Merge remote-tracking branch 'origin/topic/timw/odesc-size-t'

* origin/topic/timw/odesc-size-t:
  Use size_t for byte sizes in SerializationFormat
  Add ODesc::Size() that returns size_t, deprecate ODesc::Len()

CHANGES

@@ -1,3 +1,9 @@
+8.0.0-dev.751 | 2025-07-25 09:13:51 -0700
+
+  * Use size_t for byte sizes in SerializationFormat (Tim Wojtulewicz, Corelight)
+
+  * Add ODesc::Size() that returns size_t, deprecate ODesc::Len() (Tim Wojtulewicz, Corelight)
+
 8.0.0-dev.748 | 2025-07-25 10:58:16 +0200
 
   * RecordType: Allow field init deferral of deferrable record constructors (Arne Welzel, Corelight)

VERSION

@@ -1 +1 @@
-8.0.0-dev.748
+8.0.0-dev.751

src/Desc.cc

@@ -11,8 +11,8 @@
 #include "zeek/IPAddr.h"
 #include "zeek/Reporter.h"
 
-constexpr unsigned int DEFAULT_SIZE = 128;
+constexpr size_t DEFAULT_SIZE = 128;
-constexpr int SLOP = 10;
+constexpr size_t SLOP = 10;
 
 namespace zeek {
 
@@ -74,7 +74,7 @@ void ODesc::PopIndentNoNL() {
 }
 
 void ODesc::Add(const char* s, int do_indent) {
-    unsigned int n = strlen(s);
+    size_t n = strlen(s);
 
     if ( do_indent && IsReadable() && offset > 0 && ((const char*)base)[offset - 1] == '\n' )
         Indent();
@@ -235,7 +235,7 @@ std::pair<const char*, size_t> ODesc::FirstEscapeLoc(const char* bytes, size_t n
     return {nullptr, 0};
 }
 
-void ODesc::AddBytes(const void* bytes, unsigned int n) {
+void ODesc::AddBytes(const void* bytes, size_t n) {
     if ( ! escape ) {
         AddBytesRaw(bytes, n);
         return;
@@ -271,7 +271,7 @@ void ODesc::AddBytes(const void* bytes, unsigned int n) {
     }
 }
 
-void ODesc::AddBytesRaw(const void* bytes, unsigned int n) {
+void ODesc::AddBytesRaw(const void* bytes, size_t n) {
     if ( n == 0 )
         return;
 
@@ -304,7 +304,7 @@ void ODesc::AddBytesRaw(const void* bytes, unsigned int n) {
     }
 }
 
-void ODesc::Grow(unsigned int n) {
+void ODesc::Grow(size_t n) {
     bool size_changed = false;
     while ( offset + n + SLOP >= size ) {
         size *= 2;
@@ -319,7 +319,8 @@ void ODesc::Clear() {
     offset = 0;
 
     // If we've allocated an exceedingly large amount of space, free it.
-    if ( size > 10 * 1024 * 1024 ) {
+    constexpr size_t too_large = 10l * 1024 * 1024;
+    if ( size > too_large ) {
         free(base);
         size = DEFAULT_SIZE;
         base = util::safe_malloc(size);

src/Desc.h

@@ -144,7 +144,11 @@ public:
         return byte_vec(t);
     }
 
-    int Len() const { return offset; }
+    [[deprecated("Remove in v8.1. Use Size() that returns size_t instead.")]]
+    int Len() const {
+        return static_cast<int>(offset);
+    }
+    size_t Size() const { return offset; }
 
     void Clear();
 
@@ -157,11 +161,11 @@ public:
 protected:
     void Indent();
 
-    void AddBytes(const void* bytes, unsigned int n);
+    void AddBytes(const void* bytes, size_t n);
-    void AddBytesRaw(const void* bytes, unsigned int n);
+    void AddBytesRaw(const void* bytes, size_t n);
 
     // Make buffer big enough for n bytes beyond bufp.
-    void Grow(unsigned int n);
+    void Grow(size_t n);
 
     /**
      * Returns the location of the first place in the bytes to be hex-escaped.
@@ -188,9 +192,9 @@ protected:
     DescType type;
     DescStyle style;
 
-    void* base;          // beginning of buffer
+    void* base;    // beginning of buffer
-    unsigned int offset; // where we are in the buffer
+    size_t offset; // where we are in the buffer
-    unsigned int size;   // size of buffer in bytes
+    size_t size;   // size of buffer in bytes
 
     bool utf8;   // whether valid utf-8 sequences may pass through unescaped
     bool escape; // escape unprintable characters in output?

src/Expr.cc

@@ -4369,7 +4369,7 @@ void LambdaExpr::BuildName() {
 
     for ( ;; ) {
         hash128_t h;
-        KeyedHash::Hash128(d.Bytes(), d.Len(), &h);
+        KeyedHash::Hash128(d.Bytes(), d.Size(), &h);
 
         my_name = "lambda_<" + std::to_string(h[0]) + ">";
         auto fullname = make_full_var_name(current_module.data(), my_name.data());

src/OpaqueVal.cc

@@ -401,7 +401,7 @@ void HashVal::digest_one(detail::HashDigestState* h, const Val* v) {
     else {
         ODesc d(DESC_BINARY);
         v->Describe(&d);
-        detail::hash_update(h, (const u_char*)d.Bytes(), d.Len());
+        detail::hash_update(h, (const u_char*)d.Bytes(), d.Size());
     }
 }
 

src/SerializationFormat.cc

@@ -40,8 +40,8 @@ void SerializationFormat::StartWrite() {
     bytes_written = 0;
 }
 
-uint32_t SerializationFormat::EndWrite(char** data) {
+size_t SerializationFormat::EndWrite(char** data) {
-    uint32_t rval = output_pos;
+    size_t rval = output_pos;
     *data = output;
     output = nullptr;
     output_size = 0;

src/SerializationFormat.h

@@ -57,7 +57,7 @@ public:
      *             be reclaimed using "free()".
      * @return The number of bytes in the buffer object assigned to \a data.
      */
-    virtual uint32_t EndWrite(char** data);
+    virtual size_t EndWrite(char** data);
 
     virtual bool Write(int v, const char* tag) = 0;
     virtual bool Write(uint16_t v, const char* tag) = 0;
@@ -80,7 +80,7 @@ public:
     virtual bool WriteSeparator() = 0;
 
     // Returns number of raw bytes written since last call to StartWrite().
-    int BytesWritten() const { return bytes_written; }
+    size_t BytesWritten() const { return bytes_written; }
 
 protected:
     bool ReadData(void* buf, size_t count);
@@ -89,15 +89,15 @@ protected:
     static const uint32_t INITIAL_SIZE = 65536;
     static const float GROWTH_FACTOR;
     char* output = nullptr;
-    uint32_t output_size = 0;
+    size_t output_size = 0;
-    uint32_t output_pos = 0;
+    size_t output_pos = 0;
 
     const char* input = nullptr;
-    uint32_t input_len = 0;
+    size_t input_len = 0;
-    uint32_t input_pos = 0;
+    size_t input_pos = 0;
 
-    int bytes_written = 0;
+    size_t bytes_written = 0;
-    int bytes_read = 0;
+    size_t bytes_read = 0;
 };
 
 class BinarySerializationFormat final : public SerializationFormat {

src/Stmt.cc

@@ -345,7 +345,7 @@ void do_print_stmt(const std::vector<ValPtr>& vals) {
         d.SetStyle(style);
 
         describe_vals(vals, &d, offset);
-        f->Write(d.Description(), d.Len());
+        f->Write(d.Description(), d.Size());
     }
     else {
         ODesc d(DESC_READABLE, f);

src/Val.cc

@@ -397,7 +397,7 @@ static void BuildJSON(json::detail::NullDoubleWriter& writer, Val* val, bool onl
             ODesc d;
             d.SetStyle(RAW_STYLE);
             val->Describe(&d);
-            writer.String(reinterpret_cast<const char*>(d.Bytes()), d.Len());
+            writer.String(reinterpret_cast<const char*>(d.Bytes()), d.Size());
             break;
         }
 
@@ -408,7 +408,7 @@ static void BuildJSON(json::detail::NullDoubleWriter& writer, Val* val, bool onl
                 ODesc d;
                 d.SetStyle(RAW_STYLE);
                 val->Describe(&d);
-                writer.String(reinterpret_cast<const char*>(d.Bytes()), d.Len());
+                writer.String(reinterpret_cast<const char*>(d.Bytes()), d.Size());
             }
             break;
         }
@@ -420,7 +420,7 @@ static void BuildJSON(json::detail::NullDoubleWriter& writer, Val* val, bool onl
             ODesc d;
             d.SetStyle(RAW_STYLE);
             val->Describe(&d);
-            std::string desc(reinterpret_cast<const char*>(d.Bytes()), d.Len());
+            std::string desc(reinterpret_cast<const char*>(d.Bytes()), d.Size());
 
             // None of our function types should have surrounding
             // whitespace, but ODesc might produce it due to its

src/logging/writers/ascii/Ascii.cc

@@ -240,10 +240,7 @@ void Ascii::InitConfigOptions() {
 
     meta_prefix.assign((const char*)BifConst::LogAscii::meta_prefix->Bytes(), BifConst::LogAscii::meta_prefix->Len());
 
-    ODesc tsfmt;
+    json_timestamps = zeek::obj_desc_short(BifConst::LogAscii::json_timestamps);
-    BifConst::LogAscii::json_timestamps->Describe(&tsfmt);
-    json_timestamps.assign((const char*)tsfmt.Bytes(), tsfmt.Len());
-
     json_include_unset_fields = BifConst::LogAscii::json_include_unset_fields;
 
     gzip_file_extension.assign((const char*)BifConst::LogAscii::gzip_file_extension->Bytes(),
@@ -591,7 +588,7 @@ bool Ascii::DoWrite(int num_fields, const threading::Field* const* fields, threa
     desc.AddRaw("\n", 1);
 
     const char* bytes = (const char*)desc.Bytes();
-    int len = desc.Len();
+    size_t len = desc.Size();
 
     if ( strncmp(bytes, meta_prefix.data(), meta_prefix.size()) == 0 ) {
         // It would so escape the first character.

src/logging/writers/sqlite/SQLite.cc

@@ -318,7 +318,7 @@ int SQLite::AddParams(Value* val, int pos) {
                 }
 
             desc.RemoveEscapeSequence(set_separator);
-            return sqlite3_bind_text(st, pos, (const char*)desc.Bytes(), desc.Len(), SQLITE_TRANSIENT);
+            return sqlite3_bind_text(st, pos, (const char*)desc.Bytes(), desc.Size(), SQLITE_TRANSIENT);
         }
 
         case TYPE_VECTOR: {
@@ -338,7 +338,7 @@ int SQLite::AddParams(Value* val, int pos) {
                 }
 
             desc.RemoveEscapeSequence(set_separator);
-            return sqlite3_bind_text(st, pos, (const char*)desc.Bytes(), desc.Len(), SQLITE_TRANSIENT);
+            return sqlite3_bind_text(st, pos, (const char*)desc.Bytes(), desc.Size(), SQLITE_TRANSIENT);
         }
 
         default: Error(Fmt("unsupported field format %d", val->type)); return 0;

src/script_opt/ZAM/BuiltInSupport.h

@@ -88,7 +88,7 @@ public:
     DescCatArg(TypePtr _t) : CatArg(), t(std::move(_t)) { d.SetStyle(RAW_STYLE); }
 
     void RenderInto(const ZVal& zv, char*& res) override {
-        auto n = d.Len();
+        auto n = d.Size();
         memcpy(res, d.Bytes(), n);
         res += n;
         d.Clear();
@@ -97,7 +97,7 @@ public:
 protected:
     size_t ComputeMaxSize(const ZVal& zv) override {
         zv.ToVal(t)->Describe(&d);
-        return d.Len();
+        return d.Size();
     }
 
     ODesc d;

src/script_opt/ZAM/Support.cc

@@ -197,7 +197,7 @@ StringValPtr ZAM_val_cat(const ValPtr& v) {
 
     v->Describe(&d);
 
-    String* s = new String(true, d.TakeBytes(), d.Len());
+    String* s = new String(true, d.TakeBytes(), d.Size());
     s->SetUseFreeToDelete(true);
 
     return make_intrusive<StringVal>(s);

src/strings.bif

@@ -121,7 +121,7 @@ function join_string_vec%(vec: string_vec, sep: string%): string
 		e->Describe(&d);
 		}
 
-	zeek::String* s = new zeek::String(true, d.TakeBytes(), d.Len());
+	zeek::String* s = new zeek::String(true, d.TakeBytes(), d.Size());
 	s->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(s);
@@ -179,7 +179,7 @@ function join_string_set%(ss: string_set, sep: string%): string
 		++i;
 		}
 
-	zeek::String* str = new zeek::String(true, d.TakeBytes(), d.Len());
+	zeek::String* str = new zeek::String(true, d.TakeBytes(), d.Size());
 	str->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(str);

src/zeek.bif

@@ -266,13 +266,13 @@ static void do_fmt(const char*& fmt, zeek::Val* v, zeek::ODesc* d)
 			d->Add(" ");
 		}
 
-	d->AddN((const char*)(s.Bytes()), s.Len());
+	d->AddN((const char*)(s.Bytes()), s.Size());
 
 	// Right-padding with whitespace, if any.
 	if ( field_width > 0 && left_just )
 		{
-		int sl = s.Len();
+		size_t sl = s.Size();
-		while ( ++sl <= field_width )
+		while ( ++sl <= static_cast<size_t>(field_width) )
 			d->Add(" ");
 		}
 
@@ -1052,23 +1052,23 @@ template<typename IntType>
 IntType fnva(zeek::Val* input, IntType offset, IntType prime) {
 	zeek::ODesc desc(zeek::DESC_BINARY);
 
-	auto length = 0;
+	size_t length = 0;
 	const u_char* bytes = nullptr;
 
 	if (input->GetType()->Tag() == zeek::TYPE_STRING)
 		{
-		length = ((zeek::StringVal*) input) -> Len();
+		length = ((zeek::StringVal*) input)->Len();
-		bytes = ((zeek::StringVal*) input)-> Bytes();
+		bytes = ((zeek::StringVal*) input)->Bytes();
 		}
 	else
 		{
 		input->Describe(&desc);
 		bytes = desc.Bytes();
-		length = desc.Len();
+		length = desc.Size();
 		}
 
 	IntType rval = offset;
-	for ( auto i = 0; i < length; ++i )
+	for ( size_t i = 0; i < length; ++i )
 		{
 		rval ^= bytes[i];
 		rval *= prime;
@@ -1747,7 +1747,7 @@ function cat%(...%): string
 	for ( const auto& a : @ARG@ )
 		a->Describe(&d);
 
-	String* s = new String(true, d.TakeBytes(), d.Len());
+	String* s = new String(true, d.TakeBytes(), d.Size());
 	s->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(s);
@@ -1784,7 +1784,7 @@ function cat_sep%(sep: string, def: string, ...%): string
 		v->Describe(&d);
 		}
 
-	String* s = new String(true, d.TakeBytes(), d.Len());
+	String* s = new String(true, d.TakeBytes(), d.Size());
 	s->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(s);
@@ -1859,7 +1859,7 @@ function fmt%(...%): string
 		return zeek::val_mgr->EmptyString();
 		}
 
-	String* s = new String(true, d.TakeBytes(), d.Len());
+	String* s = new String(true, d.TakeBytes(), d.Size());
 	s->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(s);
@@ -1877,7 +1877,7 @@ function print_raw%(...%): bool
 	zeek::ODesc d(DESC_READABLE);
 	d.SetStyle(RAW_STYLE);
 	describe_vals(@ARG@, &d, 0);
-	printf("%.*s", d.Len(), d.Description());
+	printf("%.*s", static_cast<int>(d.Size()), d.Description());
 	return zeek::val_mgr->Bool(true);
 	%}
 
@@ -2074,7 +2074,7 @@ function type_name%(t: any%): string
 	zeek::ODesc d;
 	t->GetType()->Describe(&d);
 
-	String* s = new String(true, d.TakeBytes(), d.Len());
+	String* s = new String(true, d.TakeBytes(), d.Size());
 	s->SetUseFreeToDelete(true);
 
 	return zeek::make_intrusive<zeek::StringVal>(s);