Compare commits

...
Author SHA1 Message Date
Niels Lohmann ea3dca202b json_view: no save() and load() on big-endian targets
The image format is little-endian only. Instead of throwing
type_error.320 at runtime, save_image(), load_image() and the members
save() and load() do not exist on a big-endian target, so a call fails to
compile, like set() on a read-only document. type_error.320 remains for
saving a discarded document.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-25 12:12:21 +01:00
10 changed files with 53 additions and 36 deletions

No files matched your search

@@ -55,10 +55,7 @@ validate with overload 2 first, if the original vector must survive a failed loa
## Exceptions
On a big-endian target, throws [`type_error.320`](../../home/exceptions.md#jsonexceptiontype_error320) -- the same
exception [`save()`](save.md#exceptions) throws there, since the image format is little-endian only.
Otherwise throws [`parse_error.116`](../../home/exceptions.md#jsonexceptionparse_error116) if `image` is not one
Throws [`parse_error.116`](../../home/exceptions.md#jsonexceptionparse_error116) if `image` is not one
`save()` could have written, or fails the requested `check`:
| message | when |
@@ -95,6 +92,9 @@ are checked once, however many nodes refer to it.
## Notes
**Little-endian targets only.** On a big-endian target, `load()` does not exist, just like
[`save()`](save.md#notes): the image format is little-endian, so a call fails to compile.
### `image_check` { #image_check }
```cpp
@@ -31,9 +31,6 @@ Throws [`type_error.320`](../../home/exceptions.md#jsonexceptiontype_error320) i
[discarded](is_discarded.md) -- a default-constructed document, or one a failed [`parse()`](parse.md)/
[`read()`](read.md) with `allow_exceptions == false` left discarded.
On a big-endian target, throws `type_error.320` with a different message instead: the image format is little-endian
only (see [Notes](#notes)).
Throws [`out_of_range.416`](../../home/exceptions.md#jsonexceptionout_of_range416) if the node count, the text, or
the decoded strings of the image would individually reach 4 GiB -- the same 32-bit offsets
[`parse()`](parse.md#exceptions) and, for edits, [`set`](set.md)/[`push_back`](push_back.md) are already limited to.
@@ -44,9 +41,6 @@ the decoded strings of the image would individually reach 4 GiB -- the same 32-b
[json.exception.type_error.320] cannot save a discarded json_document
```
```
[json.exception.type_error.320] json_document images need a little-endian target
```
```
[json.exception.out_of_range.416] images of 4 GiB or more are not supported by json_document
```
@@ -70,9 +64,9 @@ else -- see [`load`'s Exceptions](load.md#exceptions).
the *same* build on the *same* (little-endian) machine -- not as a long-term storage format. Keep the original
JSON text if you need to read a saved document back with a future library version.
**Little-endian only.** The image is written as raw little-endian bytes, with no byte-swapping. `save()` (and
[`load`](load.md)) throw `type_error.320` on a big-endian target rather than silently produce bytes a big-endian
reader could not interpret correctly.
**Little-endian targets only.** The image is written as raw little-endian bytes, with no byte-swapping. On a
big-endian target, `save()` and [`load`](load.md) do not exist: a call fails to compile rather than produce bytes a
big-endian reader could not interpret correctly.
## Examples
+4 -3
View File
@@ -354,9 +354,10 @@ validation entirely, for an image trusted as much as the process's own memory. S
!!! warning "Experimental"
The image format is versioned but not yet stable, and may change in an incompatible way before it is declared
stable. It is little-endian only, and tied to the library build that wrote it -- use it to cache a document or to
hand one to another process running the *same* build, not as a long-term storage format; keep the original JSON
text if a saved document needs to be readable by a future library version.
stable. It is little-endian only (on big-endian targets, `save()` and `load()` do not exist), and tied to the
library build that wrote it -- use it to cache a document or to hand one to another process running the *same*
build, not as a long-term storage format; keep the original JSON text if a saved document needs to be readable by
a future library version.
The idea of a document you can read without parsing comes from zero-copy formats such as
[FlatBuffers](https://github.com/google/flatbuffers) and [YaFF](https://github.com/yandex/yaff); the check
-5
View File
@@ -853,17 +853,12 @@ from JSON text.
[`basic_json_document::save()`](../api/basic_json_document/save.md) cannot write an
[image](../features/json_view.md#images) of a [discarded](../api/basic_json_document/is_discarded.md) document.
[`save()`](../api/basic_json_document/save.md) and [`load()`](../api/basic_json_document/load.md) also throw this
exception on a big-endian target, since the image format is little-endian only.
!!! failure "Example messages"
```
[json.exception.type_error.320] cannot save a discarded json_document
```
```
[json.exception.type_error.320] json_document images need a little-endian target
```
!!! note
+1 -1
View File
@@ -102,7 +102,7 @@ enum class exception_id : int
type_not_serializable = 317, ///< a value cannot be serialized to the requested binary format
enum_key_duplicate = 318, ///< an enum-keyed map has two keys that serialize to the same string
document_binary_value = 319, ///< a binary value is stored in a json_document
document_image_unsupported = 320, ///< a json_document image cannot be saved or loaded (discarded document, big-endian target)
document_image_unsupported = 320, ///< a discarded json_document cannot be saved as an image
discarded_value_used = 321, ///< a discarded value is used
// out_of_range
array_index_out_of_range = 401, ///< an array index is out of range
+4 -6
View File
@@ -182,12 +182,10 @@ inline void compact_nodes(const document_data& d, std::size_t arena_size, std::v
}
}
#if NLOHMANN_VIEW_LITTLE_ENDIAN // (the image format is little-endian: no save() elsewhere)
/// the document as an image
inline std::vector<std::uint8_t> save_image(const document_data& d)
{
#if !NLOHMANN_VIEW_LITTLE_ENDIAN
throw_type_error(exception_id::document_image_unsupported, "json_document images need a little-endian target"); // LCOV_EXCL_LINE
#endif
const std::size_t arena_size = d.arena_size;
std::vector<node> compacted;
std::string text_tail;
@@ -243,6 +241,7 @@ inline std::vector<std::uint8_t> save_image(const document_data& d)
*o = 0;
return image;
}
#endif
/// the parts of a number token that the checks need
struct number_token
@@ -663,13 +662,11 @@ inline bool check_image(const node* nodes, std::size_t count, const unsigned cha
throw_parse_error(exception_id::invalid_document_image, concat("invalid json_document image: ", what));
}
#if NLOHMANN_VIEW_LITTLE_ENDIAN // (the image format is little-endian: no load() elsewhere)
/// Read an image into d. The text and the decoded strings stay in the image;
/// the nodes are copied (so that they are aligned, and edits can change them).
inline void load_image(document_data& d, const std::uint8_t* image, std::size_t size, image_check check)
{
#if !NLOHMANN_VIEW_LITTLE_ENDIAN
throw_type_error(exception_id::document_image_unsupported, "json_document images need a little-endian target"); // LCOV_EXCL_LINE
#endif
if (image == nullptr || size < sizeof(image_header))
{
throw_invalid_image("too short");
@@ -738,6 +735,7 @@ inline void load_image(document_data& d, const std::uint8_t* image, std::size_t
std::vector<std::uint32_t>().swap(d.large_objects); // (only needed while building)
d.discarded = false;
}
#endif
} // namespace view
} // namespace detail
+5
View File
@@ -1182,6 +1182,10 @@ class basic_json_document
/// how load() checks an image (full, bounds, or none)
using image_check = detail::view::image_check;
#if NLOHMANN_VIEW_LITTLE_ENDIAN
// (the image format is little-endian: big-endian targets have no save()
// and load(), so that a call fails to compile instead of always throwing)
/// The document as an image that load() reads without parsing: the node
/// index, the text, and the decoded strings. An edited document is
/// written in its current state (floats that are not finite become null,
@@ -1223,6 +1227,7 @@ class basic_json_document
detail::view::load_image(*d.m_data, d.m_data->owned_image.data(), d.m_data->owned_image.size(), check);
return d;
}
#endif
///////////
// edits //
+1 -1
View File
@@ -5565,7 +5565,7 @@ enum class exception_id : int
type_not_serializable = 317, ///< a value cannot be serialized to the requested binary format
enum_key_duplicate = 318, ///< an enum-keyed map has two keys that serialize to the same string
document_binary_value = 319, ///< a binary value is stored in a json_document
document_image_unsupported = 320, ///< a json_document image cannot be saved or loaded (discarded document, big-endian target)
document_image_unsupported = 320, ///< a discarded json_document cannot be saved as an image
discarded_value_used = 321, ///< a discarded value is used
// out_of_range
array_index_out_of_range = 401, ///< an array index is out of range
+9 -6
View File
@@ -5007,12 +5007,10 @@ inline void compact_nodes(const document_data& d, std::size_t arena_size, std::v
}
}
#if NLOHMANN_VIEW_LITTLE_ENDIAN // (the image format is little-endian: no save() elsewhere)
/// the document as an image
inline std::vector<std::uint8_t> save_image(const document_data& d)
{
#if !NLOHMANN_VIEW_LITTLE_ENDIAN
throw_type_error(exception_id::document_image_unsupported, "json_document images need a little-endian target"); // LCOV_EXCL_LINE
#endif
const std::size_t arena_size = d.arena_size;
std::vector<node> compacted;
std::string text_tail;
@@ -5068,6 +5066,7 @@ inline std::vector<std::uint8_t> save_image(const document_data& d)
*o = 0;
return image;
}
#endif
/// the parts of a number token that the checks need
struct number_token
@@ -5488,13 +5487,11 @@ inline bool check_image(const node* nodes, std::size_t count, const unsigned cha
throw_parse_error(exception_id::invalid_document_image, concat("invalid json_document image: ", what));
}
#if NLOHMANN_VIEW_LITTLE_ENDIAN // (the image format is little-endian: no load() elsewhere)
/// Read an image into d. The text and the decoded strings stay in the image;
/// the nodes are copied (so that they are aligned, and edits can change them).
inline void load_image(document_data& d, const std::uint8_t* image, std::size_t size, image_check check)
{
#if !NLOHMANN_VIEW_LITTLE_ENDIAN
throw_type_error(exception_id::document_image_unsupported, "json_document images need a little-endian target"); // LCOV_EXCL_LINE
#endif
if (image == nullptr || size < sizeof(image_header))
{
throw_invalid_image("too short");
@@ -5563,6 +5560,7 @@ inline void load_image(document_data& d, const std::uint8_t* image, std::size_t
std::vector<std::uint32_t>().swap(d.large_objects); // (only needed while building)
d.discarded = false;
}
#endif
} // namespace view
} // namespace detail
@@ -8911,6 +8909,10 @@ class basic_json_document
/// how load() checks an image (full, bounds, or none)
using image_check = detail::view::image_check;
#if NLOHMANN_VIEW_LITTLE_ENDIAN
// (the image format is little-endian: big-endian targets have no save()
// and load(), so that a call fails to compile instead of always throwing)
/// The document as an image that load() reads without parsing: the node
/// index, the text, and the decoded strings. An edited document is
/// written in its current state (floats that are not finite become null,
@@ -8952,6 +8954,7 @@ class basic_json_document
detail::view::load_image(*d.m_data, d.m_data->owned_image.data(), d.m_data->owned_image.size(), check);
return d;
}
#endif
///////////
// edits //
+22 -1
View File
@@ -28,11 +28,25 @@ using nlohmann::detail::view::node;
#include <random>
#include <sstream>
#include <string>
#include <type_traits>
#include <utility>
#include <vector>
#include <test_data.hpp>
namespace json_view_image_test
{
// whether a document type has save() and load() (only on little-endian targets)
template<typename T, typename = void>
struct has_save : std::false_type {};
template<typename T>
struct has_save<T, decltype(static_cast<void>(std::declval<const T&>().save()))> : std::true_type {};
template<typename T, typename = void>
struct has_load : std::false_type {};
template<typename T>
struct has_load<T, decltype(static_cast<void>(T::load(std::declval<const std::vector<std::uint8_t>&>())))> : std::true_type {};
} // namespace json_view_image_test
#if !(defined(__BYTE_ORDER__) && defined(__ORDER_BIG_ENDIAN__) && __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
namespace
@@ -171,6 +185,9 @@ std::uint32_t rng()
TEST_CASE("json_view images: round trips")
{
static_assert(json_view_image_test::has_save<json_document>::value && json_view_image_test::has_load<json_document>::value, "save() and load() on a little-endian target");
static_assert(json_view_image_test::has_save<ordered_json_editable_document>::value && json_view_image_test::has_load<ordered_json_editable_document>::value, "save() and load() on a little-endian target");
SECTION("small documents")
{
for (const char* text :
@@ -1275,8 +1292,12 @@ TEST_CASE("json_view images: damaged images")
TEST_CASE("json_view images: big-endian targets")
{
// the image format is little-endian: save() and load() do not exist here
static_assert(!json_view_image_test::has_save<json_document>::value, "no save() on a big-endian target");
static_assert(!json_view_image_test::has_load<json_document>::value, "no load() on a big-endian target");
static_assert(!json_view_image_test::has_save<json_editable_document>::value, "no save() on a big-endian target");
const json_document d = json_document::parse("[1]");
CHECK_THROWS_WITH_AS(d.save(), "[json.exception.type_error.320] json_document images need a little-endian target", json::type_error&);
CHECK(d.root().size() == 1);
}
#endif