node-marshal 0.1.2 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 9cb2222133e284b6420cd51abdcfeb7b7e266e73
-  data.tar.gz: 864b5b93258b51e1d9a9bbbf309ba655447f091b
+  metadata.gz: d564a41c56aa30cec81fe572e069cd4734637995
+  data.tar.gz: 28c7a859af9ba4cfa82c2f8138e0a9b1708f3176
 SHA512:
-  metadata.gz: 73a4e165ce77ad01f7cc8af2ae257c16f177dc509c78e783e1872fb88495efb33bb409590996d83925f9791adf3dac67bd8614e15394d8d5a6dfd83742bdecb9
-  data.tar.gz: c653aaf8b6f25e5ce4ab8f6d5c953ae0bd53d89aaf7adeec60b9172072d8430de84676366c65e133789532215986e5c913570f2ab2d3eeed79b1d61788a74985
+  metadata.gz: e7a41dbf52db03ccb69a4ccd9641ff740f45c88ba27488e17dea2897655c0920e1880edfcb29ded1801e25287358bd81d77097453e19c97ca2ff55fa46c90fa0
+  data.tar.gz: f31215d9e96c760d21927279162f48627d4bbace53ad66fe43d7335f8de403788e77022570a933fb6b4e5fee65cbd829f26ae9c3d5cd8c63903087ca10048fdd

data/{ext/node-marshal/COPYING → COPYING}

@@ -1,4 +1,4 @@
-Copyright (C) 2015 Alexey Voskov. All rights reserved.
+Copyright (C) 2015-2016 Alexey Voskov. All rights reserved.
 Copyright (C) 1993-2013 Yukihiro Matsumoto. All rights reserved.
 
 Redistribution and use in source and binary forms, with or without

data/README.rdoc

@@ -17,6 +17,15 @@ The key features of node-marshal gem:
 - 2-clause BSD license suitable for creation of custom source code protection system
 
 Changelog:
+- 11.JAN.2016 - 0.2.0
+      - Bugfix: || and && in NODE_OP_ASGN1 (e.g. in x['a'] ||= 'b' or x['b'] &&= false)
+        (this bug caused segfaults in some cases)
+      - Bugfix: NodeMarshal#dump_node_short
+      - Format version changed to NODEMARSHAL11 (support of symbols not representable in
+        the form of String is added)
+      - show_offsets property for controlling verbosity of NodeMarshal#dump_node_short output
+      - Improved information about licenses
+      - Improved rdoc documentation
 - 24.DEC.2015 - 0.1.2
       - Ruby 2.3.x preliminary support (including &. safe navigation operator)
       - Bugfix: NODE_MATCH3 (a =~ /abc/) issue (reported by Gregory Siehień)
@@ -24,4 +33,4 @@ Changelog:
       - Bugfix: failure in the case of syntax errors in the input data (now ArgumentError
         exception will be generated instead of it)
       - Ability of symbols renaming (for hiding variables and constants name during code obfuscation)
-- 04.MAY.2015 - 0.1.1 - first public version
+- 04.MAY.2015 - 0.1.1 - first public version

data/bin/noderbc

@@ -2,13 +2,12 @@
 require_relative '../lib/node-marshal.rb'
 
 help = <<-EOS
-Ruby source files compiler. It is based on NodeMarshal class.
-from NodeDump gem. Source code is transformed to the Ruby node
-(syntax tree) serialized into ASCII string. Such transformation
-is based on Ruby internal syntax tree, irreversible and can
-be used for code obfuscation.
+Ruby source files compiler from node-marshal gem. It is based
+on NodeMarshal class. Source code is irreversibly transformed to the 
+Ruby node (syntax tree) serialized into ASCII string. It can be used
+for for code obfuscation.
 
-(C) 2015 Alexey Voskov. License: 2-clause BSD.
+(C) 2015-2016 Alexey Voskov. License: BSD-2-Clause.
 
 Usage:
   noderbc inpfile outfile [options]
@@ -61,5 +60,4 @@ else
 	outfile = ARGV[1]
 	raise 'inpfile and outfile cannot be equal' if inpfile == outfile
 	NodeMarshal.compile_rb_file(outfile, inpfile, opts)
-
 end

data/ext/node-marshal/base85r.c

@@ -10,8 +10,8 @@
  * 2) big-endian 5-byte numbers (base 85)
  * 3) empty string: arbitrary two bytes
  *
- * (C) 2015 Alexey Voskov
- * License: 2-clause BSD
+ * (C) 2015-2016 Alexey Voskov
+ * License: BSD-2-Clause
  */
 #include <stdio.h>
 #include <stdlib.h>
@@ -78,8 +78,10 @@ static int base85_encode_buf_len(int len)
 	return buf_len;
 }
 
-/* Encode string to modified BASE85 ASCII.
-   Call base85_init_tables before using of this function */
+/* 
+ * Encode string to modified BASE85 ASCII.
+ * Note: call base85_init_tables before using of this function
+ */
 VALUE base85r_encode(VALUE input)
 {
 	VALUE output;
@@ -131,8 +133,10 @@ VALUE base85r_encode(VALUE input)
 }
 
 
-/* Decode string in modified BASE85 ASCII format.
-   Call base85_init_tables before using of this function */
+/* 
+ * Decode string in modified BASE85 ASCII format.
+ * Note: call base85_init_tables before using of this function
+ */
 VALUE base85r_decode(VALUE input)
 {
 	int inp_len, out_len, pos, shift;

data/ext/node-marshal/nodedump.c

@@ -2,8 +2,8 @@
  * This file contains implementation of classes for Ruby nodes
  * marshalization (i.e. loading and saving them from disk)
  * 
- * (C) 2015 Alexey Voskov
- * License: 2-clause BSD
+ * (C) 2015-2016 Alexey Voskov
+ * License: BSD-2-Clause
  */
 #define __STDC_FORMAT_MACROS
 #include <stdio.h>
@@ -204,6 +204,23 @@ int LeafTableInfo_addEntry(LeafTableInfo *lti, VALUE key, VALUE value)
 	}
 }
 
+/*
+ * Adds Ruby ID data type as the entry to the LeafTableInfo struct.
+ * Main features:
+ * 1) ID will be converted to Fixnum
+ * 2) If ID can be converted to string by rb_id2str it will be saved as
+      String object. Otherwise it will be converted to Fixnum.
+ */
+int LeafTableInfo_addIDEntry(LeafTableInfo *lti, ID id)
+{
+	VALUE r_idval = rb_id2str(id);
+	if (TYPE(r_idval) != T_STRING)
+	{
+		r_idval = INT2FIX(id);
+	}
+	return LeafTableInfo_addEntry(lti, INT2FIX(id), r_idval);
+}
+
 VALUE LeafTableInfo_getLeavesTable(LeafTableInfo *lti)
 {
 	VALUE key, keys = rb_funcall(lti->vals, rb_intern("keys"), 0);
@@ -389,7 +406,6 @@ static int dump_node_value(NODEInfo *info, char *ptr, NODE *node, int type, VALU
 		}
 		else
 		{
-			//memcpy(ptr, &id, sizeof(VALUE));
 			switch(type)
 			{
 				case NT_ENTRY: return DUMP_RAW_VALUE(VL_GVAR, id);
@@ -422,7 +438,7 @@ static VALUE dump_nodes(NODEInfo *info)
 		nt = nd_type(node);
 		rtypes = (char *) ptr; ptr += sizeof(int);
 		flags_len = value_to_bin(node->flags >> 5, (unsigned char *) ptr); ptr += flags_len;
-		//memcpy(ptr, &(node->flags), sizeof(VALUE)); ptr += sizeof(VALUE);
+
 		ut[0] = nodes_ctbl[nt * 3];
 		ut[1] = nodes_ctbl[nt * 3 + 1];
 		ut[2] = nodes_ctbl[nt * 3 + 2];
@@ -456,7 +472,6 @@ static VALUE dump_nodes(NODEInfo *info)
 		rtypes[3] = flags_len;
 	}
 	rb_str_resize(nodes_bin, (int) (ptr - bin) + 1);
-	//printf("%d", ptr - bin);
 	return nodes_bin;
 }
 
@@ -475,7 +490,7 @@ VALUE NODEInfo_toHash(NODEInfo *info)
 	VALUE args;
 	int i, j, id;
 	// Add some signatures
-	rb_hash_aset(ans, ID2SYM(rb_intern("MAGIC")), rb_str_new2("NODEMARSHAL10"));
+	rb_hash_aset(ans, ID2SYM(rb_intern("MAGIC")), rb_str_new2(NODEMARSHAL_MAGIC));
 	rb_hash_aset(ans, ID2SYM(rb_intern("RUBY_PLATFORM")),
 		rb_const_get(rb_cObject, rb_intern("RUBY_PLATFORM")));
 	rb_hash_aset(ans, ID2SYM(rb_intern("RUBY_VERSION")),
@@ -573,7 +588,7 @@ static void NODEInfo_addValue(NODEInfo *info, VALUE value)
  * Function counts number of nodes and fills NODEInfo struct
  * that is neccessary for the node saving to the HDD
  */
-int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
+static int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 {
 	int ut[3], num, offset;
 	if (node == 0)
@@ -632,7 +647,7 @@ int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 		}
 		else if (ut[0] == NT_ID)
 		{
-			LeafTableInfo_addEntry(&info->syms, INT2FIX(node->u1.id), rb_id2str(node->u1.id));
+			LeafTableInfo_addIDEntry(&info->syms, node->u1.id);
 		}
 		else if (ut[0] == NT_VALUE)
 		{
@@ -650,7 +665,7 @@ int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 			{
 				ID sym = *idtbl++;
 				rb_ary_push(idtbl_ary, INT2FIX(sym));
-				LeafTableInfo_addEntry(&info->syms, INT2FIX(sym), rb_id2str(sym));
+				LeafTableInfo_addIDEntry(&info->syms, sym);
 			}
 			LeafTableInfo_addEntry(&info->idtabs, tkey, idtbl_ary);
 		}
@@ -665,7 +680,7 @@ int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 		}
 		else if (ut[1] == NT_ID)
 		{
-			LeafTableInfo_addEntry(&info->syms, INT2FIX(node->u2.id), rb_id2str(node->u2.id));
+			LeafTableInfo_addIDEntry(&info->syms, node->u2.id);
 		}
 		else if (ut[1] == NT_VALUE)
 		{
@@ -685,7 +700,7 @@ int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 		}
 		else if (ut[2] == NT_ID)
 		{
-			LeafTableInfo_addEntry(&info->syms, INT2FIX(node->u3.id), rb_id2str(node->u3.id));
+			LeafTableInfo_addIDEntry(&info->syms, node->u3.id);
 		}
 		else if (ut[2] == NT_ARGS)
 		{
@@ -709,29 +724,29 @@ int count_num_of_nodes(NODE *node, NODE *parent, NODEInfo *info)
 
 			asym = ainfo->first_post_arg; rb_ary_push(varg, INT2FIX(asym)); // ID
 			if (asym != 0)
-				LeafTableInfo_addEntry(&info->syms, INT2FIX(asym), rb_id2str(asym));
+				LeafTableInfo_addIDEntry(&info->syms, asym);
 
 			asym = ainfo->rest_arg; rb_ary_push(varg, INT2FIX(asym)); // ID
 			if (asym != 0)
-				LeafTableInfo_addEntry(&info->syms, INT2FIX(asym), rb_id2str(asym));
+				LeafTableInfo_addIDEntry(&info->syms, asym);
 
 			asym = ainfo->block_arg; rb_ary_push(varg, INT2FIX(asym)); // ID
 			if (asym != 0)
-				LeafTableInfo_addEntry(&info->syms, INT2FIX(asym), rb_id2str(asym));
+				LeafTableInfo_addIDEntry(&info->syms, asym);
 			rb_ary_push(varg, value_to_str((VALUE) ainfo->kw_args));
 			rb_ary_push(varg, value_to_str((VALUE) ainfo->kw_rest_arg));
 			rb_ary_push(varg, value_to_str((VALUE) ainfo->opt_args));
 
 			LeafTableInfo_addEntry(&info->args, value_to_str((VALUE) ainfo), varg);
 #else
-			rb_raise(rb_eArgError, "NT_ARGS entry withour USE_RB_ARGS_INFO");
+			rb_raise(rb_eArgError, "NT_ARGS entry without USE_RB_ARGS_INFO");
 #endif
 		}
 		else if (ut[2] == NT_ENTRY)
 		{
 			ID gsym = node->u3.entry->id;
 			// Save symbol to the symbol table
-			int newid = LeafTableInfo_addEntry(&info->syms, INT2FIX(gsym), rb_id2str(gsym));
+			int newid = LeafTableInfo_addIDEntry(&info->syms, gsym);
 			LeafTableInfo_addEntry(&info->gentries, value_to_str(node->u3.value), INT2FIX(newid));
 		}
 		else if (ut[2] != NT_LONG && ut[2] != NT_NULL)
@@ -799,10 +814,16 @@ void rbstr_printf(VALUE str, const char *fmt, ...)
 }
 
 #define PRINT_NODE_TAB for (j = 0; j < tab; j++) rbstr_printf(str, "  ");
-void print_node(VALUE str, NODE *node, int tab)
+/*
+ * Recursively transforms node into Ruby string
+ *   str -- output Ruby string
+ *   node -- input Ruby NODE
+ *   tab -- number of tabulations during print
+ *   show_offsets -- 0/1 show/hide addresses and symbol IDs
+ */
+static void print_node(VALUE str, NODE *node, int tab, int show_offsets)
 {
 	int i, j, type, ut[3];
-	int show_offsets = 0;
 	VALUE uref[3];
 
 	PRINT_NODE_TAB
@@ -815,12 +836,14 @@ void print_node(VALUE str, NODE *node, int tab)
 
 	if (show_offsets)
 	{
-		rbstr_printf(str, "@ %s | %16"PRIxPTR " %16"PRIxPTR " %16"PRIxPTR "\n", ruby_node_name(type),
-			(intptr_t) node->u1.value, (intptr_t) node->u2.value, (intptr_t) node->u3.value);
+		rbstr_printf(str, "@ %s | %16"PRIxPTR " %16"PRIxPTR " %16"PRIxPTR " (line %d)\n",
+			ruby_node_name(type),
+			(intptr_t) node->u1.value, (intptr_t) node->u2.value, (intptr_t) node->u3.value,
+			nd_line(node));
 	}
 	else
 	{
-		rbstr_printf(str, "@ %s\n", ruby_node_name(type));
+		rbstr_printf(str, "@ %s (line %d)\n", ruby_node_name(type), nd_line(node));
 	}
 
 	ut[0] = nodes_ctbl[type * 3];
@@ -837,7 +860,7 @@ void print_node(VALUE str, NODE *node, int tab)
 		if (ut[i] == NT_NODE)
 		{
 			if (nd_type(node) != NODE_OP_ASGN2 || i != 2)
-				print_node(str, RNODE(uref[i]), tab + 1);
+				print_node(str, RNODE(uref[i]), tab + 1, show_offsets);
 			else
 			{
 				if (ut[i] != 0 && TYPE(ut[i]) != T_NODE)
@@ -855,7 +878,7 @@ void print_node(VALUE str, NODE *node, int tab)
 			PRINT_NODE_TAB; rbstr_printf(str, "  ");
 			if (show_offsets)
 			{
-				rbstr_printf(str, ">| ADR: %"PRIxPTR" TYPE: %d  TEXT: %s\n",
+				rbstr_printf(str, ">| ADR: %"PRIxPTR"; CLASS: %s (TYPE %d); VALUE: %s\n",
 					(intptr_t) uref[i],
 					class_name, TYPE(uref[i]),
 					RSTRING_PTR(rb_funcall(uref[i], rb_intern("to_s"), 0)));
@@ -869,11 +892,25 @@ void print_node(VALUE str, NODE *node, int tab)
 		}
 		else if (ut[i] == NT_ID)
 		{
+			const char *str_null = "<NULL>", *str_intern = "<NONAME>";
+			const char *str_sym;
 			PRINT_NODE_TAB; rbstr_printf(str, "  ");
+
+			if (uref[i] == 0)
+				str_sym = str_null;
+			else
+			{
+				VALUE rbstr_sym = rb_id2str(uref[i]);
+				if (TYPE(rbstr_sym) == T_STRING)
+					str_sym = RSTRING_PTR(rb_id2str(uref[i]));
+				else
+					str_sym = str_intern;
+			}
+
 			if (show_offsets)
-				rbstr_printf(str, ">| ID: %d; SYMBOL: %s\n", (ID) uref[i], RSTRING_PTR(rb_id2str(uref[i])));
+				rbstr_printf(str, ">| ID: %d; SYMBOL: :%s\n", (ID) uref[i], str_sym);
 			else
-				rbstr_printf(str, ">| SYMBOL: :%s\n", RSTRING_PTR(rb_id2str(uref[i])));
+				rbstr_printf(str, ">| SYMBOL: :%s\n", str_sym);
 		}
 		else if (ut[i] == NT_LONG)
 		{
@@ -927,7 +964,21 @@ void resolve_syms_ords(VALUE data, NODEObjAddresses *relocs)
 	relocs->syms_len = RARRAY_LEN(tbl_val);
 	relocs->syms_adr = ALLOC_N(ID, relocs->syms_len);
 	for (i = 0; i < relocs->syms_len; i++)
-		relocs->syms_adr[i] = rb_intern(RSTRING_PTR(RARRAY_PTR(tbl_val)[i]));
+	{
+		VALUE r_sym = RARRAY_PTR(tbl_val)[i];
+		if (TYPE(r_sym) == T_STRING)
+		{
+			relocs->syms_adr[i] = rb_intern(RSTRING_PTR(r_sym));
+		}
+		else if (TYPE(r_sym) == T_FIXNUM)
+		{
+			relocs->syms_adr[i] = (ID) FIX2INT(r_sym);
+		}
+		else
+		{
+			rb_raise(rb_eArgError, "Symbols table is corrupted");
+		}
+	}
 }
 
 void resolve_lits_ords(VALUE data, NODEObjAddresses *relocs)
@@ -1053,16 +1104,16 @@ void resolve_args_ords(VALUE data, NODEObjAddresses *relocs)
 			rb_raise(rb_eArgError, "args entry %d is corrupted", i);
 		}
 		// Load unresolved values
-		ainfo->pre_init = (NODE *) FIX2LONG(aiptr[0]); // Node ordinal
-		ainfo->post_init = (NODE *) FIX2LONG(aiptr[1]); // Node ordinal
+		ainfo->pre_init = (NODE *) (uintptr_t) FIX2LONG(aiptr[0]); // Node ordinal
+		ainfo->post_init = (NODE *) (uintptr_t) FIX2LONG(aiptr[1]); // Node ordinal
 		ainfo->pre_args_num = FIX2INT(aiptr[2]); // No ordinal resolving
 		ainfo->post_args_num = FIX2INT(aiptr[3]); // No ordinal resolving
 		ainfo->first_post_arg = FIX2INT(aiptr[4]); // Symbolic ordinal
 		ainfo->rest_arg = FIX2INT(aiptr[5]); // Symbolic ordinal
 		ainfo->block_arg = FIX2INT(aiptr[6]); // Symbolic ordinal
-		ainfo->kw_args = (NODE *) FIX2LONG(aiptr[7]); // Node ordinal
-		ainfo->kw_rest_arg = (NODE *) FIX2LONG(aiptr[8]); // Node ordinal
-		ainfo->opt_args = (NODE *) FIX2LONG(aiptr[9]); // Node ordinal
+		ainfo->kw_args = (NODE *) (uintptr_t) FIX2LONG(aiptr[7]); // Node ordinal
+		ainfo->kw_rest_arg = (NODE *) (uintptr_t) FIX2LONG(aiptr[8]); // Node ordinal
+		ainfo->opt_args = (NODE *) (uintptr_t) FIX2LONG(aiptr[9]); // Node ordinal
 		// Resolve nodes
 		ord = (int) (((VALUE) ainfo->pre_init) & 0xFFFFFFFF);
 		if (ord < -1 || ord >= relocs->nodes_len)
@@ -1107,6 +1158,23 @@ void resolve_args_ords(VALUE data, NODEObjAddresses *relocs)
 }
 #endif
 
+/*
+ * Transforms binary data with nodes descriptions into Ruby AST (i.e. 
+ * ternary tree of nodes). Each node is represented in the next binary format:
+ *
+ * [4 bytes -- pointers info] [node flags] [child ORD1] [child ORD2] [child ORD3]
+ * 
+ * Pointers info:
+ *   BYTE -- child 1 info (bits 7..4 -- ordinal type, bits 3..0 -- ordinal size, bytes)
+ *   BYTE -- child 2 info
+ *   BYTE -- child 3 info
+ *   BYTE -- node flags length, bytes
+ * Node flags:
+ *   node->flags field packed by bin_to_value function
+ *   child ORDi  Ordinal of ith node child packed by bin_to_value_function
+ *               (it will be transformed to the real address in memory, i.e. pointer
+ *                or symbol ID during data loading)
+ */
 void load_nodes_from_str(VALUE data, NODEObjAddresses *relocs)
 {
 	int i, j;
@@ -1215,7 +1283,7 @@ static VALUE check_hash_magic(VALUE data)
 	VALUE val, refval;
 	// MAGIC signature must be valid
 	val = get_hash_strfield(data, "MAGIC");
-	if (strcmp("NODEMARSHAL10", RSTRING_PTR(val)))
+	if (strcmp(NODEMARSHAL_MAGIC, RSTRING_PTR(val)))
 		rb_raise(rb_eArgError, "Bad value of MAGIC signature");
 	// RUBY_PLATFORM signature must match the current platform
 	val = get_hash_strfield(data, "RUBY_PLATFORM"); 
@@ -1307,6 +1375,9 @@ static VALUE m_nodedump_from_memory(VALUE self, VALUE dump)
 
 
 /*
+ * call-seq:
+ *   obj.symbols
+ *
  * Return array with the list of symbols
  */
 static VALUE m_nodedump_symbols(VALUE self)
@@ -1343,8 +1414,12 @@ static VALUE m_nodedump_symbols(VALUE self)
 }
 
 /*
- * Replace one symbol by another
- * (to be used for code obfuscation)
+ * call-seq:
+ *   obj.change_symbol(old_sym, new_sym)
+ *
+ * Replace one symbol by another (to be used for code obfuscation)
+ * - +old_sym+ -- String that contains symbol name to be replaced 
+ * - +new_sym+ -- String that contains new name of the symbol
  */
 static VALUE m_nodedump_change_symbol(VALUE self, VALUE old_sym, VALUE new_sym)
 {
@@ -1540,6 +1615,7 @@ static VALUE m_nodedump_from_string(VALUE self, VALUE str)
 static VALUE m_nodedump_init(VALUE self, VALUE source, VALUE info)
 {
 	ID id_usr;
+	rb_iv_set(self, "@show_offsets", Qfalse);
 	Check_Type(source, T_SYMBOL);
 	id_usr = SYM2ID(source);
 	if (id_usr == rb_intern("srcfile"))
@@ -1571,8 +1647,8 @@ static VALUE m_nodedump_init(VALUE self, VALUE source, VALUE info)
  * call-seq:
  *   obj.dump_tree
  * 
- * Transforms Ruby syntax tree (NODE) to the text string using
- * rb_parser_dump_tree function from node.c (see Ruby source code).
+ * Transforms Ruby syntax tree (NODE) to the String using
+ * +rb_parser_dump_tree+ function from +node.c+ (see Ruby source code).
  */
 static VALUE m_nodedump_parser_dump_tree(VALUE self)
 {
@@ -1581,16 +1657,52 @@ static VALUE m_nodedump_parser_dump_tree(VALUE self)
 }
 
 /*
- * Prints the node tree in the short variant
+ * call-seq:
+ *   obj.dump_tree_short
+ *
+ * Transforms Ruby syntax tree (NODE) to the String using custom function
+ * instead of +rb_parser_dump_tree+ function.
+ *
+ * See also #show_offsets, #show_offsets=
  */
 static VALUE m_nodedump_dump_tree_short(VALUE self)
 {
 	VALUE str = rb_str_new2(""); // Output string
 	NODE *node = RNODE(rb_iv_get(self, "@node"));
-	print_node(str, node, 0);
+	int show_offsets = (rb_iv_get(self, "@show_offsets") == Qtrue) ? 1 : 0;
+	print_node(str, node, 0, show_offsets);
 	return str;
 }
 
+/*
+ * call-seq:
+ *   obj.show_offsets
+ *
+ * Returns show_offsets property (used by NodeMarshal#dump_tree_short)
+ * It can be either true or false
+ */
+static VALUE m_nodedump_show_offsets(VALUE self)
+{
+	return rb_iv_get(self, "@show_offsets");
+}
+
+/*
+ * call-seq:
+ *   obj.show_offsets=
+ *
+ * Sets show_offsets property (used by NodeMarshal#dump_tree_short)
+ * It can be either true or false
+ */
+static VALUE m_nodedump_set_show_offsets(VALUE self, VALUE value)
+{
+	if (value != Qtrue && value != Qfalse)
+	{
+		rb_raise(rb_eArgError, "show_offsets property must be either true or false");
+	}
+	return rb_iv_set(self, "@show_offsets", value);
+}
+
+
 /*
  * call-seq:
  *   obj.to_hash
@@ -1604,7 +1716,7 @@ static VALUE m_nodedump_dump_tree_short(VALUE self)
  * 
  * <i>Part 1: Signatures</i>
  *
- * - <tt>MAGIC</tt> -- NODEMARSHAL10
+ * - <tt>MAGIC</tt> -- NODEMARSHAL11
  * - <tt>RUBY_PLATFORM</tt> -- saved <tt>RUBY_PLATFORM</tt> constant value
  * - <tt>RUBY_VERSION</tt>  -- saved <tt>RUBY_VERSION</tt> constant value
  *
@@ -1614,7 +1726,8 @@ static VALUE m_nodedump_dump_tree_short(VALUE self)
  * its identifier that is used in the node tree.
  *
  * - <tt>literals</tt> -- program literals (strings, ranges etc.)
- * - <tt>symbols</tt> -- program symbols
+ * - <tt>symbols</tt> -- program symbols (values have either String or Fixnum
+ *   data type; numbers are used for symbols that cannot be represented as strings)
  * - <tt>global_entries</tt> -- global variables information
  * - <tt>id_tables</tt> -- array of arrays. Each array contains symbols IDs
  * - <tt>args</tt> -- information about code block argument(s)
@@ -1867,7 +1980,7 @@ static VALUE m_base85r_decode(VALUE obj, VALUE input)
 }
 
 /* call-seq:
- *   obj.to_bin
+ *   obj.to_text
  * 
  * Converts NodeMarshal class example to the text string (modified Base85 encoding) that
  * can be saved to the file and used for loading the node from the file.
@@ -1888,7 +2001,6 @@ static VALUE m_nodedump_node(VALUE self)
 	return rb_iv_get(self, "@node");
 }
 
-
 /*
  * This class can load and save Ruby code in the form of the
  * platform-dependent syntax tree (made of NODEs). Such function
@@ -1917,6 +2029,8 @@ void Init_nodemarshal()
 	rb_define_method(cNodeMarshal, "dump_tree", RUBY_METHOD_FUNC(m_nodedump_parser_dump_tree), 0);
 	rb_define_method(cNodeMarshal, "dump_tree_short", RUBY_METHOD_FUNC(m_nodedump_dump_tree_short), 0);
 	rb_define_method(cNodeMarshal, "compile", RUBY_METHOD_FUNC(m_nodedump_compile), 0);
+	rb_define_method(cNodeMarshal, "show_offsets", RUBY_METHOD_FUNC(m_nodedump_show_offsets), 0);
+	rb_define_method(cNodeMarshal, "show_offsets=", RUBY_METHOD_FUNC(m_nodedump_set_show_offsets), 1);
 	// Methods for working with the information about the node
 	// a) literals, symbols, generic information
 	rb_define_method(cNodeMarshal, "symbols", RUBY_METHOD_FUNC(m_nodedump_symbols), 0);

data/ext/node-marshal/nodedump.h

@@ -14,6 +14,8 @@
 
 
 /* Some constants */
+// Magic value with the version of the format
+#define NODEMARSHAL_MAGIC "NODEMARSHAL11"
 // Type of the node "Child"
 #define NT_NULL 0
 #define NT_UNKNOWN 1

data/ext/node-marshal/nodeinfo.c

@@ -8,7 +8,7 @@
  * Fragments from Ruby source code are used here
  * (mainly from node.c, gc.c)
  *
- * License: 2-clause BSD
+ * License: BSD-2-Clause
  *
  */
 #include <stdio.h>
@@ -165,9 +165,10 @@ static int nodes_child_info[][4] =
 
 
 
-/** 
- * @brief Check the correctness of nodes table from the viewpoint
- * This function is based on Ruby 2.2.1 source code (node.c) */
+/*
+ * Check the correctness of nodes table from the viewpoint
+ * This function is based on Ruby source code (node.c)
+ */
 void check_nodes_child_info(int pos)
 {
 	int type = nodes_child_info[pos][0];
@@ -565,22 +566,36 @@ void check_nodes_child_info(int pos)
 	
 }
 
-
+/*
+ * Converts nodes_child_info 2D array of int into nodes_ctbl 1D array of int
+ *
+ * nodes_child_info is similar to hash by the structure: each 1D subarray
+ * has the next structure:
+ *   {NODE_ID, NT_CHILD1_TYPE, NT_CHILD2_TYPE, NT_CHILD3_TYPE}
+ */
 void init_nodes_table(int *nodes_ctbl, int num_of_entries)
 {
-	int pos, offset, i;
+	int pos, i;
+	/* Check the input array using information from Ruby source code */
 	for (pos = 0; nodes_child_info[pos][0] != -1; pos++)
 	{
 		check_nodes_child_info(pos);
 	}
-	
+	/* Initialize output array by NT_UNKNOWN (if node is not defined
+      in the input table the types of its childs are unknown) */	
 	for (i = 0; i < num_of_entries * 3; i++)
 	{
 		nodes_ctbl[i] = NT_UNKNOWN;
 	}
+	/* Fill output array */
 	for (pos = 0; nodes_child_info[pos][0] != -1; pos++)
 	{
-		offset = (nodes_child_info[pos][0]) * 3;
+		int index = nodes_child_info[pos][0], offset;
+		if (index < 0 || index > num_of_entries)
+		{
+			rb_raise(rb_eArgError, "NODE ID %d is out or nodes_ctbl array boundaries", index);
+		}
+		offset = index * 3;
 		nodes_ctbl[offset++] = nodes_child_info[pos][1];
 		nodes_ctbl[offset++] = nodes_child_info[pos][2];
 		nodes_ctbl[offset++] = nodes_child_info[pos][3];

data/lib/node-marshal.rb

@@ -31,7 +31,9 @@ class NodeMarshal
 	# - +opts+ -- Hash with options (+:compress+, +:so_path+)
 	#   +:compress+ can be +true+ or +false+, +:so_path+ is a test string 
 	#   with the command for nodemarshal.so inclusion (default is 
-	#   require_relative '../ext/node-marshal/nodemarshal.so')
+	#   <tt>require_relative '../ext/node-marshal/nodemarshal.so'</tt>)
+	#
+	# See also NodeMarshal::compile_rb_file
 	def to_compiled_rb(outfile, *args)
 		compress = true
 		so_path = "require_relative '../ext/node-marshal/nodemarshal.so'"
@@ -80,6 +82,12 @@ EOS
 		return txt
 	end
 
+	# call-seq:
+	#   NodeMarshal::compile_rb_file(outfile, inpfile, opts)
+	# 
+   # Reads +.rb+ file (Ruby source) and compiles it to .rb file containing
+	# compressed AST node and its loader. This functions is an envelope for
+	# NodeMarshal#to_compiled_rb
 	def self.compile_rb_file(outfile, inpfile, *args)
 		node = NodeMarshal.new(:srcfile, inpfile)
 		node.to_compiled_rb(outfile, *args)
@@ -117,7 +125,7 @@ EOS
 	end
 
 	# call-seq:
-	#   obj.get_safe_symbols
+	#   obj.get_safe_symbols(our_symbols)
 	#
 	# Returns an array that contains strings with the names of symbols that are safe
 	# to change. It excludes symbols that are present in the table of literals (and their derivatives
@@ -189,7 +197,7 @@ EOS
 	end
 
 	# call-seq:
-	#   obj.rebuld
+	#   obj.rebuild
 	#
 	# Rebuilds the node by converting it to the binary dump and further restoring
 	# of it from this dump. It doesn't change the original node and returns rebuilt

data/test/test_base.rb

@@ -37,7 +37,6 @@ end
 ni.map {|x| fact(x) }
 EOS
 		assert_node_compiler(program)		
-#		test_string_eval(program, [1, 2, 6, 24, 120, 720, 5040, 40320, 362880], "Factorial");
 	end
 
 	# Simple ROT13 task that combines several language construction
@@ -163,6 +162,27 @@ EOS
 		assert_node_compiler(program)
 	end
 
+	# Test expressions like x['a'] &&= true or x['b'] ||= false
+	# (they use NODE_OP_ASGN1 node and symbols that cannot be represented
+	# by String value)
+	def test_node_op_asgn1
+		program = <<EOS
+x = {a: [1234], b: [5678, 2],
+	and1: true, and2: true, and3: false, and4: false,
+	or1:  true, or2:  true, or3:  false, or4:  false}
+x[:a] &&= 'test'
+x[:b] ||= 'qqq'
+
+x[:and1] &&= false; x[:and2] &&= true
+x[:and3] &&= false; x[:and4] &&= true
+
+x[:or1] ||= false; x[:or2] ||= true
+x[:or3] ||= false; x[:or4] ||= true
+x
+EOS
+		assert_node_compiler(program)
+	end
+
 	# Check the reaction on the parsing errors during the node creation
 	# In the case of syntax error ArgumentError exception should be generated
 	def test_syntax_error

data/test/test_complex.rb

@@ -35,7 +35,7 @@ class TestHashTree
 			@depth = depth
 			@value = {:depth => depth, :data => [], :leaves =>[] }
 			data_size.times do
-				case rnd.rand(30000) % 6
+				case rnd.rand(30000) % 7
 				when 0
 					@value[:data] << 10**(rnd.rand(600000) * 0.001 - 300)
 				when 1
@@ -51,6 +51,9 @@ class TestHashTree
 				when 5
 					a, b = rnd.rand(100), rnd.rand(10)
 					@value[:data] << (a...(a + b))
+				when 6
+					rndsym = (Random.rand(1_000_000_000).to_s(36)).to_sym
+					@value[:data] << rndsym
 				else
 					nil
 				end
@@ -122,7 +125,7 @@ class TestComplex < Test::Unit::TestCase
 		puts "  Source code size: %d bytes" % tree_str.length
 		puts "  Binary data size: %d bytes" % tree_bin.length
 		# Clear the memory
-		node = nil; GC.start
+		node = nil; GC.enable; GC.start
 		# Load the node from the disk and turn it to tree
 		puts 'Loading the node...'
 		node = NodeMarshal.new(:binfile, 'node.bin')

data/test/test_qcall.rb

@@ -35,12 +35,17 @@ puts "'#{b&.owner_name}'"
 puts "'#{b&.owner_info&.address}'"
 [a&.owner_name, a&.owner_info&.address, b&.owner_name, b&.owner_info&.address]
 }
-
-		node = NodeMarshal.new(:srcmemory, qcall_program)
-		bindump =  node.to_bin
-		node = NodeMarshal.new(:binmemory, bindump)
-		res_node = node.compile.eval
-		res_text = eval(qcall_program)
-		assert_equal(res_text, res_node)
+		ver = RUBY_VERSION
+		ver = (ver[0] + ver[2] + ver[4]).to_i
+		if ver >= 230
+			node = NodeMarshal.new(:srcmemory, qcall_program)
+			bindump =  node.to_bin
+			node = NodeMarshal.new(:binmemory, bindump)
+			res_node = node.compile.eval
+			res_text = eval(qcall_program)
+			assert_equal(res_text, res_node)
+		else
+			assert_true(false, "Ruby 2.3 or higher is required for &. operator test")
+		end
 	end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: node-marshal
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.2.0
 platform: ruby
 authors:
 - Alexey Voskov
@@ -9,7 +9,7 @@ autorequire:
 bindir:
 - bin
 cert_chain: []
-date: 2015-12-24 00:00:00.000000000 Z
+date: 2016-01-11 00:00:00.000000000 Z
 dependencies: []
 description: "This gem is designed for transformation of Ruby source code (eiher in
   the form of files or strings) to the \nRuby nodes (syntax trees) used by Ruby MRI
@@ -25,10 +25,10 @@ extensions:
 extra_rdoc_files:
 - README.rdoc
 files:
+- COPYING
 - README.rdoc
 - bin/noderbc
 - bin/noderbc.bat
-- ext/node-marshal/COPYING
 - ext/node-marshal/base85r.c
 - ext/node-marshal/extconf.rb
 - ext/node-marshal/libobj/readme.txt
@@ -46,9 +46,9 @@ files:
 - test/test_obfuscator.rb
 - test/test_qcall.rb
 - test/tinytet.rb
-homepage: 
+homepage: https://github.com/dig386/node-marshal
 licenses:
-- 2-clause BSD
+- BSD-2-Clause
 metadata: {}
 post_install_message: 
 rdoc_options: