JSON.dump: write directly into the provided IO

Rakefile

@@ -161,7 +161,7 @@ if defined?(RUBY_ENGINE) and RUBY_ENGINE == 'jruby'
   file JRUBY_PARSER_JAR => :compile do
     cd 'java/src' do
       parser_classes = FileList[
-        "json/ext/ByteListTranscoder*.class",
+        "json/ext/ByteList*.class",
         "json/ext/OptionsReader*.class",
         "json/ext/Parser*.class",
         "json/ext/RuntimeInfo*.class",
@@ -179,7 +179,7 @@ if defined?(RUBY_ENGINE) and RUBY_ENGINE == 'jruby'
   file JRUBY_GENERATOR_JAR => :compile do
     cd 'java/src' do
       generator_classes = FileList[
-        "json/ext/ByteListTranscoder*.class",
+        "json/ext/ByteList*.class",
         "json/ext/OptionsReader*.class",
         "json/ext/Generator*.class",
         "json/ext/RuntimeInfo*.class",

ext/json/ext/fbuffer/fbuffer.h

@@ -46,9 +46,11 @@ typedef struct FBufferStruct {
     unsigned long len;
     unsigned long capa;
     char *ptr;
+    VALUE io;
 } FBuffer;
 
 #define FBUFFER_STACK_SIZE 512
+#define FBUFFER_IO_BUFFER_SIZE (16384 - 1)
 #define FBUFFER_INITIAL_LENGTH_DEFAULT 1024
 
 #define FBUFFER_PTR(fb) ((fb)->ptr)
@@ -66,7 +68,7 @@ static void fbuffer_append_long(FBuffer *fb, long number);
 #endif
 static inline void fbuffer_append_char(FBuffer *fb, char newchr);
 #ifdef JSON_GENERATOR
-static VALUE fbuffer_to_s(FBuffer *fb);
+static VALUE fbuffer_finalize(FBuffer *fb);
 #endif
 
 static void fbuffer_stack_init(FBuffer *fb, unsigned long initial_length, char *stack_buffer, long stack_buffer_size)
@@ -86,24 +88,19 @@ static void fbuffer_free(FBuffer *fb)
     }
 }
 
-#ifndef JSON_GENERATOR
 static void fbuffer_clear(FBuffer *fb)
 {
     fb->len = 0;
 }
-#endif
 
-static void fbuffer_do_inc_capa(FBuffer *fb, unsigned long requested)
+static void fbuffer_flush(FBuffer *fb)
 {
-    unsigned long required;
-
-    if (RB_UNLIKELY(!fb->ptr)) {
-        fb->ptr = ALLOC_N(char, fb->initial_length);
-        fb->capa = fb->initial_length;
-    }
-
-    for (required = fb->capa; requested > required - fb->len; required <<= 1);
+    rb_io_write(fb->io, rb_utf8_str_new(fb->ptr, fb->len));
+    fbuffer_clear(fb);
+}
 
+static void fbuffer_realloc(FBuffer *fb, unsigned long required)
+{
     if (required > fb->capa) {
         if (fb->type == FBUFFER_STACK_ALLOCATED) {
             const char *old_buffer = fb->ptr;
@@ -117,6 +114,32 @@ static void fbuffer_do_inc_capa(FBuffer *fb, unsigned long requested)
     }
 }
 
+static void fbuffer_do_inc_capa(FBuffer *fb, unsigned long requested)
+{
+    if (RB_UNLIKELY(fb->io)) {
+        if (fb->capa < FBUFFER_IO_BUFFER_SIZE) {
+            fbuffer_realloc(fb, FBUFFER_IO_BUFFER_SIZE);
+        } else {
+            fbuffer_flush(fb);
+        }
+
+        if (RB_LIKELY(requested < fb->capa)) {
+            return;
+        }
+    }
+
+    unsigned long required;
+
+    if (RB_UNLIKELY(!fb->ptr)) {
+        fb->ptr = ALLOC_N(char, fb->initial_length);
+        fb->capa = fb->initial_length;
+    }
+
+    for (required = fb->capa; requested > required - fb->len; required <<= 1);
+
+    fbuffer_realloc(fb, required);
+}
+
 static inline void fbuffer_inc_capa(FBuffer *fb, unsigned long requested)
 {
     if (RB_UNLIKELY(requested > fb->capa - fb->len)) {
@@ -174,11 +197,18 @@ static void fbuffer_append_long(FBuffer *fb, long number)
     fbuffer_append(fb, buffer_end - len, len);
 }
 
-static VALUE fbuffer_to_s(FBuffer *fb)
+static VALUE fbuffer_finalize(FBuffer *fb)
 {
-    VALUE result = rb_utf8_str_new(FBUFFER_PTR(fb), FBUFFER_LEN(fb));
-    fbuffer_free(fb);
-    return result;
+    if (fb->io) {
+        fbuffer_flush(fb);
+        fbuffer_free(fb);
+        rb_io_flush(fb->io);
+        return fb->io;
+    } else {
+        VALUE result = rb_utf8_str_new(FBUFFER_PTR(fb), FBUFFER_LEN(fb));
+        fbuffer_free(fb);
+        return result;
+    }
 }
 #endif
 #endif

ext/json/ext/generator/generator.c

@@ -54,7 +54,7 @@ struct generate_json_data {
 };
 
 static VALUE cState_from_state_s(VALUE self, VALUE opts);
-static VALUE cState_partial_generate(VALUE self, VALUE obj, generator_func);
+static VALUE cState_partial_generate(VALUE self, VALUE obj, generator_func, VALUE io);
 static void generate_json(FBuffer *buffer, struct generate_json_data *data, JSON_Generator_State *state, VALUE obj);
 static void generate_json_object(FBuffer *buffer, struct generate_json_data *data, JSON_Generator_State *state, VALUE obj);
 static void generate_json_array(FBuffer *buffer, struct generate_json_data *data, JSON_Generator_State *state, VALUE obj);
@@ -453,7 +453,7 @@ static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_object);
+    return cState_partial_generate(Vstate, self, generate_json_object, Qfalse);
 }
 
 /*
@@ -467,7 +467,7 @@ static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self)
 static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self) {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_array);
+    return cState_partial_generate(Vstate, self, generate_json_array, Qfalse);
 }
 
 #ifdef RUBY_INTEGER_UNIFICATION
@@ -480,7 +480,7 @@ static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_integer);
+    return cState_partial_generate(Vstate, self, generate_json_integer, Qfalse);
 }
 
 #else
@@ -493,7 +493,7 @@ static VALUE mFixnum_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_fixnum);
+    return cState_partial_generate(Vstate, self, generate_json_fixnum, Qfalse);
 }
 
 /*
@@ -505,7 +505,7 @@ static VALUE mBignum_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_bignum);
+    return cState_partial_generate(Vstate, self, generate_json_bignum, Qfalse);
 }
 #endif
 
@@ -518,7 +518,7 @@ static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_float);
+    return cState_partial_generate(Vstate, self, generate_json_float, Qfalse);
 }
 
 /*
@@ -543,7 +543,7 @@ static VALUE mString_to_json(int argc, VALUE *argv, VALUE self)
 {
     rb_check_arity(argc, 0, 1);
     VALUE Vstate = cState_from_state_s(cState, argc == 1 ? argv[0] : Qnil);
-    return cState_partial_generate(Vstate, self, generate_json_string);
+    return cState_partial_generate(Vstate, self, generate_json_string, Qfalse);
 }
 
 /*
@@ -638,7 +638,7 @@ static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self)
     rb_scan_args(argc, argv, "01", &state);
     Check_Type(string, T_STRING);
     state = cState_from_state_s(cState, state);
-    return cState_partial_generate(state, string, generate_json_string);
+    return cState_partial_generate(state, string, generate_json_string, Qfalse);
 }
 
 static void State_mark(void *ptr)
@@ -1045,12 +1045,14 @@ static VALUE generate_json_rescue(VALUE d, VALUE exc)
     return Qundef;
 }
 
-static VALUE cState_partial_generate(VALUE self, VALUE obj, generator_func func)
+static VALUE cState_partial_generate(VALUE self, VALUE obj, generator_func func, VALUE io)
 {
     GET_STATE(self);
 
     char stack_buffer[FBUFFER_STACK_SIZE];
-    FBuffer buffer = {0};
+    FBuffer buffer = {
+        .io = RTEST(io) ? io : Qfalse,
+    };
     fbuffer_stack_init(&buffer, state->buffer_initial_length, stack_buffer, FBUFFER_STACK_SIZE);
 
     struct generate_json_data data = {
@@ -1062,19 +1064,12 @@ static VALUE cState_partial_generate(VALUE self, VALUE obj, generator_func func)
     };
     rb_rescue(generate_json_try, (VALUE)&data, generate_json_rescue, (VALUE)&data);
 
-    return fbuffer_to_s(&buffer);
+    return fbuffer_finalize(&buffer);
 }
 
-/*
- * call-seq: generate(obj)
- *
- * Generates a valid JSON document from object +obj+ and returns the
- * result. If no valid JSON document can be created this method raises a
- * GeneratorError exception.
- */
-static VALUE cState_generate(VALUE self, VALUE obj)
+static VALUE cState_generate(VALUE self, VALUE obj, VALUE io)
 {
-    VALUE result = cState_partial_generate(self, obj, generate_json);
+    VALUE result = cState_partial_generate(self, obj, generate_json, io);
     GET_STATE(self);
     (void)state;
     return result;
@@ -1502,14 +1497,16 @@ static VALUE cState_configure(VALUE self, VALUE opts)
     return self;
 }
 
-static VALUE cState_m_generate(VALUE klass, VALUE obj, VALUE opts)
+static VALUE cState_m_generate(VALUE klass, VALUE obj, VALUE opts, VALUE io)
 {
     JSON_Generator_State state = {0};
     state_init(&state);
     configure_state(&state, opts);
 
     char stack_buffer[FBUFFER_STACK_SIZE];
-    FBuffer buffer = {0};
+    FBuffer buffer = {
+        .io = RTEST(io) ? io : Qfalse,
+    };
     fbuffer_stack_init(&buffer, state.buffer_initial_length, stack_buffer, FBUFFER_STACK_SIZE);
 
     struct generate_json_data data = {
@@ -1521,7 +1518,7 @@ static VALUE cState_m_generate(VALUE klass, VALUE obj, VALUE opts)
     };
     rb_rescue(generate_json_try, (VALUE)&data, generate_json_rescue, (VALUE)&data);
 
-    return fbuffer_to_s(&buffer);
+    return fbuffer_finalize(&buffer);
 }
 
 /*
@@ -1583,9 +1580,9 @@ void Init_generator(void)
     rb_define_method(cState, "depth=", cState_depth_set, 1);
     rb_define_method(cState, "buffer_initial_length", cState_buffer_initial_length, 0);
     rb_define_method(cState, "buffer_initial_length=", cState_buffer_initial_length_set, 1);
-    rb_define_method(cState, "generate", cState_generate, 1);
+    rb_define_private_method(cState, "_generate", cState_generate, 2);
 
-    rb_define_singleton_method(cState, "generate", cState_m_generate, 2);
+    rb_define_singleton_method(cState, "generate", cState_m_generate, 3);
 
     VALUE mGeneratorMethods = rb_define_module_under(mGenerator, "GeneratorMethods");
 

java/src/json/ext/ByteListDirectOutputStream.java

@@ -0,0 +1,16 @@
+package json.ext;
+
+import org.jcodings.Encoding;
+import org.jruby.util.ByteList;
+
+import java.io.ByteArrayOutputStream;
+
+public class ByteListDirectOutputStream extends ByteArrayOutputStream {
+    ByteListDirectOutputStream(int size) {
+        super(size);
+    }
+
+    public ByteList toByteListDirect(Encoding encoding) {
+        return new ByteList(buf, 0, count, encoding, false);
+    }
+}

java/src/json/ext/ByteListTranscoder.java

@@ -9,6 +9,9 @@
 import org.jruby.runtime.ThreadContext;
 import org.jruby.util.ByteList;
 
+import java.io.IOException;
+import java.io.OutputStream;
+
 /**
  * A class specialized in transcoding a certain String format into another,
  * using UTF-8 ByteLists as both input and output.
@@ -23,7 +26,7 @@ abstract class ByteListTranscoder {
     /** Position of the next character to read */
     protected int pos;
 
-    private ByteList out;
+    private OutputStream out;
     /**
      * When a character that can be copied straight into the output is found,
      * its index is stored on this variable, and copying is delayed until
@@ -37,11 +40,11 @@ protected ByteListTranscoder(ThreadContext context) {
         this.context = context;
     }
 
-    protected void init(ByteList src, ByteList out) {
+    protected void init(ByteList src, OutputStream out) {
         this.init(src, 0, src.length(), out);
     }
 
-    protected void init(ByteList src, int start, int end, ByteList out) {
+    protected void init(ByteList src, int start, int end, OutputStream out) {
         this.src = src;
         this.pos = start;
         this.charStart = start;
@@ -142,19 +145,19 @@ protected void quoteStart() {
      *               recently read character, or {@link #charStart} to quote
      *               until the character before it.
      */
-    protected void quoteStop(int endPos) {
+    protected void quoteStop(int endPos) throws IOException {
         if (quoteStart != -1) {
-            out.append(src, quoteStart, endPos - quoteStart);
+            out.write(src.bytes(), quoteStart, endPos - quoteStart);
             quoteStart = -1;
         }
     }
 
-    protected void append(int b) {
-        out.append(b);
+    protected void append(int b) throws IOException {
+        out.write(b);
     }
 
-    protected void append(byte[] origin, int start, int length) {
-        out.append(origin, start, length);
+    protected void append(byte[] origin, int start, int length) throws IOException {
+        out.write(origin, start, length);
     }