THRIFT-5910: Add UUID support in Ruby

.github/workflows/build.yml

@@ -641,6 +641,45 @@ jobs:
           path: lib/cpp/test/*.xml
           retention-days: 3
 
+  lib-ruby:
+    needs: compiler
+    runs-on: ubuntu-24.04
+    strategy:
+      matrix:
+        ruby-version: ["2.7", "3.0", "3.1", "3.2", "3.3", "3.4", "4.0.0-preview2"]
+      fail-fast: false
+    steps:
+      - uses: actions/checkout@v5
+
+      - name: Set up Ruby
+        uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby-version }}
+          bundler-cache: true
+          working-directory: lib/rb
+
+      - name: Run bootstrap
+        run: ./bootstrap.sh
+
+      - name: Run configure
+        run: |
+          ./configure $(echo $CONFIG_ARGS_FOR_LIBS | sed 's/without-ruby/with-ruby/')
+
+      - uses: actions/download-artifact@v6
+        with:
+          name: thrift-compiler
+          path: compiler/cpp
+
+      - name: Run thrift-compiler
+        run: |
+          chmod a+x compiler/cpp/thrift
+          compiler/cpp/thrift -version
+
+      - name: Run tests
+        run: |
+          cd lib/rb
+          bundle exec rake spec
+
   cross-test:
     needs:
       - lib-java-kotlin

LANGUAGES.md

@@ -312,8 +312,8 @@ Thrift's core protocol is TBinary, supported by all languages except for JavaScr
 <td align=left><a href="https://github.com/apache/thrift/blob/master/lib/rb/README.md">Ruby</a></td>
 <!-- Since -----------------><td>0.2.0</td>
 <!-- Build Systems ---------><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cred.png" alt=""/></td>
-<!-- Language Levels -------><td>2.3.1p112</td><td>2.5.1p57</td>
-<!-- Field types -----------><td><img src="/doc/images/cred.png" alt=""/></td>
+<!-- Language Levels -------><td>2.7.8</td><td>4.0.0</td>
+<!-- Field types -----------><td><img src="/doc/images/cgrn.png" alt="Yes"/></td>
 <!-- Low-Level Transports --><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cred.png" alt=""/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cred.png" alt=""/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td>
 <!-- Transport Wrappers ----><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cred.png" alt=""/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cred.png" alt=""/></td>
 <!-- Protocols -------------><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td><td><img src="/doc/images/cgrn.png" alt="Yes"/></td>

compiler/cpp/src/thrift/generate/t_rb_generator.cc

@@ -430,6 +430,9 @@ t_rb_ofstream& t_rb_generator::render_const_value(t_rb_ofstream& out,
     case t_base_type::TYPE_STRING:
       out << "%q\"" << get_escaped_string(value) << '"';
       break;
+    case t_base_type::TYPE_UUID:
+      out << "%q\"" << get_escaped_string(value) << '"';
+      break;
     case t_base_type::TYPE_BOOL:
       out << (value->get_integer() > 0 ? "true" : "false");
       break;
@@ -1175,6 +1178,8 @@ string t_rb_generator::type_to_enum(t_type* type) {
       return "::Thrift::Types::I64";
     case t_base_type::TYPE_DOUBLE:
       return "::Thrift::Types::DOUBLE";
+    case t_base_type::TYPE_UUID:
+      return "::Thrift::Types::UUID";
     default:
       throw "compiler error: unhandled type";
     }

lib/rb/ext/binary_protocol_accelerated.c

@@ -20,6 +20,7 @@
 #include <ruby.h>
 #include <stdbool.h>
 #include <stdint.h>
+#include <string.h>
 #include <constants.h>
 #include <struct.h>
 #include <macros.h>
@@ -86,6 +87,19 @@ static void write_string_direct(VALUE trans, VALUE str) {
   rb_funcall(trans, write_method_id, 1, str);
 }
 
+// Efficient hex character to integer conversion
+static inline int hex_char_to_int(char c) {
+  if (c >= '0' && c <= '9') return c - '0';
+  if (c >= 'a' && c <= 'f') return c - 'a' + 10;
+  if (c >= 'A' && c <= 'F') return c - 'A' + 10;
+  return -1;  // invalid hex character
+}
+
+// Efficient integer to hex character conversion
+static inline char int_to_hex_char(int val) {
+  return val < 10 ? ('0' + val) : ('a' + val - 10);
+}
+
 //--------------------------------
 // interface writing methods
 //--------------------------------
@@ -228,6 +242,55 @@ VALUE rb_thrift_binary_proto_write_binary(VALUE self, VALUE buf) {
   return Qnil;
 }
 
+VALUE rb_thrift_binary_proto_write_uuid(VALUE self, VALUE uuid) {
+  CHECK_NIL(uuid);
+
+  if (TYPE(uuid) != T_STRING) {
+    rb_raise(rb_eStandardError, "UUID must be a string");
+  }
+
+  VALUE trans = GET_TRANSPORT(self);
+  char bytes[16];
+  const char* str = RSTRING_PTR(uuid);
+  long len = RSTRING_LEN(uuid);
+
+  // Parse UUID string (format: "550e8400-e29b-41d4-a716-446655440000")
+  // Expected length: 36 characters (32 hex + 4 hyphens)
+  if (len != 36 || str[8] != '-' || str[13] != '-' || str[18] != '-' || str[23] != '-') {
+    VALUE exception_class = rb_const_get(thrift_module, rb_intern("ProtocolException"));
+    VALUE invalid_data = rb_const_get(exception_class, rb_intern("INVALID_DATA"));
+    VALUE args[2];
+    args[0] = invalid_data;
+    args[1] = rb_str_new2("Invalid UUID format");
+    rb_exc_raise(rb_class_new_instance(2, args, exception_class));
+  }
+
+  // Parse hex string to bytes using direct conversion, skipping hyphens
+  int byte_idx = 0;
+  for (int i = 0; i < len && byte_idx < 16; i++) {
+    if (str[i] == '-') continue;
+
+    // Convert two hex characters to one byte
+    int high = hex_char_to_int(str[i]);
+    int low = hex_char_to_int(str[i + 1]);
+
+    if (high < 0 || low < 0) {
+      VALUE exception_class = rb_const_get(thrift_module, rb_intern("ProtocolException"));
+      VALUE invalid_data = rb_const_get(exception_class, rb_intern("INVALID_DATA"));
+      VALUE args[2];
+      args[0] = invalid_data;
+      args[1] = rb_str_new2("Invalid hex character in UUID");
+      rb_exc_raise(rb_class_new_instance(2, args, exception_class));
+    }
+
+    bytes[byte_idx++] = (unsigned char)((high << 4) | low);
+    i++; // skip next char since we processed two
+  }
+
+  WRITE(trans, bytes, 16);
+  return Qnil;
+}
+
 //---------------------------------------
 // interface reading methods
 //---------------------------------------
@@ -400,6 +463,28 @@ VALUE rb_thrift_binary_proto_read_binary(VALUE self) {
   return READ(self, size);
 }
 
+VALUE rb_thrift_binary_proto_read_uuid(VALUE self) {
+  unsigned char bytes[16];
+  VALUE data = READ(self, 16);
+  memcpy(bytes, RSTRING_PTR(data), 16);
+
+  // Format as UUID string: "XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX"
+  char uuid_str[37];
+  char* p = uuid_str;
+
+  for (int i = 0; i < 16; i++) {
+    *p++ = int_to_hex_char((bytes[i] >> 4) & 0x0F);
+    *p++ = int_to_hex_char(bytes[i] & 0x0F);
+    if (i == 3 || i == 5 || i == 7 || i == 9) {
+      *p++ = '-';
+    }
+  }
+
+  *p = '\0';
+
+  return rb_str_new(uuid_str, 36);
+}
+
 void Init_binary_protocol_accelerated() {
   VALUE thrift_binary_protocol_class = rb_const_get(thrift_module, rb_intern("BinaryProtocol"));
 
@@ -425,6 +510,7 @@ void Init_binary_protocol_accelerated() {
   rb_define_method(bpa_class, "write_double",        rb_thrift_binary_proto_write_double, 1);
   rb_define_method(bpa_class, "write_string",        rb_thrift_binary_proto_write_string, 1);
   rb_define_method(bpa_class, "write_binary",        rb_thrift_binary_proto_write_binary, 1);
+  rb_define_method(bpa_class, "write_uuid",          rb_thrift_binary_proto_write_uuid, 1);
   // unused methods
   rb_define_method(bpa_class, "write_message_end", rb_thrift_binary_proto_write_message_end, 0);
   rb_define_method(bpa_class, "write_struct_begin", rb_thrift_binary_proto_write_struct_begin, 1);
@@ -447,6 +533,7 @@ void Init_binary_protocol_accelerated() {
   rb_define_method(bpa_class, "read_double",         rb_thrift_binary_proto_read_double, 0);
   rb_define_method(bpa_class, "read_string",         rb_thrift_binary_proto_read_string, 0);
   rb_define_method(bpa_class, "read_binary",         rb_thrift_binary_proto_read_binary, 0);
+  rb_define_method(bpa_class, "read_uuid",           rb_thrift_binary_proto_read_uuid, 0);
   // unused methods
   rb_define_method(bpa_class, "read_message_end", rb_thrift_binary_proto_read_message_end, 0);
   rb_define_method(bpa_class, "read_struct_begin", rb_thrift_binary_proto_read_struct_begin, 0);

lib/rb/ext/compact_protocol.c

@@ -58,6 +58,7 @@ static int CTYPE_LIST           = 0x09;
 static int CTYPE_SET            = 0x0A;
 static int CTYPE_MAP            = 0x0B;
 static int CTYPE_STRUCT         = 0x0C;
+static int CTYPE_UUID           = 0x0D;
 
 VALUE rb_thrift_compact_proto_write_i16(VALUE self, VALUE i16);
 
@@ -86,6 +87,8 @@ static int get_compact_type(VALUE type_value) {
     return CTYPE_MAP;
   } else if (type == TTYPE_STRUCT) {
     return CTYPE_STRUCT;
+  } else if (type == TTYPE_UUID) {
+    return CTYPE_UUID;
   } else {
     char str[50];
     sprintf(str, "don't know what type: %d", type);
@@ -357,6 +360,8 @@ static int8_t get_ttype(int8_t ctype) {
     return TTYPE_MAP;
   } else if (ctype == CTYPE_STRUCT) {
     return TTYPE_STRUCT;
+  } else if (ctype == CTYPE_UUID) {
+    return TTYPE_UUID;
   } else {
     char str[50];
     sprintf(str, "don't know what type: %d", ctype);

lib/rb/ext/constants.h

@@ -29,6 +29,7 @@ extern int TTYPE_MAP;
 extern int TTYPE_SET;
 extern int TTYPE_LIST;
 extern int TTYPE_STRUCT;
+extern int TTYPE_UUID;
 
 extern ID validate_method_id;
 extern ID write_struct_begin_method_id;
@@ -49,6 +50,7 @@ extern ID write_list_begin_method_id;
 extern ID write_list_end_method_id;
 extern ID write_set_begin_method_id;
 extern ID write_set_end_method_id;
+extern ID write_uuid_method_id;
 extern ID read_bool_method_id;
 extern ID read_byte_method_id;
 extern ID read_i16_method_id;
@@ -63,6 +65,7 @@ extern ID read_list_begin_method_id;
 extern ID read_list_end_method_id;
 extern ID read_set_begin_method_id;
 extern ID read_set_end_method_id;
+extern ID read_uuid_method_id;
 extern ID read_struct_begin_method_id;
 extern ID read_struct_end_method_id;
 extern ID read_field_begin_method_id;

lib/rb/ext/struct.c

@@ -75,6 +75,11 @@ VALUE default_write_string(VALUE protocol, VALUE value) {
   return Qnil;
 }
 
+VALUE default_write_uuid(VALUE protocol, VALUE value) {
+  rb_funcall(protocol, write_uuid_method_id, 1, value);
+  return Qnil;
+}
+
 VALUE default_write_binary(VALUE protocol, VALUE value) {
   rb_funcall(protocol, write_binary_method_id, 1, value);
   return Qnil;
@@ -195,6 +200,10 @@ VALUE default_read_string(VALUE protocol) {
   return rb_funcall(protocol, read_string_method_id, 0);
 }
 
+VALUE default_read_uuid(VALUE protocol) {
+  return rb_funcall(protocol, read_uuid_method_id, 0);
+}
+
 VALUE default_read_binary(VALUE protocol) {
   return rb_funcall(protocol, read_binary_method_id, 0);
 }
@@ -342,6 +351,8 @@ static void write_anything(int ttype, VALUE value, VALUE protocol, VALUE field_i
     } else {
       default_write_binary(protocol, value);
     }
+  } else if (ttype == TTYPE_UUID) {
+    default_write_uuid(protocol, value);
   } else if (IS_CONTAINER(ttype)) {
     write_container(ttype, field_info, value, protocol);
   } else if (ttype == TTYPE_STRUCT) {
@@ -452,6 +463,8 @@ static VALUE read_anything(VALUE protocol, int ttype, VALUE field_info) {
     }
   } else if (ttype == TTYPE_DOUBLE) {
     result = default_read_double(protocol);
+  } else if (ttype == TTYPE_UUID) {
+    result = default_read_uuid(protocol);
   } else if (ttype == TTYPE_STRUCT) {
     VALUE klass = rb_hash_aref(field_info, class_sym);
     result = rb_class_new_instance(0, NULL, klass);

lib/rb/ext/thrift_native.c

@@ -43,6 +43,7 @@ int TTYPE_MAP;
 int TTYPE_SET;
 int TTYPE_LIST;
 int TTYPE_STRUCT;
+int TTYPE_UUID;
 
 // method ids
 ID validate_method_id;
@@ -57,6 +58,7 @@ ID write_i32_method_id;
 ID write_i64_method_id;
 ID write_double_method_id;
 ID write_string_method_id;
+ID write_uuid_method_id;
 ID write_binary_method_id;
 ID write_map_begin_method_id;
 ID write_map_end_method_id;
@@ -70,6 +72,7 @@ ID read_i16_method_id;
 ID read_i32_method_id;
 ID read_i64_method_id;
 ID read_string_method_id;
+ID read_uuid_method_id;
 ID read_binary_method_id;
 ID read_double_method_id;
 ID read_map_begin_method_id;
@@ -138,6 +141,7 @@ void Init_thrift_native() {
   TTYPE_SET = FIX2INT(rb_const_get(thrift_types_module, rb_intern("SET")));
   TTYPE_LIST = FIX2INT(rb_const_get(thrift_types_module, rb_intern("LIST")));
   TTYPE_STRUCT = FIX2INT(rb_const_get(thrift_types_module, rb_intern("STRUCT")));
+  TTYPE_UUID = FIX2INT(rb_const_get(thrift_types_module, rb_intern("UUID")));
 
   // method ids
   validate_method_id = rb_intern("validate");
@@ -152,6 +156,7 @@ void Init_thrift_native() {
   write_i64_method_id = rb_intern("write_i64");
   write_double_method_id = rb_intern("write_double");
   write_string_method_id = rb_intern("write_string");
+  write_uuid_method_id = rb_intern("write_uuid");
   write_binary_method_id = rb_intern("write_binary");
   write_map_begin_method_id = rb_intern("write_map_begin");
   write_map_end_method_id = rb_intern("write_map_end");
@@ -165,10 +170,11 @@ void Init_thrift_native() {
   read_i32_method_id = rb_intern("read_i32");
   read_i64_method_id = rb_intern("read_i64");
   read_string_method_id = rb_intern("read_string");
+  read_uuid_method_id = rb_intern("read_uuid");
   read_binary_method_id = rb_intern("read_binary");
   read_double_method_id = rb_intern("read_double");
   read_map_begin_method_id = rb_intern("read_map_begin");
-  read_map_end_method_id = rb_intern("read_map_end");  
+  read_map_end_method_id = rb_intern("read_map_end");
   read_list_begin_method_id = rb_intern("read_list_begin");
   read_list_end_method_id = rb_intern("read_list_end");
   read_set_begin_method_id = rb_intern("read_set_begin");
@@ -192,7 +198,7 @@ void Init_thrift_native() {
   fields_const_id = rb_intern("FIELDS");
   transport_ivar_id = rb_intern("@trans");
   strict_read_ivar_id = rb_intern("@strict_read");
-  strict_write_ivar_id = rb_intern("@strict_write");  
+  strict_write_ivar_id = rb_intern("@strict_write");
 
   // cached symbols
   type_sym = ID2SYM(rb_intern("type"));

lib/rb/lib/thrift.rb

@@ -23,7 +23,6 @@
 $:.unshift File.dirname(__FILE__)
 
 require 'thrift/bytes'
-require 'thrift/core_ext'
 require 'thrift/exceptions'
 require 'thrift/types'
 require 'thrift/processor'