ragel environment parser.

mpw/CMakeLists.txt

@@ -6,6 +6,11 @@ add_definitions(-I ${CMAKE_SOURCE_DIR}/)
 
 set(MPW_SRC mpw.cpp mpw_io.cpp mpw_close.cpp mpw_access.cpp mpw_ioctl.cpp environ.cpp)
 
+add_custom_command(
+	OUTPUT environ.cpp
+	COMMAND ragel -p -G2 -o environ.cpp "${CMAKE_CURRENT_SOURCE_DIR}/environ.rl"
+	MAIN_DEPENDENCY environ.rl 
+)
 
 
 add_library(MPW_LIB ${MPW_SRC})

mpw/environ.cpp

@@ -1,356 +0,0 @@
-/* Generated by re2c 0.13.5 on Sat May 18 16:26:02 2013 */
-#include <string>
-#include <unordered_map>
-
-#include <cstdio>
-
-/*
- * #...comment
- * var=value
- * where value may contain $var and ${var} interpolations.
- *
- */
-
-namespace MPW {
-std::string EvalString(std::string &s, std::unordered_map<std::string, std::string> &env)
-{
-
-	std::string rv;
-	const char *cp = s.c_str();
-	const char *marker = NULL;
-
-	while (*cp)
-	{
-		const char *begin = cp;
-
-		
-	{
-		char yych;
-
-		yych = (char)*cp;
-		switch (yych) {
-		case '\n':	goto yy5;
-		case '$':	goto yy2;
-		default:	goto yy4;
-		}
-yy2:
-		yych = (char)*(marker = ++cp);
-		switch (yych) {
-		case '0':
-		case '1':
-		case '2':
-		case '3':
-		case '4':
-		case '5':
-		case '6':
-		case '7':
-		case '8':
-		case '9':
-		case 'A':
-		case 'B':
-		case 'C':
-		case 'D':
-		case 'E':
-		case 'F':
-		case 'G':
-		case 'H':
-		case 'I':
-		case 'J':
-		case 'K':
-		case 'L':
-		case 'M':
-		case 'N':
-		case 'O':
-		case 'P':
-		case 'Q':
-		case 'R':
-		case 'S':
-		case 'T':
-		case 'U':
-		case 'V':
-		case 'W':
-		case 'X':
-		case 'Y':
-		case 'Z':
-		case '_':
-		case 'a':
-		case 'b':
-		case 'c':
-		case 'd':
-		case 'e':
-		case 'f':
-		case 'g':
-		case 'h':
-		case 'i':
-		case 'j':
-		case 'k':
-		case 'l':
-		case 'm':
-		case 'n':
-		case 'o':
-		case 'p':
-		case 'q':
-		case 'r':
-		case 's':
-		case 't':
-		case 'u':
-		case 'v':
-		case 'w':
-		case 'x':
-		case 'y':
-		case 'z':	goto yy9;
-		case '{':	goto yy7;
-		default:	goto yy3;
-		}
-yy3:
-		{
-			rv.push_back(*begin);
-			continue;
-		}
-yy4:
-		yych = (char)*++cp;
-		goto yy3;
-yy5:
-		++cp;
-		{
-			break;
-		}
-yy7:
-		yych = (char)*++cp;
-		switch (yych) {
-		case '}':	goto yy8;
-		default:	goto yy13;
-		}
-yy8:
-		cp = marker;
-		goto yy3;
-yy9:
-		++cp;
-		yych = (char)*cp;
-		switch (yych) {
-		case '0':
-		case '1':
-		case '2':
-		case '3':
-		case '4':
-		case '5':
-		case '6':
-		case '7':
-		case '8':
-		case '9':
-		case 'A':
-		case 'B':
-		case 'C':
-		case 'D':
-		case 'E':
-		case 'F':
-		case 'G':
-		case 'H':
-		case 'I':
-		case 'J':
-		case 'K':
-		case 'L':
-		case 'M':
-		case 'N':
-		case 'O':
-		case 'P':
-		case 'Q':
-		case 'R':
-		case 'S':
-		case 'T':
-		case 'U':
-		case 'V':
-		case 'W':
-		case 'X':
-		case 'Y':
-		case 'Z':
-		case '_':
-		case 'a':
-		case 'b':
-		case 'c':
-		case 'd':
-		case 'e':
-		case 'f':
-		case 'g':
-		case 'h':
-		case 'i':
-		case 'j':
-		case 'k':
-		case 'l':
-		case 'm':
-		case 'n':
-		case 'o':
-		case 'p':
-		case 'q':
-		case 'r':
-		case 's':
-		case 't':
-		case 'u':
-		case 'v':
-		case 'w':
-		case 'x':
-		case 'y':
-		case 'z':	goto yy9;
-		default:	goto yy11;
-		}
-yy11:
-		{
-			std::string tmp(begin + 1, cp);
-
-			// lookup value...
-			// append to rv.
-			auto iter = env.find(tmp);
-			if (iter != env.end())
-			{
-				rv.append(iter->second);
-			}
-			continue;
-		}
-yy12:
-		++cp;
-		yych = (char)*cp;
-yy13:
-		switch (yych) {
-		case '0':
-		case '1':
-		case '2':
-		case '3':
-		case '4':
-		case '5':
-		case '6':
-		case '7':
-		case '8':
-		case '9':
-		case 'A':
-		case 'B':
-		case 'C':
-		case 'D':
-		case 'E':
-		case 'F':
-		case 'G':
-		case 'H':
-		case 'I':
-		case 'J':
-		case 'K':
-		case 'L':
-		case 'M':
-		case 'N':
-		case 'O':
-		case 'P':
-		case 'Q':
-		case 'R':
-		case 'S':
-		case 'T':
-		case 'U':
-		case 'V':
-		case 'W':
-		case 'X':
-		case 'Y':
-		case 'Z':
-		case '_':
-		case 'a':
-		case 'b':
-		case 'c':
-		case 'd':
-		case 'e':
-		case 'f':
-		case 'g':
-		case 'h':
-		case 'i':
-		case 'j':
-		case 'k':
-		case 'l':
-		case 'm':
-		case 'n':
-		case 'o':
-		case 'p':
-		case 'q':
-		case 'r':
-		case 's':
-		case 't':
-		case 'u':
-		case 'v':
-		case 'w':
-		case 'x':
-		case 'y':
-		case 'z':	goto yy12;
-		case '}':	goto yy14;
-		default:	goto yy8;
-		}
-yy14:
-		++cp;
-		{
-		
-			std::string tmp(begin + 2, cp - 1);
-
-			auto iter = env.find(tmp);
-			if (iter != env.end())
-			{
-				rv.append(iter->second);
-			}
-			continue;
-		}
-	}
-
-
-	}
-
-	return rv;
-}
-
-
-void LoadEnvironment(std::string &envfile, std::unordered_map<std::string, std::string> &env)
-{
-
-	FILE *fp;
-
-
-	fp = fopen(envfile.c_str(), "r");
-
-	if (!fp) return;
-
-	for (;;)
-	{
-		char *begin;
-		char *end;
-		char *iter;
-
-		size_t length;		
-
-		begin = fgetln(fp, &length);
-		if (!begin) break; // eof or error.
-
-		if (!length) continue;
-		if (begin[0] == '#') continue;
-
-		while (length && isspace(begin[length - 1])) --length;
-		if (!length) continue;
-
-		// key=value
-
-		// ehh, this could really check for [A-Za-z0-9_]+ '='
-
-		end = begin + length;
-		iter = std::find(begin, end, '=');
-
-		if (iter == end || iter == begin)
-		{
-			fprintf(stderr, "Invalid Environment entry: %.*s\n", (int)length, begin);
-			continue;
-		}
-
-		std::string key(begin, iter);
-		std::string value(iter + 1, end);
-
-		value = EvalString(value, env);
-
-		//fprintf(stdout, "%s = %s\n", key.c_str(), value.c_str());
-
-		// todo -- does this replace an existing value?
-		env.emplace(key, value);
-	}
-
-	fclose(fp);
-}
-
-}

mpw/environ.rl

@@ -0,0 +1,145 @@
+
+%%{
+	
+	machine lexer;
+
+
+	ws = [ \t\r\n];
+
+	word = [A-Za-z0-9_];
+
+	action emplace {
+		//printf("emplacing %s\n", name.c_str());
+		// trim any whitespace.
+		while (value.length() && isspace(value.back()))
+			value.pop_back();	
+		env[std::move(name)] = std::move(value);
+	}
+
+	action error {
+		fprintf(stderr, "Bad environment: %.*s\n", (int)length, line);	
+	}
+
+	value := |*
+
+		'$' word+  => {
+			std::string name(ts + 1, te);
+			auto iter = env.find(name);
+			if (iter != env.end())
+				value.append(iter->second);
+		};
+		
+		'${' word+ '}'  => {
+			std::string name(ts + 2, te - 1);
+			auto iter = env.find(name);
+			if (iter != env.end())
+				value.append(iter->second);
+		};
+
+		'$$' => {
+			value.push_back('$');
+		};
+
+		any $eof(emplace)  => {
+			value.push_back(fc);
+		} ;
+
+		*|;
+
+
+	# exit w/ cs = lexer_en_comment.
+
+	comment := any* ${ fbreak; };
+
+	assignment = 
+	word+ ${ name.push_back(fc); }
+	ws* %eof(error)
+	'='
+	ws*
+	(any - ws)? ${ fhold; fgoto value; }
+	%eof(emplace)
+	;
+
+	main := ws* ( '#' ${ fgoto comment; } | assignment);
+
+}%%
+
+#include <string>
+#include <unordered_map>
+#include <cstdio>
+#include <cctype>
+
+namespace MPW {
+
+void LoadEnvironment(std::string &envfile, std::unordered_map<std::string, std::string> &env)
+{
+	%% write data;
+
+	FILE *fp;
+
+
+	fp = fopen(envfile.c_str(), "r");
+
+	if (!fp) return;
+
+	for(;;)
+	{
+		const char *line;
+		size_t length;
+
+		line = fgetln(fp, &length);
+		if (!line) break; // eof or error.
+
+		// strip any trailign space.
+		while (length && isspace(line[length - 1]))
+			length--;
+
+		std::string name;
+		std::string value;
+
+		const char *p = line;
+		const char *pe = line + length;
+		const char *eof = line + length;
+		const char *te;
+		const char *ts;
+
+		int cs, act;
+
+		%%write init;
+		%%write exec;
+
+		if (cs == lexer_error)
+		{
+			fprintf(stderr, "Bad environment: %.*s\n", (int)length, line);
+			continue;
+		}
+	}
+
+	fclose(fp);
+}
+
+} // namespace
+
+#ifdef TEST
+
+int main(int argc, char **argv)
+{
+
+	for (int i = 1; i < argc; ++i)
+	{
+		std::unordered_map<std::string, std::string> env;
+		std::string f(argv[i]);
+
+		MPW::LoadEnvironment(f, env);
+
+		for (const auto &kv : env)
+		{
+			printf("%s -> %s\n", kv.first.c_str(), kv.second.c_str());
+		} 
+	}
+
+	return 0;
+}
+
+
+#endif