Rework SpyInput buffer handling

SpyInput uses a fixed-size buffer and explicitly zeros memory which is good for
catching logic errors but defeats valgrind's memory tracking. Use a separate
buffer of exactly the correct size for each request. This correctly catches the
problem under valgrind:

```
==8694== Invalid read of size 2
==8694==    at 0x54EFFB: utf16_iterate (utf16.c:10)
==8694==    by 0x551126: ts_lexer__get_lookahead (lexer.c:54)
==8694==    by 0x5515CD: ts_lexer_start (lexer.c:154)
==8694==    by 0x54699F: parser(long,...)(long long) (parser.c:297)
==8694==    by 0x54788A: parser__get_lookahead (parser.c:439)
==8694==    by 0x54B2D3: parser__advance (parser.c:1150)
==8694==    by 0x54C2AA: parser_parse (parser.c:1348)
==8694==    by 0x53F063: ts_document_parse_with_options (document.c:136)
==8694==    by 0x53EF43: ts_document_parse (document.c:107)
==8694==    by 0x4AED11: {lambda()#1}::operator()() const::{lambda()#1}::operator()() const::{lambda()#4}::operator()() const::{lambda()#4}::operator()() const (document_test.cc:82)
==8694==    by 0x4B56B6: std::_Function_handler<void (), {lambda()#1}::operator()() const::{lambda()#1}::operator()() const::{lambda()#4}::operator()() const::{lambda()#4}>::_M_invoke(std::_Any_data const&) (functional:1871)
==8694==    by 0x40F8C5: std::function<void ()>::operator()() const (functional:2267)
==8694==  Address 0x5d08be0 is 0 bytes inside a block of size 1 alloc'd
==8694==    at 0x4C2E80F: operator new[](unsigned long) (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==8694==    by 0x507C3E: SpyInput::read(void*, unsigned int*) (spy_input.cc:66)
==8694==    by 0x55103D: ts_lexer__get_chunk (lexer.c:29)
==8694==    by 0x5515B6: ts_lexer_start (lexer.c:152)
==8694==    by 0x54699F: parser(long,...)(long long) (parser.c:297)
==8694==    by 0x54788A: parser__get_lookahead (parser.c:439)
==8694==    by 0x54B2D3: parser__advance (parser.c:1150)
==8694==    by 0x54C2AA: parser_parse (parser.c:1348)
==8694==    by 0x53F063: ts_document_parse_with_options (document.c:136)
==8694==    by 0x53EF43: ts_document_parse (document.c:107)
==8694==    by 0x4AED11: {lambda()#1}::operator()() const::{lambda()#1}::operator()() const::{lambda()#4}::operator()() const::{lambda()#4}::operator()() const (document_test.cc:82)
==8694==    by 0x4B56B6: std::_Function_handler<void (), {lambda()#1}::operator()() const::{lambda()#1}::operator()() const::{lambda()#4}::operator()() const::{lambda()#4}>::_M_invoke(std::_Any_data const&) (functional:1871)
```

test/helpers/spy_input.cc

@@ -12,8 +12,7 @@ static const size_t UTF8_MAX_CHAR_SIZE = 4;
 
 SpyInput::SpyInput(string content, size_t chars_per_chunk) :
   chars_per_chunk(chars_per_chunk),
-  buffer_size(UTF8_MAX_CHAR_SIZE * chars_per_chunk),
-  buffer(new char[buffer_size]),
+  buffer(nullptr),
   byte_offset(0),
   content(content),
   encoding(TSInputEncodingUTF8),
@@ -57,12 +56,19 @@ const char * SpyInput::read(void *payload, uint32_t *bytes_read) {
    * This class stores its entire `content` in a contiguous buffer, but we want
    * to ensure that the code under test cannot accidentally read more than
    * `*bytes_read` bytes past the returned pointer. To make sure that this type
-   * of error does not fly, we copy the chunk into a zeroed-out buffer and
+   * of error does not fly, we allocate a separate buffer for each request and
    * return a reference to that buffer, rather than a pointer into the main
-   * content.
+   * content. The temporary buffer only fits `*bytes_read` bytes so valgrind
+   * can detect code reading too many bytes from the buffer.
    */
-  memset(spy->buffer, 0, spy->buffer_size);
-  memcpy(spy->buffer, result.data(), byte_count);
+  delete[] spy->buffer;
+  if (byte_count) {
+    spy->buffer = new char[byte_count];
+    memcpy(spy->buffer, result.data(), byte_count);
+  } else {
+    spy->buffer = nullptr;
+  }
+
   return spy->buffer;
 }
 

test/helpers/spy_input.h

@@ -13,7 +13,6 @@ struct SpyInputEdit {
 
 class SpyInput {
   uint32_t chars_per_chunk;
-  uint32_t buffer_size;
   char *buffer;
   uint32_t byte_offset;
   std::vector<SpyInputEdit> undo_stack;

test/runtime/document_test.cc

@@ -72,6 +72,16 @@ describe("Document", [&]() {
         "(array (true) (false))");
     });
 
+    it("handles truncated UTF16 data", [&]() {
+      const char content[1] = { '\0' };
+      spy_input->content = string(content, sizeof(content));
+      spy_input->encoding = TSInputEncodingUTF16;
+
+      ts_document_set_input(document, spy_input->input());
+      ts_document_invalidate(document);
+      ts_document_parse(document);
+    });
+
     it("allows columns to be measured in either bytes or characters", [&]() {
       const char16_t content[] = u"[true, false]";
       spy_input->content = string((const char *)content, sizeof(content));

test/runtime/lexer_test.cc

@@ -1,18 +0,0 @@
-#include "test_helper.h"
-#include "runtime/utf16.h"
-
-START_TEST
-
-describe("Lexer", [&]() {
-    it("handles truncated UTF16 data", [&]() {
-      uint8_t *content = new uint8_t[1];
-      *content = 'A';
-
-      int32_t code_point = 0;
-      utf16_iterate(content, 1, &code_point);
-
-      delete[] content;
-    });
-});
-
-END_TEST