java-llvm/classfile.h

// types for Java's Classfile structure
// https://docs.oracle.com/javase/specs/jvms/se22/html/jvms-4.html

#include <stdint.h>

#pragma pack(1)

typedef uint8_t u1;
typedef uint16_t u2;
typedef uint32_t u4;
typedef uint64_t u8;

typedef enum {
  ACC_PUBLIC = 0x0001,
  ACC_PRIVATE = 0x0002,
  ACC_PROTECTED = 0x0004,
  ACC_STATIC = 0x0008,
  ACC_FINAL = 0x0010,
  ACC_SUPER = 0x0020,
  ACC_SYNCHRONIZED = 0x0020,
  ACC_VOLATLE = 0x0040,
  ACC_BRIDGE = 0x0040,
  ACC_TRANSIENT = 0x0080,
  ACC_VARARGS = 0x0080,
  ACC_NATIVE = 0x0100,
  ACC_INTERFACE = 0x0200,
  ACC_ABSTRACT = 0x0400,
  ACC_STRICT = 0x0800,
  ACC_SYNTHETIC = 0x1000,
  ACC_ANNOTATION = 0x2000,
  ACC_ENUM = 0x4000,
  ACC_MODULE = 0x8000,
} ACCESS_FLAGS;

typedef enum {
  CONSTANT_Class = 7,
  CONSTANT_Fieldref = 9,
  CONSTANT_Methodref = 10,
  CONSTANT_InterfaceMethodref = 11,
  CONSTANT_String = 8,
  CONSTANT_Integer = 3,
  CONSTANT_Float = 4,
  CONSTANT_Long = 5,
  CONSTANT_Double = 6,
  CONSTANT_NameAndType = 12,
  CONSTANT_Utf8 = 1,
  CONSTANT_MethodHandle = 15,
  CONSTANT_MethodType = 16,
  CONSTANT_Dynamic = 17,
  CONSTANT_InvokeDynamic = 18,
  CONSTANT_Module = 19,
  CONSTANT_Package = 20,
} CONSTANT_TAGS;

static const char* CONSTANT_TAGS_STRINGS[] = {"",
                                              "UTF-8",
                                              "",
                                              "Interger",
                                              "Float",
                                              "Long",
                                              "Double",
                                              "Class",
                                              "String",
                                              "Fieldref",
                                              "Methodref",
                                              "InterfaceMethodref",
                                              "NameAndType",
                                              "",
                                              "",
                                              "MethodHandle",
                                              "MethodType",
                                              "Dynamic",
                                              "InvokeDynamic",
                                              "Module",
                                              "Package"};

typedef struct {
  u2 name_index;
} CONSTANT_Class_info;

typedef struct {
  u2 class_index;
  u2 name_and_type_index;
} CONSTANT_Fieldref_info;
typedef CONSTANT_Fieldref_info CONSTANT_Methodref_info;
typedef CONSTANT_Fieldref_info CONSTANT_InterfaceMethodref_info;

typedef struct {
  u2 string_index;
} CONSTANT_String_info;

typedef struct {
  u4 bytes;
} CONSTANT_Integer_info;
typedef CONSTANT_Integer_info CONSTANT_Float_info;

typedef struct {
  u4 high_bytes;
  u4 low_bytes;
} CONSTANT_Long_info;
typedef CONSTANT_Long_info CONSTANT_Double_info;

typedef struct {
  u2 name_index;
  u2 descriptor_index;
} CONSTANT_NameAndType_info;

typedef struct {
  u2 length;
  u1* bytes;
} CONSTANT_Utf8_info;

typedef struct {
  u1 reference_kind;
  u2 reference_index;
} CONSTANT_MethodHandle_info;

typedef struct {
  u2 descriptor_index;
} CONSTANT_MethodType_info;

typedef struct {
  u2 bootstrap_method_attr_index;
  u2 name_and_type_index;
} CONSTANT_Dynamic_info;
typedef CONSTANT_Dynamic_info CONSTANT_InvokeDynamic_info;

typedef struct {
  u2 name_index;
} CONSTANT_Module_info;

typedef struct {
  u2 name_index;
} CONSTANT_Package_info;

typedef struct {
  u1 tag;
  union {
    CONSTANT_Class_info class_info;
    CONSTANT_Fieldref_info fieldref_info;
    CONSTANT_Methodref_info methodref_info;
    CONSTANT_InterfaceMethodref_info interface_methodref_info;
    CONSTANT_String_info string_info;
    CONSTANT_Integer_info integer_info;
    CONSTANT_Float_info float_info;
    CONSTANT_Long_info long_info;
    CONSTANT_Double_info double_info;
    CONSTANT_NameAndType_info name_and_type_info;
    CONSTANT_Utf8_info utf8_info;
    CONSTANT_MethodHandle_info method_handle_info;
    CONSTANT_MethodType_info method_type_info;
    CONSTANT_Dynamic_info dynamic_info;
    CONSTANT_InvokeDynamic_info invoke_dynamic_info;
    CONSTANT_Module_info module_info;
    CONSTANT_Package_info package_info;
  } info;
} cp_info;

typedef struct {
  u2 constantvalue_index;
} ConstantValue_attribute;

typedef struct {
  u2 max_stack;
  u2 max_locals;
  u4 code_length;
  u1* code;
  u2 exception_table_length;
  struct {
    u2 start_pc;
    u2 end_pc;
    u2 handler_pc;
    u2 catch_type;
  } * exception_table;
  u2 attributes_count;
  void* attributes;
} Code_attribute;

typedef enum {
  ITEM_Top = 0,
  ITEM_Integer = 1,
  ITEM_Float = 2,
  ITEM_Double = 3,
  ITEM_Long = 4,
  ITEM_Null = 5,
  ITEM_UninitializedThis = 6,
  ITEM_Object = 7,
  ITEM_Uninitialized = 8,
} VERIFICATION_TAG;

typedef struct {
  u1 tag;
  union {
    struct {
    } Top_variable_info;
    struct {
    } Integer_variable_info;
    struct {
    } Float_variable_info;
    struct {
    } Long_variable_info;
    struct {
    } Double_variable_info;
    struct {
    } Null_variable_info;
    struct {
    } UninitializedThis_variable_info;
    struct {
      u2 cpool_index;
    } Object_variable_info;
    struct {
      u2 offset;
    } Uninitialized_variable_info;
  };
} verification_type_info;

typedef enum {
  SAME,
  SAME_LOCALS_1_STACK_ITEM_FRAME,
  SAME_LOCALS_1_STACK_ITEM_FRAME_EXTENDED,
  CHOP,
  SAME_FRAME_EXTENDED,
  APPEND,
  FULL_FRAME,
} stack_frame_type;

typedef struct {
  u1 frame_type;
  union {
    struct {
    } same_frame; // 0-63
    struct {
      verification_type_info stack[1];
    } same_locals_1_stack_item_frame; // 64-127
    struct {
      u2 offset_delta;
      verification_type_info stack[1];
    } same_locals_1_stack_item_frame_extended; // 247
    struct {
      u2 offset_delta;
    } chop_frame; // 248-250
    struct {
      u2 offset_delta;
    } same_frame_extended; // 251
    struct {
      u2 offset_delta;
      verification_type_info* locals; // [frame_type - 251]
    } append_frame; // 252-254
    struct {
      u2 offset_delta;
      u2 number_of_locals;
      verification_type_info* locals;
      u2 number_of_stack_items;
      verification_type_info* stack;
    } full_frame; // 255
  };
} stack_map_frame;

typedef struct {
  u2 number_of_entries;
  stack_map_frame* entries;
} StackMapTable_attribute;

typedef struct {
  u2 bootstrap_method_ref;
  u2 num_bootstrap_arguments;
  u2* bootstrap_arguments;
} BootstrapMethod;

typedef struct {
  u2 num_bootstrap_methods;
  BootstrapMethod* bootstrap_methods;
} BootstrapMethods_attribute;

typedef struct {
  u2 host_class_index;
} NestHost_attribute;

typedef struct {
  u2 number_of_classes;
  u2* classes;
} NestMembers_attribute;

typedef struct {
  u2 number_of_classes;
  u2* classes;
} PermittedSubclasses_attribute;

typedef enum {
  UNKNOWN,
  CONSTANT_VALUE,
  CODE,
  STACK_MAP_TABLE,
  BOOTSTRAP_METHODS,
  NEST_HOST,
  NEST_MEMBERS,
  PERMITTED_SUBCLASSES,
  ATTRIBUTE_TAGS_END
} ATTRIBUTE_TAGS;

static const char* ATTRIBUTE_TAGS_STRINGS[] = {
    "ConstantValue", "Code", "StackMapTable", "BootstrapMethods", "NestHost", "NestMembers", "PermittedSubclasses",
};

typedef struct {
  u2 attribute_name_index;
  u4 attribute_length;
  ATTRIBUTE_TAGS _tag;
  union {
    ConstantValue_attribute constant_value_attribute;
    Code_attribute code_attribute;
    StackMapTable_attribute stack_map_table_attribute;
    BootstrapMethods_attribute bootstrap_methods_attribute;
    NestHost_attribute nest_host_attribute;
    NestMembers_attribute nest_members_attribute;
    PermittedSubclasses_attribute permitted_subclasses_attribute;
  } info;
} attribute_info;

typedef struct {
  u2 access_flags;
  u2 name_index;
  u2 descriptor_index;
  u2 attributes_count;
  attribute_info* attributes;
} field_info;

typedef struct {
  u2 access_flags;
  u2 name_index;
  u2 descriptor_index;
  u2 attributes_count;
  attribute_info* attributes;
} method_info;

typedef struct {
  u4 magic;
  u2 minor_version;
  u2 major_version;
  u2 constant_pool_count;
  cp_info* constant_pool;
  u2 access_flags;
  u2 this_class;
  u2 super_class;
  u2 interfaces_count;
  u2* interfaces;
  u2 fields_count;
  field_info* fields;
  u2 methods_count;
  method_info* methods;
  u2 attributes_count;
  attribute_info* attributes;
} ClassFile;

ClassFile* ClassFile_load(const char* path);
void ClassFile_info(const ClassFile* cf);

char* ClassFile_resolve_Fieldref_Methodref_InterfaceMethodref(const ClassFile* cf, CONSTANT_Fieldref_info info);
char* ClassFile_resolve_NameAnyType(const ClassFile* cf, CONSTANT_NameAndType_info info);

#pragma pack()