yencode 1.0.8 → 1.1.2

package/test/testenc.js

@@ -0,0 +1,163 @@
+// a basic script to test that raw yEnc works as expected
+
+var assert = require('assert');
+var y = (function() {
+	try {
+		return require('../build/Debug/yencode.node');
+	} catch(x) {}
+	return require('../build/Release/yencode.node');
+})();
+var maxSize = require('./_maxsize');
+
+// slow reference yEnc implementation
+var refYEnc = function(src, line_size, col) {
+	var ret = [];
+	for (var i = 0; i < src.length; i++) {
+		var c = (src[i] + 42) & 0xFF;
+		switch(String.fromCharCode(c)) {
+			case '.':
+				if(col > 0) break;
+			case '\t': case ' ':
+				if(col > 0 && col < line_size-1) break;
+			case '\0': case '\r': case '\n': case '=':
+				ret.push('='.charCodeAt(0));
+				c += 64;
+				col++;
+		}
+		ret.push(c);
+		col++;
+		if(col >= line_size) {
+			ret.push('\r'.charCodeAt(0));
+			ret.push('\n'.charCodeAt(0));
+			col = 0;
+		}
+	}
+	
+	// if there's a trailing newline, trim it
+	if(ret[ret.length-2] == '\r'.charCodeAt(0) && ret[ret.length-1] == '\n'.charCodeAt(0)) {
+		ret.pop();
+		ret.pop();
+	}
+	
+	// if the last character is tab/space, escape it
+	if(ret[ret.length-1] == '\t'.charCodeAt(0) || ret[ret.length-1] == ' '.charCodeAt(0)) {
+		var c = ret[ret.length-1];
+		ret.pop();
+		ret.push('='.charCodeAt(0));
+		ret.push((c+64) & 0xFF);
+	}
+	
+	return new Buffer(ret);
+};
+var doTest = function(msg, test, expected) {
+	test[0] = new Buffer(test[0]);
+	if(!test[1]) test[1] = 128; // line size
+	if(!test[2]) test[2] = 0; // column offset
+	
+	if(!expected && expected !== '') expected = refYEnc.apply(null, test).toString('hex');
+	else expected = expected.replace(/ /g, '');
+	var actual = y.encode.apply(null, test).toString('hex');
+	if(actual != expected) {
+		console.log('Actual:', actual);
+		console.log('Expect:', expected);
+		console.log('Source:', test[0].toString('hex'));
+		assert.equal(actual, expected, msg);
+	}
+	
+	var buf = new Buffer(maxSize(test[0].length, test[1]));
+	var len = y.encodeTo.apply(null, [test[0], buf].concat(test.slice(1)));
+	assert.equal(buf.slice(0, len).toString('hex'), expected, msg);
+};
+
+var range = function(start, end) {
+	var ret = Array(end-start);
+	for(var i=start; i<end; i++)
+		ret[i-start] = i;
+	return ret;
+}
+var multiRange = function(list) {
+	var ret = [];
+	for(var i=0; i<list.length; i+=2) {
+		ret = ret.concat(range(list[i], list[i+1]));
+	}
+	return ret;
+}
+
+var lineSizes = // 48 sizes to test
+	range(1, 18).concat(range(23, 26)).concat(range(30, 35)).concat(range(46, 51)).concat(range(62, 67)).concat(range(126, 131)).concat([136, 145, 159]).concat(range(254, 259));
+var runLineSizes = function(fn) {
+	lineSizes.forEach(function(ls) {
+		lineSizes.forEach(function(offs) {
+			if(offs >= ls) return;
+			fn(ls, offs);
+		});
+	});
+};
+
+// simple tests
+runLineSizes(function(ls, offs) {
+	var infoStr = ' [ls='+ls+', offs='+offs+']';
+	var b = new Buffer(256);
+	b.fill(0);
+	doTest('Long no escaping' + infoStr, [b, ls]);
+	b.fill(227);
+	doTest('Long all escaping' + infoStr, [b, ls]);
+	b.fill(4);
+	doTest('Long all dots' + infoStr, [b, ls]);
+	b.fill(223);
+	doTest('Long all tabs' + infoStr, [b, ls]);
+});
+
+// case tests
+var padLens = range(0, 35).concat(range(46, 51)).concat(range(62, 67));
+var padding = new Buffer(128);
+padding.fill(97); // 'a'
+[ ['Empty test', [[]], ''],
+  ['Simple test', [[0,1,2,3,224,4]]],
+  ['Partial tab line', [[223, 223, 223]]],
+  ['Dot first should escape', [[4,3,224,2,1,0]]],
+  ['Short line', [[0,1,2,3,4], 2]],
+  ['Short line (2)', [[0,1,224,3,4], 2]],
+  ['Short line + offset', [[0,1,2,3,4], 2, 1]],
+  ['Short line (2) + offset', [[0,1,224,3,4], 2, 1]],
+  ['Tab & lf around line break', [[223,224], 128, 127], '3d 49 0d 0a 3d 4a'],
+].forEach(function(test) {
+	doTest.apply(null, test);
+	if(test.length == 2) {
+		// if no defined result, try padding the data to make it long enough to invoke SIMD behaviour
+		padLens.forEach(function(prepad) {
+			padLens.forEach(function(postpad) {
+				var newBuf = Buffer.concat([padding.slice(0, prepad), new Buffer(test[1][0]), padding.slice(0, postpad)]);
+				var newTest = test.slice();
+				newTest[1] = [newBuf].concat(newTest[1].slice(1));
+				doTest.apply(null, newTest);
+			});
+		});
+	}
+});
+
+// random tests
+for(var i=0; i<32; i++) {
+	var rand = require('crypto').pseudoRandomBytes(4*1024);
+	runLineSizes(function(ls, offs) {
+		doTest('Random [ls='+ls+', offs='+offs+']', [rand, ls, offs]);
+	});
+}
+
+// targeted random tests
+var charset = 'a\xd6\xdf\xe0\xe3\xf6\x04\x13';
+var randStr = function(n) {
+	var ret = new Buffer(n);
+	for(var i=0; i<n; i++)
+		ret[i] = charset[(Math.random() * charset.length) | 0].charCodeAt(0);
+	return ret;
+};
+for(var i=0; i<32; i++) {
+	var rand = randStr(2048);
+	runLineSizes(function(ls, offs) {
+		doTest('Random2 [ls='+ls+', offs='+offs+']', [rand, ls, offs]);
+	});
+}
+
+
+console.log('All tests passed');

package/test/testpostdec.js

@@ -0,0 +1,126 @@
+
+var assert = require('assert');
+var y = require('../index.js');
+
+var assertObjectHas = function(obj, required) {
+	for(var k in required) {
+		if(required[k] === undefined || required[k] === null)
+			assert(obj[k] === required[k]);
+		else if(typeof required[k] === 'object') // recurse to sub-objects
+			assertObjectHas(obj[k], required[k]);
+		else
+			assert.equal(obj[k], required[k], k);
+	}
+};
+
+var post, postData;
+
+// test basic post
+postData = [
+	'=ybegin part=5 line=128 size=500000 name=myBinary.dat',
+	'=ypart begin=499991 end=500000',
+	'..... data',
+	'=yend size=10 part=5 pcrc32=97f4bd52',
+	''
+].join('\r\n');
+post = y.from_post(new Buffer(postData));
+assertObjectHas(post, {
+	yencStart: 0,
+	dataStart: postData.indexOf('...'),
+	dataEnd: postData.indexOf('\r\n=yend'),
+	yencEnd: postData.length,
+	props: {
+		begin: {
+			part: '5',
+			line: '128',
+			size: '500000',
+			name: 'myBinary.dat'
+		},
+		part: {
+			begin: '499991',
+			end: '500000'
+		},
+		end: {
+			size: '10',
+			part: '5',
+			pcrc32: '97f4bd52'
+		}
+	},
+	warnings: undefined
+});
+assert.equal(post.data.length, post.dataEnd - post.dataStart); // there's no escaped characters in this post, so length should equal input length
+// won't bother testing actual encoding or CRC32 (leave this to other tests)
+
+
+// test post with extra data
+postData = [
+	'ignored data',
+	'=ybegin part=5a some_prop=hello line=0 size=0 name=name with space and = chars',
+	'.... data',
+	'=yend size=2 pcrc32=invalid pcrc32=invalid invalid_prop',
+].join('\r\n');
+post = y.from_post(new Buffer(postData));
+assertObjectHas(post, {
+	yencStart: postData.indexOf('=ybegin'),
+	dataStart: postData.indexOf('...'),
+	dataEnd: postData.indexOf('\r\n=yend'),
+	yencEnd: postData.length,
+	props: {
+		begin: {
+			part: '5a',
+			line: '0',
+			some_prop: 'hello',
+			name: 'name with space and = chars'
+		},
+		end: {
+			pcrc32: 'invalid'
+		}
+	}
+});
+assert.equal(post.data.length, post.dataEnd - post.dataStart);
+// check that expected warnings are there
+var seenWarnings = {};
+post.warnings.forEach(function(warn) {
+	seenWarnings[warn.code] = warn.message;
+});
+assert(seenWarnings['ignored_line_data']);
+assert(seenWarnings['duplicate_property']);
+assert(seenWarnings['missing_yend_newline']);
+assert(seenWarnings['invalid_prop_part']);
+assert(seenWarnings['zero_prop_line']);
+assert(seenWarnings['invalid_prop_pcrc32']);
+assert(seenWarnings['size_mismatch']);
+
+// empty post
+postData = [
+	'=ybegin size=0 line=1 name=',
+	'=yend size=0',
+	''
+].join('\r\n');
+post = y.from_post(new Buffer(postData));
+assertObjectHas(post, {
+	yencStart: 0,
+	dataStart: undefined,
+	dataEnd: undefined,
+	yencEnd: postData.length,
+	data: undefined,
+	warnings: undefined,
+	props: {
+		begin: {
+			size: '0',
+			line: '1',
+			name: ''
+		},
+		end: { size: '0' }
+	}
+});
+
+
+// test parse errors
+post = y.from_post(new Buffer('invalid post'));
+assert.equal(post.code, 'no_start_found');
+post = y.from_post(new Buffer('=ybegin abc=def'));
+assert.equal(post.code, 'no_end_found');
+
+
+console.log('All tests passed');

package/test.js

@@ -1,91 +0,0 @@
-// a basic script to test that raw yEnc works as expected
-
-var assert = require('assert');
-var y = require('./build/Release/yencode.node');
-
-// slow reference yEnc implementation
-var refYEnc = function(src, line_size, col) {
-	var ret = [];
-	for (var i = 0; i < src.length; i++) {
-		var c = (src[i] + 42) & 0xFF;
-		switch(String.fromCharCode(c)) {
-			case '.':
-				if(col > 0) break;
-			case '\t': case ' ':
-				if(col > 0 && col < line_size-1) break;
-			case '\0': case '\r': case '\n': case '=':
-				ret.push('='.charCodeAt(0));
-				c += 64;
-				col++;
-		}
-		ret.push(c);
-		col++;
-		if(col >= line_size) {
-			ret.push('\r'.charCodeAt(0));
-			ret.push('\n'.charCodeAt(0));
-			col = 0;
-		}
-	}
-	
-	// if there's a trailing newline, trim it
-	if(ret[ret.length-2] == '\r'.charCodeAt(0) && ret[ret.length-1] == '\n'.charCodeAt(0)) {
-		ret.pop();
-		ret.pop();
-	}
-	
-	// if the last character is tab/space, escape it
-	if(ret[ret.length-1] == '\t'.charCodeAt(0) || ret[ret.length-1] == ' '.charCodeAt(0)) {
-		var c = ret[ret.length-1];
-		ret.pop();
-		ret.push('='.charCodeAt(0));
-		ret.push((c+64) & 0xFF);
-	}
-	
-	return new Buffer(ret);
-};
-var doTest = function(msg, test, expected) {
-	test[0] = new Buffer(test[0]);
-	if(!test[1]) test[1] = 128; // line size
-	if(!test[2]) test[2] = 0; // column offset
-	
-	if(!expected) expected = refYEnc.apply(null, test).toString('hex');
-	else expected = expected.replace(/ /g, '');
-	assert.equal(y.encode.apply(null, test).toString('hex'), expected, msg);
-	var buf = new Buffer(require('./').maxSize(test[0].length, test[1]));
-	var len = y.encodeTo.apply(null, [test[0], buf].concat(test.slice(1)));
-	assert.equal(buf.slice(0, len).toString('hex'), expected, msg);
-};
-
-
-doTest('Empty test', [[]], '');
-doTest('Simple test', [[0,1,2,3,224,4]]);
-doTest('Partial tab line', [[223, 223, 223]]);
-doTest('Dot first should escape', [[4,3,224,2,1,0]]);
-doTest('Short line', [[0,1,2,3,4], 2]);
-doTest('Short line (2)', [[0,1,224,3,4], 2]);
-doTest('Short line + offset', [[0,1,2,3,4], 2, 1]);
-doTest('Short line (2) + offset', [[0,1,224,3,4], 2, 1]);
-doTest('Tab & lf around line break', [[223,224], 128, 127], '3d 49 0d 0a 3d 4a');
-
-// longer tests
-var b = new Buffer(256);
-b.fill(0);
-doTest('Long no escaping', [b]);
-b.fill(227);
-doTest('Long all escaping', [b]);
-b.fill(4);
-doTest('Long all dots', [b]);
-b.fill(223);
-doTest('Long all tabs', [b]);
-
-
-// random tests
-for(var i=0; i<32; i++) {
-	var rand = require('crypto').pseudoRandomBytes(128*1024);
-	doTest('Random', [rand]);
-	doTest('Random + short line', [rand, 3]);
-	doTest('Random + offset', [rand, 128, 1]);
-	doTest('Random + offset (end)', [rand, 128, 127]);
-}
-
-console.log('All tests passed');