我怎样写在C或C ++一个简单的正前pression模式匹配功能? [英] How do I write a simple regular expression pattern matching function in C or C++?
本文介绍了我怎样写在C或C ++一个简单的正前pression模式匹配功能?的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!
问题描述
这是我今天的笔试问题,函数签名
This is a question in my paper test today, the function signature is
int is_match(char* pattern,char* string)
该模式仅限于ASCII字符和定量 *
和?
,所以它是比较简单的。 is_match
如果匹配应该返回1,否则为0。
The pattern is limited to only ASCII chars and the quantification *
and ?
, so it is relatively simple. is_match
should return 1 if matched, otherwise 0.
我如何做到这一点?
推荐答案
下面是递归扩展实现。测试图案复杂的一阶。
Here is recursive extendable implementation. Tested for first order of pattern complexity.
#include <string.h>
#include <string>
#include <vector>
#include <iostream>
struct Match {
Match():_next(0) {}
virtual bool match(const char * pattern, const char * input) const {
return !std::strcmp(pattern, input);
}
bool next(const char * pattern, const char * input) const {
if (!_next) return false;
return _next->match(pattern, input);
}
const Match * _next;
};
class MatchSet: public Match {
typedef std::vector<Match *> Set;
Set toTry;
public:
virtual bool match(const char * pattern, const char * input) const {
for (Set::const_iterator i = toTry.begin(); i !=toTry.end(); ++i) {
if ((*i)->match(pattern, input)) return true;
}
return false;
}
void add(Match * m) {
toTry.push_back(m);
m->_next = this;
}
~MatchSet() {
for (Set::const_iterator i = toTry.begin(); i !=toTry.end(); ++i)
if ((*i)->_next==this) (*i)->_next = 0;
}
};
struct MatchQuestion: public Match {
virtual bool match(const char * pattern, const char * input) const {
if (pattern[0] != '?')
return false;
if (next(pattern+1, input))
return true;
if (next(pattern+1, input+1))
return true;
return false;
}
};
struct MatchEmpty: public Match {
virtual bool match(const char * pattern, const char * input) const {
if (pattern[0]==0 && input[0]==0)
return true;
return false;
}
};
struct MatchAsterisk: public Match {
virtual bool match(const char * pattern, const char * input) const {
if (pattern[0] != '*')
return false;
if (pattern[1] == 0) {
return true;
}
for (int i = 0; input[i] != 0; ++i) {
if (next(pattern+1, input+i))
return true;
}
return false;
}
};
struct MatchSymbol: public Match {
virtual bool match(const char * pattern, const char * input) const {
// TODO: consider cycle here to prevent unnecessary recursion
// Cycle should detect special characters and call next on them
// Current implementation abstracts from that
if (pattern[0] != input[0])
return false;
return next(pattern+1, input+1);
}
};
class DefaultMatch: public MatchSet {
MatchEmpty empty;
MatchQuestion question;
MatchAsterisk asterisk;
MatchSymbol symbol;
public:
DefaultMatch() {
add(&empty);
add(&question);
add(&asterisk);
add(&symbol);
}
void test(const char * p, const char * input) const {
testOneWay(p, input);
if (!std::strcmp(p, input)) return;
testOneWay(input, p);
}
bool testOneWay(const char * p, const char * input) const {
const char * eqStr = " == ";
bool rv = match(p, input);
if (!rv) eqStr = " != ";
std::cout << p << eqStr << input << std::endl;
return rv;
}
};
int _tmain(int argc, _TCHAR* argv[])
{
using namespace std;
typedef vector<string> Strings;
Strings patterns;
patterns.push_back("*");
patterns.push_back("*hw");
patterns.push_back("h*w");
patterns.push_back("hw*");
patterns.push_back("?");
patterns.push_back("?ab");
patterns.push_back("a?b");
patterns.push_back("ab?");
patterns.push_back("c");
patterns.push_back("cab");
patterns.push_back("acb");
patterns.push_back("abc");
patterns.push_back("*this homework?");
patterns.push_back("Is this homework?");
patterns.push_back("This is homework!");
patterns.push_back("How is this homework?");
patterns.push_back("hw");
patterns.push_back("homework");
patterns.push_back("howork");
DefaultMatch d;
for (unsigned i = 0; i < patterns.size(); ++i)
for (unsigned j =i; j < patterns.size(); ++j)
d.test(patterns[i].c_str(), patterns[j].c_str());
return 0;
}
如果有任何不明白,问。
If something is unclear, ask.
这篇关于我怎样写在C或C ++一个简单的正前pression模式匹配功能?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!
查看全文