ANBF语法中的数值类型有3种:二进制、十进制和十六进制,可以是一个以点号分隔的数列,也可以是一个数值的范围。例如,%d11.22.33.44.55表示五个有次序的十进制数字“11、22、33、44、55”,而%x80-ff表示一个字节,这个字节的数值可以是在0x80至0xff之间。
我把以点号分隔的数列定义为NumVal,把范围类型的数值定义为RangedNumVal。这两个类实现了Element,其实我觉得应该定义一个接口NumVal(继承Element),然后一个SerialNumVal和一个RangedNumVal(实现NumVal),这样看起来更漂亮?作为一个完美主义者看到现在这个定义真是很蛋疼,有时间再重新考虑吧。
由于二进制、十进制和十六进制的构成都是很相似的,只是进制符号(b、d、x)以及数字符号(01、0123456789、0123456789abcdef)不同而已,为了避免重复地写三个很相像的方法,我投机取巧的定义了一个Matcher接口,这个接口是用来判断字符是否在预设的符号集里面的,没什么技术含量,看代码就明白了。
先来看看解析代码:
/* This file is one of the component a Context-free Grammar Parser Generator, which accept a piece of text as the input, and generates a parser for the inputted context-free grammar. Copyright (C) 2013, Junbiao Pan (Email: [email protected]) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>. */ // bin-val = "b" 1*BIT // [ 1*("." 1*BIT) / ("-" 1*BIT) ] // BIT = "0" / "1" // 二进制解析器 protected Element bin_val() throws IOException, MatchException { // 真正的解析工作由val方法完成,只要把二进制数的符号集{0、1}通过Matcher实例传递给它就OK了。 return val('b', new Matcher() { @Override public boolean match(int value) { // 如果符号是0或1就匹配 return value == '0' || value == '1'; } @Override public String expected() { // 提示符号不在符号集内(仅用于异常情况) return "['0', '1']"; } }); } // dec-val = "d" 1*DIGIT // [ 1*("." 1*DIGIT) / ("-" 1*DIGIT) ] protected Element dec_val() throws IOException, MatchException { // 同上,把十进制的符号集0~9传递给val方法 return val('d', new Matcher() { @Override public boolean match(int value) { // 直到写博客才发现这段代码错了,符号集不应该包含A~F的情形啊,居然单元测试已经通过了,尼玛这是什么测试质量! // PS:单元测试代码也是我自己写的。。。 return (value >= 0x30 && value <= 0x39) || (value >= 'A' && value <= 'F') || (value >= 'a' && value <= 'f'); } @Override public String expected() { // 错误代码,无语了。。。 return "['0'-'9', 'A'-'F', 'a'-'f']"; } }); } // hex-val = "x" 1*HEXDIG // [ 1*("." 1*HEXDIG) / ("-" 1*HEXDIG) ] protected Element hex_val() throws IOException, MatchException { // 将十六进制的符号集通过Matcher实例传递给val方法进行解析 return val('x', new Matcher() { @Override public boolean match(int value) { return (value >= 0x30 && value <= 0x39) || (value >= 'A' && value <= 'F') || (value >= 'a' && value <= 'f'); } @Override public String expected() { return "['0'-'9', 'A'-'F', 'a'-'f']"; } }); } // 解析各个进制 protected Element val(char base, Matcher matcher) throws IOException, MatchException { // 检查进制符号 assertMatch(is.peek(), base); int baseValue = is.read(); String from = ""; String val = ""; // 进制符号之后的第一个字符,必须在Matcher定义的字符集内,否则异常 if (matcher.match(is.peek())) { // 连续读入符合字符的字符,构成NumVal的第一个数值。 while (matcher.match(is.peek())) { from += (char)is.read(); } // 第一个数值后面如果是跟着点号,则是一个数列NumVal,如果是-破折号,则是一个范围型数值RangedNumVal,如果都不是,则是单一个数值 if (match(is.peek(), '.')) { NumVal numval = new NumVal(String.valueOf((char)baseValue)); // 将刚才匹配到的数值作为第一个数值加到将要返回的NumVal中 numval.addValue(from); // 如果后面跟着点号,则继续加入新的数值到NumVal中 while (match(is.peek(), '.')) { int next = is.peek(1); if (!(matcher.match(next))) { break; } is.read(); val = ""; while (matcher.match(is.peek())) { val += (char)is.read(); } numval.addValue(val); } // 直到不能匹配到点号,数列结束,返回 return numval; } else if (match(is.peek(), '-')) { // 这里向前读取两个字符,因此即使破折号后面跟着的不是数字,也能返回单一个数字而且将破折号留给后面的分析程序 // 这是本程序里为数不多的能够具备回溯的代码段之一,嘿嘿。 int next = is.peek(1); if (!(matcher.match(next))) { // 如果破折号后面跟的不是数字,则破折号不读入,返回单一数值 NumVal numval = new NumVal(String.valueOf((char)baseValue)); numval.addValue(from); return numval; } // 否则,破折号后面是数值,读取之,并返回RangedNumVal类型 is.read(); val =""; val += (char)is.read(); while (matcher.match(is.peek())) { val += (char)is.read(); } return new RangedNumVal(String.valueOf((char)baseValue), from, val); } else { // 第一个数值之后跟的既不是点号,也不是破折号,则返回单一数值 NumVal numval = new NumVal(String.valueOf((char)baseValue)); numval.addValue(from); return numval; } } else { throw new MatchException(matcher.expected(), is.peek(), is.getPos(), is.getLine()); } } // num-val = "%" (bin-val / dec-val / hex-val) // 解析num-val protected Element num_val() throws IOException, MatchException { String base = "", from ="", val =""; // 百分号开头 assertMatch(is.peek(), '%'); is.read(); // 根据进制符号选择相应的解析方法(函数) switch ((char)is.peek()) { case 'b': case 'B': return bin_val(); case 'd': case 'D': return dec_val(); case 'x': case 'X': return hex_val(); default: throw new MatchException("['b', 'd', 'x']", is.peek(), is.getPos(), is.getLine()); } }
接下来看看单元测试部分,不详细说了,其中有一句注释说明为什么上面有错误代码不能测试出来:
/* This file is one of the component a Context-free Grammar Parser Generator, which accept a piece of text as the input, and generates a parser for the inputted context-free grammar. Copyright (C) 2013, Junbiao Pan (Email: [email protected]) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>. */ // bin-val = "b" 1*BIT // [ 1*("." 1*BIT) / ("-" 1*BIT) ] // BIT = "0" / "1" // 测试二进制数的解析 @Test public void testBin_val() throws Exception { Tester<String> tester = new Tester<String>() { @Override public String test(AbnfParser parser) throws MatchException, IOException { return parser.bin_val().toString(); } }; String input; input = "b1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); input = "b1010"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); input = "B1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); input = "b1.1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); input = "b0101.1111"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); input = "b0000-1111"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".00").bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"-1234").bin_val().toString()); input = "b00.11.00.01.10.00.11.00.11"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".").bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"..").bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".bb").bin_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"-00").bin_val().toString()); Assertion.assertMatchException("", tester, 1, 1); Assertion.assertMatchException("b", tester, 2,1); Assertion.assertMatchException("bg", tester, 2, 1); Assertion.assertMatchException("b.", tester, 2, 1); Assertion.assertMatchException("b-", tester, 2, 1); } // dec-val = "d" 1*DIGIT // [ 1*("." 1*DIGIT) / ("-" 1*DIGIT) ] // 测试十进制数的解析 @Test public void testDec_val() throws Exception { Tester<String> tester = new Tester<String>() { @Override public String test(AbnfParser parser) throws MatchException, IOException { return parser.dec_val().toString(); } }; String input; input = "d1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); input = "d1234"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); input = "D1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); input = "d1.2"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); input = "d1234.5678"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); input = "d1234-5678"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".00").dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"-1234").dec_val().toString()); input = "d12.34.56.78.9a.bc.de.f0"; // 看看这里,就明白为什么单元测试测不出十进制数带有a~f符号的问题了,竟然有这样错误的测试用例!!! Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".").dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"..").dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+".##").dec_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"-00").dec_val().toString()); Assertion.assertMatchException("", tester, 1, 1); Assertion.assertMatchException("d", tester, 2, 1); Assertion.assertMatchException("dg", tester, 2, 1); Assertion.assertMatchException("d.", tester, 2, 1); Assertion.assertMatchException("d-", tester, 2, 1); } // hex-val = "x" 1*HEXDIG // [ 1*("." 1*HEXDIG) / ("-" 1*HEXDIG) ] // 测试十六进制数的解析 @Test public void testHex_val() throws Exception { Tester<String> tester = new Tester<String>() { @Override public String test(AbnfParser parser) throws MatchException, IOException { return parser.hex_val().toString(); } }; String input; input = "x1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "x1234"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "X1"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "x1.2"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "x1234.5678"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "xabcd.ef"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "xA1.2B"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); input = "x1234-abCD"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input+"-").hex_val().toString()); input = "x12.34.56.78.9a.bc.de.f0"; Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input).hex_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input + ".").hex_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input + ".g0").hex_val().toString()); Assert.assertEquals("%" + input, AbnfParserFactory.newInstance(input + "-00").hex_val().toString()); Assertion.assertMatchException("", tester, 1, 1); Assertion.assertMatchException("x", tester, 2, 1); Assertion.assertMatchException("xg", tester, 2, 1); Assertion.assertMatchException("x.", tester, 2, 1); Assertion.assertMatchException("x-", tester, 2, 1); } // num-val = "%" (bin-val / dec-val / hex-val) // 综合情况测试 @Test public void testNum_val() throws Exception { String input; input = "%b0101"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%b0101.1010.1111"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%b0101-1111"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%d1234"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%d0123.4567.8901"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%d12345-67890"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%x0123"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%x0123.4567.89ab.CDEF"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); input = "%x0123456789-ABCDEFabcdef09"; Assert.assertEquals(input, AbnfParserFactory.newInstance(input).num_val().toString()); }
本系列文章索引:基于预测的ABNF文法分析器