start)(unquoted_string)(num)(prefix)(ratio)(opcode)(param))
    }

  private:
    qi::rule<It, AST::ratio_t(),  Skipper> ratio;
    qi::rule<It, AST::opcode_t(), Skipper> opcode;
    qi::rule<It, AST::Node(),     Skipper> start;
    qi::rule<It, uint32_t(),      Skipper> prefix;

    //lexemes
    qi::rule<It, std::string()> unquoted_string;
    qi::rule<It, uint32_t()> num;
    qi::rule<It, std::pair<std::string, uint32_t>> param;
};

int main() {
    using It = std::string::const_iterator;
    P<It> const p;

    for (std::string const input : {
            "10: 0x1/2 (8, INC) rd API:2 SI:100",
            "10: 0x1/2 (8, INC) rd API:2",
            "10: 0x1/2 (8, INC) rd",
            "10: 0x1/2 rd API:2 SI:100",
            "10: rd API:2 SI:100",
            "0x1/2 rd API:2 SI:100",
            "rd API:2 SI:100",
        })
    {
        It f = input.begin(), l = input.end();
        AST::Node data;
        bool ok = qi::phrase_parse(f, l, p, qi::ascii::space, data);
        if (ok) {
            std::cout << "Parse success: " << data << "\n";
        }
        else {
            std::cout << "Parse failure ('" <<  input << "')\n";
        }

        if (f!=l) {
            std::cout << "Remaining unparsed input: '" << std::string(f,l) << "'\n";
        }
    }
}
#包括
#包括
#包括
#包括
名称空间qi=boost::spirit::qi;
名称空间AST{
结构比率{uint32\uta,b;};
结构操作码{uint32\ut id;std::string name;};
结构节点{
uint32\u t标签;//前缀:
boost::可选比率;//a/b
boost::可选操作码;//(id、名称)
std::字符串额外;
std::多映射参数;
};
std::ostream&operator>unquoted_字符串
>>*param;
BOOST_SPIRIT_DEBUG_节点((开始)(无引号的字符串)(num)(前缀)(比率)(操作码)(参数))
}
私人:
齐:规则比率;
qi::规则操作码;
qi::规则开始;
qi::规则前缀;
//词素
qi::规则不带引号的字符串;
qi::规则数;
qi::规则参数;
};
int main(){
使用它=std::string::const_迭代器;
常数P;
对于(标准::字符串常量输入:{
“10:0x1/2(8,INC)rd API:2 SI:100”,
“10:0x1/2(8,INC)rd API:2”,
“10:0x1/2(8,INC)rd”,
“10:0x1/2第三方API:2 SI:100”,
“10:rd API:2 SI:100”,
“0x1/2 rd API:2 SI:100”,
“研发API:2 SI:100”,
})
{
它f=input.begin(),l=input.end();
AST::节点数据;
bool ok=qi::短语解析(f,l,p,qi::ascii::空间,数据);
如果(确定){

std::cout为AST节点建模以反映解析器树:

struct ratio_t { uint32_t a,b; };
struct opcode_t { uint32_t id; std::string name; };

struct Node {
    uint32_t label; // prefix:

    boost::optional<ratio_t> ratio; // a/b
    boost::optional<opcode_t> opcode; // (id, name)

    std::string extra;
    std::multimap<std::string, uint32_t> params;
};
您可以使用类似的解析树简单地解析到其中:

    // lexemes
    unquoted_string = +(graph - ')');
    num = (no_case[ "0x" ] >> hex) | uint_;
    param = +(graph - ':') >> ':' >> num;

    // skipping productions
    opcode = '(' >> num >> ',' >> unquoted_string >> ')';
    ratio  = num >> '/' >> num;
    prefix = (num >> ':') | attr(0);                      // defaults to 0
    start  = prefix
            >> -ratio
            >> -opcode
            >> unquoted_string
            >> *param;
现在,当您分析这些测试用例时:

for (std::string const input : {
        "10: 0x1/2 (8, INC) rd API:2 SI:100",
        "10: 0x1/2 (8, INC) rd API:2",
        "10: 0x1/2 (8, INC) rd",
        "10: 0x1/2 rd API:2 SI:100",
        "10: rd API:2 SI:100",
        "0x1/2 rd API:2 SI:100",
        "rd API:2 SI:100",
    })
{
    It f = input.begin(), l = input.end();
    AST::Node data;
    bool ok = qi::phrase_parse(f, l, p, qi::ascii::space, data);
    if (ok) {
        std::cout << "Parse success: " << data << "\n";
    }
    else {
        std::cout << "Parse failure ('" <<  input << "')\n";
    }

    if (f!=l) {
        std::cout << "Remaining unparsed input: '" << std::string(f,l) << "'\n";
    }
}
完整演示

#include <boost/spirit/include/qi.hpp>
#include <boost/fusion/adapted.hpp>
#include <boost/optional/optional_io.hpp>
#include <map>

namespace qi = boost::spirit::qi;

namespace AST {
    struct ratio_t  { uint32_t a,b; };
    struct opcode_t { uint32_t id; std::string name; };

    struct Node {
        uint32_t label; // prefix:

        boost::optional<ratio_t> ratio; // a/b
        boost::optional<opcode_t> opcode; // (id, name)

        std::string extra;
        std::multimap<std::string, uint32_t> params;
    };

    std::ostream& operator<<(std::ostream& os, ratio_t const& v) {
        return os << v.a << "/" << v.b;
    }
    std::ostream& operator<<(std::ostream& os, opcode_t const& v) {
        return os << "(" << v.id << ", '" << v.name << "')";
    }
    std::ostream& operator<<(std::ostream& os, Node const& v) {
        os << v.label << ": " << v.ratio << " " << v.opcode << " " << v.extra;
        for (auto& p : v.params) os << " " << p.first << ":" << p.second;
        return os;
    }
}

BOOST_FUSION_ADAPT_STRUCT(AST::ratio_t, a, b)
BOOST_FUSION_ADAPT_STRUCT(AST::opcode_t, id, name)
BOOST_FUSION_ADAPT_STRUCT(AST::Node, label, ratio, opcode, extra, params)

template <typename It, typename Skipper = qi::ascii::space_type>
struct P : qi::grammar<It, AST::Node(), Skipper> {
    P() : P::base_type(start) 
    {
        using namespace qi;

        // lexemes
        unquoted_string = +(graph - ')');
        num = (no_case[ "0x" ] >> hex) | uint_;
        param = +(graph - ':') >> ':' >> num;

        // skipping productions
        opcode = '(' >> num >> ',' >> unquoted_string >> ')';
        ratio  = num >> '/' >> num;
        prefix = (num >> ':') | attr(0);                      // defaults to 0
        start  = prefix
                >> -ratio
                >> -opcode
                >> unquoted_string
                >> *param;

        BOOST_SPIRIT_DEBUG_NODES((start)(unquoted_string)(num)(prefix)(ratio)(opcode)(param))
    }

  private:
    qi::rule<It, AST::ratio_t(),  Skipper> ratio;
    qi::rule<It, AST::opcode_t(), Skipper> opcode;
    qi::rule<It, AST::Node(),     Skipper> start;
    qi::rule<It, uint32_t(),      Skipper> prefix;

    //lexemes
    qi::rule<It, std::string()> unquoted_string;
    qi::rule<It, uint32_t()> num;
    qi::rule<It, std::pair<std::string, uint32_t>> param;
};

int main() {
    using It = std::string::const_iterator;
    P<It> const p;

    for (std::string const input : {
            "10: 0x1/2 (8, INC) rd API:2 SI:100",
            "10: 0x1/2 (8, INC) rd API:2",
            "10: 0x1/2 (8, INC) rd",
            "10: 0x1/2 rd API:2 SI:100",
            "10: rd API:2 SI:100",
            "0x1/2 rd API:2 SI:100",
            "rd API:2 SI:100",
        })
    {
        It f = input.begin(), l = input.end();
        AST::Node data;
        bool ok = qi::phrase_parse(f, l, p, qi::ascii::space, data);
        if (ok) {
            std::cout << "Parse success: " << data << "\n";
        }
        else {
            std::cout << "Parse failure ('" <<  input << "')\n";
        }

        if (f!=l) {
            std::cout << "Remaining unparsed input: '" << std::string(f,l) << "'\n";
        }
    }
}
#包括
#包括
#包括
#包括
名称空间qi=boost::spirit::qi;
名称空间AST{
结构比率{uint32\uta,b;};
结构操作码{uint32\ut id;std::string name;};
结构节点{
uint32\u t标签;//前缀:
boost::可选比率;//a/b
boost::可选操作码;//(id、名称)
std::字符串额外;
std::多映射参数;
};
std::ostream&operator>unquoted_字符串
>>*param;
BOOST_SPIRIT_DEBUG_节点((开始)(无引号的字符串)(num)(前缀)(比率)(操作码)(参数))
}
私人:
齐:规则比率;
qi::规则操作码;
qi::规则开始;
qi::规则前缀;
//词素
qi::规则不带引号的字符串;
qi::规则数;
qi::规则参数;
};
int main(){
使用它=std::string::const_迭代器;
常数P;
对于(标准::字符串常量输入:{
“10:0x1/2(8,INC)rd API:2 SI:100”,
“10:0x1/2(8,INC)rd API:2”,
“10:0x1/2(8,INC)rd”,
“10:0x1/2第三方API:2 SI:100”,
“10:rd API:2 SI:100”,
“0x1/2 rd API:2 SI:100”,
“研发API:2 SI:100”,
})
{
它f=input.begin(),l=input.end();
AST::节点数据;
bool ok=qi::短语解析(f,l,p,qi::ascii::空间,数据);
如果(确定){

std::我不能谢谢你@sehe它帮了我很多。我没有用正确的方式使用它。谢谢你的帮助。谢谢@sehe它帮了我很多。我没有用正确的方式使用它。谢谢你的帮助。