preprocess tft files

src/parser.cpp

@@ -5,13 +5,30 @@
 #include "utils.hpp"
 
 using std::cerr;
-using std::string_view;
+/* using std::string_view; */
 using std::string;
+using std::vector;
+using std::pair;
 
 namespace /* internal */
 {
-	char* data;
-	char* data_end;				/* past the end pointer */
+	struct alloced
+	{
+		char* const data;
+		char* const end;
+		alloced(U32 len)
+			: data(new char[len])
+			, end(data + len)
+		{
+		}
+		~alloced()
+		{
+			delete[] data;
+		}
+	};
+
+	char* it_data;
+	char* end_data;				/* past the end pointer */
 	unsigned line = 1;
 
 	/* state */
@@ -39,13 +56,13 @@ namespace /* internal */
 
 	void proceedToNameEnd()
 	{
-		while(data < data_end &&
-			  *data != ',' &&
-			  *data != '\n')
+		while(it_data < end_data &&
+			  *it_data != ',' &&
+			  *it_data != '\n')
 		{
-			if(*data == ':' /* && *data == '-' */)
-				unexpected_char(*data);
-			++data;
+			if(*it_data == ':' /* && *it_data == '-' */)
+				unexpected_char(*it_data);
+			++it_data;
 		}
 
 	}
@@ -53,80 +70,196 @@ namespace /* internal */
 	Sex findSex()
 	{
 		Sex res = M;
-		if(*data == ',')
+		if(*it_data == ',')
 		{
-			++data;
-			if(*data == 'M')
+			++it_data;
+			if(*it_data == 'M')
 				res = M;
-			else if(*data == 'F')
+			else if(*it_data == 'F')
 				res = F;
 			else
-				unexpected_char(*data);
+				unexpected_char(*it_data);
 
-			++data;
-			if(*data != '\n')
-				unexpected_char(*data);
+			++it_data;
+			if(*it_data != '\n')
+				unexpected_char(*it_data);
 		}
 		return res;
 	}
 
-	void processAtribute()
+	I32 processDatePart()
 	{
+		I32 sign = 1;
+		I32 val = 0;
+		for(; it_data < end_data; ++it_data)
+		{
+			switch(*it_data)
+			{
+			case '-':
+				sign = -1;
+				break;
+			case '0' ... '9':
+				val = val*10 + *it_data;
+				break;
+			case '.':
+				if(val == 0)
+					unexpected_char('.');
+				else
+					return sign * val;
+				break;
+			case '\n':
+				++line;
+				return sign * val;
+				/* case '?': todo */
+			default:
+				unexpected_char(*it_data);
+			}
+		}
+		unexpected_end();
+	}
 
+	EventTime processDate()		/* todo validate  */
+	{
+		EventTime date;
+		date.year = processDatePart();
+		if(*it_data == '.')
+		{
+			++it_data;
+			date.month = processDatePart();
+
+			if(*it_data == '.')
+			{
+				++it_data;
+				date.day = processDatePart();
+			}
+		}
+		return date;
+	}
+
+	pair<person_id, person_id> processParents()
+	{
+		pair<person_id, person_id> res(Nobody, Nobody);
+		string name;			/* todo dont use string */
+		for(; it_data < end_data; ++it_data)
+		{
+			switch(*it_data)
+			{
+			/* case '-':todo */
+			case ',':
+				if(res.first == Nobody)
+					res.first = cur_tree->findId(name.data());
+				else if(res.second == Nobody)
+					res.second = cur_tree->findId(name.data());
+				else
+					/* todo */exit(1);
+				name.clear();
+				break;
+			case '\n':
+				++line;
+				if(res.first == Nobody)
+					res.first = cur_tree->findId(name.data());
+				else
+					res.second = cur_tree->findId(name.data());
+				return res;
+			default:
+				name.push_back(*it_data);
+			}
+		}
+
+	}
+
+	vector<person_id> parseSpouses()
+	{
+		vector<person_id> res;
+		string name;
+		for(; it_data < end_data; ++it_data)
+		{
+			switch(*it_data)
+			{
+			case ',':
+				res.push_back( cur_tree->findId(name.data()) );
+				name.clear();
+				break;
+			case '\n':
+				++line;
+				res.push_back( cur_tree->findId(name.data()) );
+				return res;
+			default:
+				name.push_back(*it_data);
+			}
+		}
+		/* return res;todo */
 	}
 
 	void state_neutral()
 	{
-		for(; data < data_end; ++data)
+		for(; it_data < end_data; ++it_data)
 		{
-			switch(*data)
+			switch(*it_data)
 			{
 			default:
 				state_person();
-				return;
 				break;
 			case '\n':
 				++line;
 				break;
-			/* case '#': */
-			/* 	state = state_comment; */
-			/* 	return; */
 			case ',':
 			case '-':
 			case 0 ... 9:
 			case ':':
-				unexpected_char(*data);
+				unexpected_char(*it_data);
 			}
 		}
 	}
 
 	void state_person()
 	{
-		char* name_start = data;
-		person_id parent1;
-		person_id parent2;
-		EventTime birth;
-		EventTime death;
+		char* name_start = it_data;
+		pair<person_id, person_id> parents;
+		person_id spouse;
+		EventTime birth{};
+		EventTime death{};
 
 		proceedToNameEnd();
-		/* string_view name(name_start, data); */
-		string name(name_start, data);
+		/* string_view name(name_start, it_data); */
+		string name(name_start, it_data);
 
 		Sex sex = findSex();
 
-		for(; data < data_end; ++data)
+		for(; it_data < end_data; ++it_data)
 		{
-			switch(*data)
+			switch(*it_data)
 			{
 			case '\n':
 				++line;
-				if(data[1] == ':') /* todo */
+				++it_data;
+				if(it_data[1] == ':') /* todo */
 				{
-					processAtribute();
+					switch(*it_data)
+					{
+					case 'b':
+						it_data += 2;
+						birth = processDate();
+						break;
+					case 'd':
+						it_data += 2;
+						death = processDate();
+						break;
+					case 'p':
+						it_data += 2;
+						processParents();
+					case 's':
+						it_data += 2;
+						parseSpouses();
+					default:
+						unexpected_char(it_data[-1]);
+					}
+
 				}
 				else
 				{
-					cur_tree->addPerson(name.c_str(), sex/* , parent1, parent2, birth, death */);
+					cur_tree->addPerson(name.c_str(), sex, parents.first, parents.second, birth, death);
+					return;
 				}
 				break;
 			/* case '#': */
@@ -134,23 +267,51 @@ namespace /* internal */
 				/* return; */
 			/* case 0...9: */
 			/* case ':': */
-			/* 	unexpected_char(*data); */
+			/* 	unexpected_char(*it_data); */
 			default:
 				break;
 			}
 		}
 	}
+
+	alloced preprocess(MappedFile file) /* remove whitespace and comments */
+	{
+		alloced res(file.len);
+
+		char* end = file.data + file.len;
+		for(char* it = file.data, * itres = res.data; it < end; ++it)
+		{
+			switch(*it)
+			{
+			case '#':
+				while(++it != end && *it != '\n'); /* skip comment */
+				*itres++ = '\n';				   /* add \n */
+				break;
+			case ' ':
+			case '\t':
+			case '\r':
+				break;
+			default:
+				*itres = *it;
+				++itres;
+			}
+		}
+		return res;
+	}
+
 }
 
 
 Tree parseTftFile(MappedFile file)
 {
-	data = file.data;
-	data_end = file.data + file.len;
 
 	Tree result("name");
 	cur_tree = &result;
 
+	alloced preprocessed = preprocess(file);
+	it_data = preprocessed.data;
+	end_data = preprocessed.end;
+
 	state_neutral();
 
 	return result;