1. 1 Java.class File Format 陳正佳

2. 2 Java Virtual Machine the cornerstone of Sun's Java programming language. a component of the Java technology responsible for A. Java's cross-platform delivery, B. the small size of its compiled code, C. Java's ability to protect users from malicious programs. JVM knows nothing of the Java programming language, only of a particular file format, the class file format.

3. 3 2 Class file format l A class file contains »1. Java Virtual Machine instructions (bytecodes) »2. a symbol table (Constant pool) »3. other ancillary information. l Javap options »1. -c bytecode »2. -s internal type signatures »3. -verbose stack size and number of local variables and args.

4. 4 2 Class file format l Each class file contains one Java type, either a class or an interface. l A class file consists of a stream of 8-bit bytes. l All 16-bit, 32-bit, and 64-bit quantities are constructed by reading in two, four, and eight consecutive 8-bit bytes, respectively.

5. 5 3 Class File Format l Multibyte data items are always stored in big- endian order, where the high bytes come first. l u1, u2,and u4 represent an unsigned one-, two-, or four-byte quantity, respectively. l These types may be read by methods such as readUnsignedByte, readUnsignedShort, and readInt of the interface java.io.DataInput.

6. 6 class SumI { public static void main (String[] args) { int count=10; int sum =0; for (int index=1;index<count;index++) sum=sum+index; System.out.println("Sum="+sum); } // method main }

7. 7 4 Class File Structure ClassFile { u4 magic; u2 minor_version; u2 major_version; u2 constant_pool_count; cp_info constant_pool[constant_pool_count-1]; u2 access_flags; u2 this_class; u2 super_class; u2 interfaces_count; u2 interfaces[interfaces_count]; u2 fields_count; field_info fields[fields_count]; u2 methods_count; method_info methods[methods_count]; u2 attributes_count; attribute_info attributes[attributes_count]; }

8. 8 5 Magic magic :u4 The magic item supplies the magic number identifying the class file format; it has the value 0xCAFEBABE.

9. 9 5 Version minor_version :u2, major_version:u2 The values of the minor_version and major_version items are the minor and major version numbers of the compiler that produced this class file.

10. 10 5 Constant Pool constant_pool_count :u2 »must > 0. »indicates the number of entries in the constant_pool table of the class file, »where the constant_pool entry at index zero is included in the count but is not present in the constant_pool table of the class file. »i.e., if count = 13 => pool[1] … pool[12]

11. 11 5 Constant Pool constant_pool[] »a table of variable-length structures »representing various –numeric literals –string constants, –class/interface/type names, –field reference, –method reference and –other constants that are referred to within the ClassFile structure and its substructures.

12. 12 5 Constant Pool l constant_pool[0], »reserved for internal use by a JVM implementation. That entry is not present in the class file. l The first entry in the class file is constant_pool[1]. l Each constant_pool[i] is a variable-length structure whose format is indicated by its first "tag" byte.

13. 13 5 Access Flags access_flags :u2 »a mask of modifiers used with class and interface declarations.

14. 14 5 Access Flags l The access_flags modifiers Flag Name Value Meaning Used By ACC_PUBLIC 0x0001 Is public; Class, interface ACC_FINAL 0x0010 Is final; Class ACC_SUPER 0x0020 superclass Class, =1 for new JVM interface ACC_INTERFACE 0x0200 interface. Interface ACC_ABSTRACT 0x0400 Is abstract; Class, interface

15. 15 5 Access Flags l The access_flags modifiers Flag Name Value Meaning Used By ACC_PUBLIC 0x0001 Is public; Class, interface ACC_FINAL 0x0010 Is final; Class ACC_SUPER 0x0020 superclass Class, =1 for new JVM interface ACC_INTERFACE 0x0200 interface. Interface ACC_ABSTRACT 0x0400 Is abstract; Class, interface

16. 16 5 Access Flags (continued) l The access_flags modifiers Flag Name Value Meaning Used By ACC_SYNTHETIC 0x1000 synthetic ; Not present in the source code. ACC_ANNOTATION 0x2000 annotation type. ACC_ENUM 0x4000 enum type.

17. 17 5 this class this_class :u2 »a valid index into the constant_pool table. »The entry at that index must be a CONSTANT_Class_info structure representing the class or interface defined by this class file.

18. 18 5 super_class » For a class, the value of the super_class item either must be zero or must be a valid index into the constant_pool table. »If the value of the super_class item is nonzero, the constant_pool entry at that index must be a CONSTANT_Class_info structure representing the superclass of the class defined by this class file.

19. 19 5 super_class »Neither the superclass nor any of its superclasses may be a final class. »If the value of super_class is zero, then this class file must represent the class java.lang.Object, »the only class or interface without a superclass.

20. 20 5 interfaces interfaces_count »the number of direct superinterfaces of this class or interface type. interfaces[] » Each value in the interfaces array must be a valid index into the constant_pool table. »The constant_pool entry at each value of interfaces[i], where i < interfaces_count, must be a CONSTANT_Class_info structure

21. 21 5 Fields fields_count » gives the number of field_info structures in the fields table. fields[] » Each entry a variable-length field_info structure giving a complete description of a field in the class or interface type. »includes only those fields that are declared by this class or interface. »does not include fields inherited from superclasses or superinterfaces.

22. 22 5 Methods methods_count »gives the number of method_info structures in the methods table. methods[] »Each entry a variable-length method_info structure giving a complete description of Java Virtual Machine code for a method in the class or interface. »The method_info structures represent all methods, both instance methods, class (static) methods, and constructor methods declared by this class or interface type.

23. 23 5 Methods methods[] »includes only those methods explicitly declared by this class. »Interfaces have only the single method, the interface initialization method. »Constructor methods have the a common name. »does not include items representing methods that are inherited from superclasses or superinterfaces.

24. 24 5 Class Attributes attributes_count »the number of attributes in the attributes table of this class attributes[] » Each value of the attributes table must be a variable-length attribute structure. »A ClassFile structure can have any number of attributes associated with it.

25. 25 Internal form of fully qualified names l Replace all dots with /. EX: a.b.C ==> a/b/C

26. 26 Descriptor l A descriptor is a string representing the type of a field or method. Descriptors are represented in the class file format using UTF-8 strings. l Grammar: FieldType ::= BaseType | ObjectType | ArrayType

27. 27 Fields Descriptors l BaseType ::= B | C | D | F | I | J | S | Z »B for byte; Z for boolean »C D F I J S for char, double, float, int, long and short, respectively. l ObjectType ::= Lclassname; l ArrayType ::= [ComponentType l Ex: »int[][] ==> [[I »Object[] ==> [java/lang/Object;

28. 28 Field and method descriptors l FieldDescriptor ::= FieldType l ComponentType ::= FieldType l MethodDescriptor ::= ( ParameterDescriptor* ) ReturnDescriptor l ParameterDescriptor ::= FieldDescriptor l ReturnDescriotor ::= FieldDescriptor | V »V for void method l Ex: 1. Object mymethod(int i, double d, Thread t) ==> (IDLjava/lang/Thread;)Ljava/lang/Object; 2. void com.Clazz.m(int i) ==> (I)V.

29. 29 The Constant Pool l JVM instructions do not rely on the runtime layout of classes, interfaces, class instances, or arrays. Instead, instructions refer to symbolic information in the constant_pool table. All constant_pool table entries have the following general format: cp_info { u1 tag; u1 info[]; }

30. 30 Constant pool Tags Constant Type Value CONSTANT_Utf8 1 CONSTANT_Integer 3 CONSTANT_Float 4 CONSTANT_Long 5 CONSTANT_Double 6 CONSTANT_Class 7 CONSTANT_String 8 CONSTANT_Fieldref 9 CONSTANT_Methodref 10 CONSTANT_InterfaceMethodref 11 CONSTANT_NameAndType 12

31. 31 The CONSTANT_Utf8_info Structure l Used to represent constant string values. l utf-8 encoding : 1. 1~127 (xxx xxxx) ==> 1 bytes : 0xxx xxxx 2. 0, \u0080~\u07ff ( xxx xxxx xxxx) ==> 2 bytes: 110x xxxx 10xx xxxx 0 => 1100 0000 1000 0000 3. \u0800 ~ \uffff (xxxx xxxx xxxx xxxx) ==> 3 bytes: 1110 xxxx 10xx xxxx 10xx xxxx

32. 32 Format CONSTANT_Utf8_info { u1 tag; // = 1 u2 length; u1 bytes[length]; } Notes: l tag = 1. l length : number of bytes in the encoded utf8 string. bytes[] l bytes[] : contains the bytes of the string. l No byte may have the value (byte)0 or lie in the range (byte)0xf0-(byte)0xff.

33. 33 The CONSTANT_String_info Structure CONSTANT_String_info { u1 tag; // = 8 u2 string_index; } l string_index : »a valid index into a CONSTANT_Utf8_info

34. 34 CONSTANT_Integer_info and CONSTANT_Float_info l CONSTANT_Integer_info { u1 tag; // = 3 u4 bytes; } l CONSTANT_Float_info { u1 tag; // =4 u4 bytes; }

35. 35 CONSTANT_Long_info and CONSTANT_Double_info l CONSTANT_Long_info { u1 tag; // = 5 u8 bytes; } l CONSTANT_Double_info { u1 tag; // =6 u8 bytes; }

36. 36 CONSTANT_Class_info l Used to represent a class or an interface: l CONSTANT_Class_info { u1 tag; // = 7 u2 name_index; } l name_index is an index into an utf-8 entry encoding a fully qualified class name. l include also array type: [[I, [Lcom/Cl;, etc.

37. 37 CONSTANT_NameAndType_info l used to represent a field or method, without indicating which class or interface type it belongs to: CONSTANT_NameAndType_info { u1 tag; // = 12 u2 name_index; // index into utf8 string of a simple name u2 descriptor_index; // index into utf8 string of a filed/method descriptor }

38. 38 Constant fieldref, methodRef and interfaceMethodRef info l CONSTANT_Fieldref_info { u1 tag; // = 9 u2 class_index; // index into a class_info entry u2 name_and_type_index; } l CONSTANT_Methodref_info { u1 tag; // = 10 u2 class_index; u2 name_and_type_index; } l CONSTANT_InterfaceMethodref_info { u1 tag; //11 u2 class_index; u2 name_and_type_index; }

39. 39 Fields l Each field is described by a field_info structure. l No two fields in one class file may have the same name and descriptor. l Format: l field_info { » u2 access_flags; » u2 name_index; // index to utf8 entry for name » u2 descriptor_index;// index to uft8 entry for field type » u2 attributes_count; » attribute_info attributes[attributes_count]; }

40. 40 Methods l Each method, including each instance initialization method( ) and the class or interface initialization method( ), is described by a method_info structure. l Format: method_info { u2 access_flags; u2 name_index; // index into utf8 entry for name u2 descriptor_index; // index into utf8 entry u2 attributes_count; attribute_info attributes[attributes_count]; }

41. 41 Attributes l Attributes are used in the ClassFile, field_info, method_info, and Code_attribute structures of the class file format. l General Format: Attribute_info { u2 attribute_name_index; // into utf8 u4 attribute_length; // excluding init 6 bytes u1 info[attribute_length]; }

42. 42 Types of attributes l class attributes: »synthetic, deprecated, sourceFile »innerClasses l field attributes: »constantValue, deprecated l method attributes: »code attributes, deprecated, »Exception attributes l code attributes: »lineNumberTable, localVariableTable

43. 43 ConstantValue Attribute l ConstantValue_attribute { u2 attribute_name_index; // index to utf8 : “ conatantValue ” u4 attribute_length; // = 2 u2 constantvalue_index; // into primitive(Integer_entry, … ) // or string_entry }

44. 44 Code Attribute l a variable-length attribute used in the attribute table of method_info structures. l A Code attribute contains »the Java virtual machine instructions and »auxiliary information for a single method, »instance initialization method, or class or interface initialization method. l Every JVM implementation must recognize Code attributes. l native or abstract, => no this attribute in its method_info structure »O/W=> has exactly one Code attribute.

45. 45 Code_attribute { u2 attribute_name_index; // into utf8: “ Code ” u4 attribute_length; // exclude initial 6 bytes u2 max_stack;// max # of stack slots needed u2 max_locals; // maximum # of local variables needed // Note: double and long count 2. u4 code_length; u1 code[code_length];

46. 46 u2 exception_table_length; { u2 start_pc; u2 end_pc; //exclusive u2 handler_pc; u2 catch_type; // index to cp entry f type constant_class_info // zero means called by all exceptions // used to implement finally-clause } exception_table[exception_table_length]; u2 attributes_count; attribute_info attributes[attributes_count]; }

47. 47 The (uncaught checked) Exception Attribute l appear in the attribute table of a method_info structure. l indicates which checked exceptions a method may throw. l  1 Exception attribute in each method_info structure. l Format: Exceptions_attribute { u2 attribute_name_index; // into utf8: ” Exception ” u4 attribute_length; u2 number_of_exceptions; u2 exception_index_table[number_of_exceptions]; // each entry an index to cp of class_info }

48. 48 The InnerClasses Attribute l a variable-length attribute in the attributes table of the ClassFile structure. l record any inner class/interface referred (or declared) in this class/interface »If the constant pool of a class or interface refers to any class or interface that is not a member of a package, its ClassFile structure must have exactly one InnerClasses attribute in its attributes table. »If a class has members that are classes or interfaces, its constant_pool table (and hence its InnerClasses attribute) must refer to each such member, even if that member is not otherwise mentioned by the class.

49. 49 Format: InnerClasses_attribute { u2 attribute_name_index; // to utf8: “ innerClasses u4 attribute_length; u2 number_of_classes; { u2 inner_class_info_index; // into class_info // for each innerclass C in cp u2 outer_class_info_index; // into class_info // containing class of C u2 inner_name_index; // into utf8 for simple // name (no / no $); anonymous => zero u2 inner_class_access_flags; } classes[number_of_classes]; }

50. 50 The Syntheticand Deprecated Attribute l A class member that does not appear in the source code must be marked using a Synthetic attribute. Synthetic_attribute { »u2 attribute_name_index; // utf8(“synthetic”) »u4 attribute_length; // = zero } l The Deprecated attribute has the following format: Deprecated_attribute { »u2 attribute_name_index; // utf8( “ Deprecated ” ) »u4 attribute_length; // = 0 }

51. 51 The SourceFile Attribute SourceFile_attribute { u2 attribute_name_index; // into utf8( “ SourceFile ” ) u4 attribute_length; // = 2 u2 sourcefile_index; // into utf8 entry }

52. 52 LineNumberTable Attribute l LineNumberTable_attribute { »u2 attribute_name_index ; // utf8(“LineNumberTable”) »u4 attribute_length; »u2 line_number_table_length; »{ u2 start_pc; » u2 line_number; } line_number_table[line_number_table_length]; » }

53. 53 The LocalVariableTable Attribute l LocalVariableTable_attribute { »u2 attribute_name_index; // utf8(“LocalVariableTable”) »u4 attribute_length; »u2 local_variable_table_length; »{ u2 start_pc; u2 length; // scope must have » // value in [start_pc, start_pc + length] » u2 name_index; // utf8 for var name » u2 descriptor_index; // uft8 for type » u2 index; // local var index » } local_variable_table[local_variable_table_length]; }