|
|
|
// Protocol Buffers - Google's data interchange format
|
|
|
|
// Copyright 2023 Google LLC. All rights reserved.
|
|
|
|
//
|
|
|
|
// Use of this source code is governed by a BSD-style
|
|
|
|
// license that can be found in the LICENSE file or at
|
|
|
|
// https://developers.google.com/open-source/licenses/bsd
|
|
|
|
|
|
|
|
#include "upb/message/promote.h"
|
|
|
|
|
|
|
|
#include <stdbool.h>
|
|
|
|
#include <stdint.h>
|
|
|
|
#include <string.h>
|
|
|
|
|
|
|
|
#include "upb/base/descriptor_constants.h"
|
|
|
|
#include "upb/mem/arena.h"
|
|
|
|
#include "upb/message/accessors.h"
|
|
|
|
#include "upb/message/array.h"
|
|
|
|
#include "upb/message/internal/array.h"
|
|
|
|
#include "upb/message/internal/extension.h"
|
|
|
|
#include "upb/message/internal/message.h"
|
|
|
|
#include "upb/message/internal/tagged_ptr.h"
|
|
|
|
#include "upb/message/map.h"
|
|
|
|
#include "upb/message/message.h"
|
|
|
|
#include "upb/message/tagged_ptr.h"
|
|
|
|
#include "upb/mini_table/extension.h"
|
|
|
|
#include "upb/mini_table/field.h"
|
|
|
|
#include "upb/mini_table/message.h"
|
|
|
|
#include "upb/mini_table/sub.h"
|
|
|
|
#include "upb/wire/decode.h"
|
|
|
|
#include "upb/wire/eps_copy_input_stream.h"
|
|
|
|
#include "upb/wire/reader.h"
|
|
|
|
|
|
|
|
// Must be last.
|
|
|
|
#include "upb/port/def.inc"
|
|
|
|
|
|
|
|
// Parses unknown data by merging into existing base_message or creating a
|
|
|
|
// new message usingg mini_table.
|
|
|
|
static upb_UnknownToMessageRet upb_MiniTable_ParseUnknownMessage(
|
|
|
|
const char* unknown_data, size_t unknown_size,
|
|
|
|
const upb_MiniTable* mini_table, upb_Message* base_message,
|
|
|
|
int decode_options, upb_Arena* arena) {
|
|
|
|
upb_UnknownToMessageRet ret;
|
|
|
|
ret.message =
|
|
|
|
base_message ? base_message : _upb_Message_New(mini_table, arena);
|
|
|
|
if (!ret.message) {
|
|
|
|
ret.status = kUpb_UnknownToMessage_OutOfMemory;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
// Decode sub message using unknown field contents.
|
|
|
|
const char* data = unknown_data;
|
|
|
|
uint32_t tag;
|
|
|
|
uint64_t message_len = 0;
|
|
|
|
data = upb_WireReader_ReadTag(data, &tag);
|
|
|
|
data = upb_WireReader_ReadVarint(data, &message_len);
|
|
|
|
upb_DecodeStatus status = upb_Decode(data, message_len, ret.message,
|
|
|
|
mini_table, NULL, decode_options, arena);
|
|
|
|
if (status == kUpb_DecodeStatus_OutOfMemory) {
|
|
|
|
ret.status = kUpb_UnknownToMessage_OutOfMemory;
|
|
|
|
} else if (status == kUpb_DecodeStatus_Ok) {
|
|
|
|
ret.status = kUpb_UnknownToMessage_Ok;
|
|
|
|
} else {
|
|
|
|
ret.status = kUpb_UnknownToMessage_ParseError;
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
upb_GetExtension_Status upb_MiniTable_GetOrPromoteExtension(
|
|
|
|
upb_Message* msg, const upb_MiniTableExtension* ext_table,
|
|
|
|
int decode_options, upb_Arena* arena, const upb_Extension** extension) {
|
|
|
|
UPB_ASSERT(upb_MiniTableExtension_CType(ext_table) == kUpb_CType_Message);
|
|
|
|
*extension = _upb_Message_Getext(msg, ext_table);
|
|
|
|
if (*extension) {
|
|
|
|
return kUpb_GetExtension_Ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Check unknown fields, if available promote.
|
|
|
|
int field_number = upb_MiniTableExtension_Number(ext_table);
|
|
|
|
upb_FindUnknownRet result = upb_Message_FindUnknown(msg, field_number, 0);
|
|
|
|
if (result.status != kUpb_FindUnknown_Ok) {
|
|
|
|
return kUpb_GetExtension_NotPresent;
|
|
|
|
}
|
|
|
|
size_t len;
|
|
|
|
size_t ofs = result.ptr - upb_Message_GetUnknown(msg, &len);
|
|
|
|
// Decode and promote from unknown.
|
|
|
|
const upb_MiniTable* extension_table =
|
|
|
|
upb_MiniTableExtension_GetSubMessage(ext_table);
|
|
|
|
upb_UnknownToMessageRet parse_result = upb_MiniTable_ParseUnknownMessage(
|
|
|
|
result.ptr, result.len, extension_table,
|
|
|
|
/* base_message= */ NULL, decode_options, arena);
|
|
|
|
switch (parse_result.status) {
|
|
|
|
case kUpb_UnknownToMessage_OutOfMemory:
|
|
|
|
return kUpb_GetExtension_OutOfMemory;
|
|
|
|
case kUpb_UnknownToMessage_ParseError:
|
|
|
|
return kUpb_GetExtension_ParseError;
|
|
|
|
case kUpb_UnknownToMessage_NotFound:
|
|
|
|
return kUpb_GetExtension_NotPresent;
|
|
|
|
case kUpb_UnknownToMessage_Ok:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
upb_Message* extension_msg = parse_result.message;
|
|
|
|
// Add to extensions.
|
|
|
|
upb_Extension* ext = _upb_Message_GetOrCreateExtension(msg, ext_table, arena);
|
|
|
|
if (!ext) {
|
|
|
|
return kUpb_GetExtension_OutOfMemory;
|
|
|
|
}
|
|
|
|
memcpy(&ext->data, &extension_msg, sizeof(extension_msg));
|
|
|
|
*extension = ext;
|
|
|
|
const char* delete_ptr = upb_Message_GetUnknown(msg, &len) + ofs;
|
|
|
|
upb_Message_DeleteUnknown(msg, delete_ptr, result.len);
|
|
|
|
return kUpb_GetExtension_Ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
static upb_FindUnknownRet upb_FindUnknownRet_ParseError(void) {
|
|
|
|
return (upb_FindUnknownRet){.status = kUpb_FindUnknown_ParseError};
|
|
|
|
}
|
|
|
|
|
|
|
|
upb_FindUnknownRet upb_Message_FindUnknown(const upb_Message* msg,
|
|
|
|
uint32_t field_number,
|
|
|
|
int depth_limit) {
|
|
|
|
depth_limit = depth_limit ? depth_limit : 100;
|
|
|
|
|
|
|
|
size_t size;
|
|
|
|
upb_FindUnknownRet ret;
|
|
|
|
|
|
|
|
const char* ptr = upb_Message_GetUnknown(msg, &size);
|
|
|
|
upb_EpsCopyInputStream stream;
|
|
|
|
upb_EpsCopyInputStream_Init(&stream, &ptr, size, true);
|
|
|
|
|
|
|
|
while (!upb_EpsCopyInputStream_IsDone(&stream, &ptr)) {
|
|
|
|
uint32_t tag;
|
|
|
|
const char* unknown_begin = ptr;
|
|
|
|
ptr = upb_WireReader_ReadTag(ptr, &tag);
|
|
|
|
if (!ptr) return upb_FindUnknownRet_ParseError();
|
|
|
|
if (field_number == upb_WireReader_GetFieldNumber(tag)) {
|
|
|
|
ret.status = kUpb_FindUnknown_Ok;
|
|
|
|
ret.ptr = upb_EpsCopyInputStream_GetAliasedPtr(&stream, unknown_begin);
|
|
|
|
ptr = _upb_WireReader_SkipValue(ptr, tag, depth_limit, &stream);
|
|
|
|
// Because we know that the input is a flat buffer, it is safe to perform
|
|
|
|
// pointer arithmetic on aliased pointers.
|
|
|
|
ret.len = upb_EpsCopyInputStream_GetAliasedPtr(&stream, ptr) - ret.ptr;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
ptr = _upb_WireReader_SkipValue(ptr, tag, depth_limit, &stream);
|
|
|
|
if (!ptr) return upb_FindUnknownRet_ParseError();
|
|
|
|
}
|
|
|
|
ret.status = kUpb_FindUnknown_NotPresent;
|
|
|
|
ret.ptr = NULL;
|
|
|
|
ret.len = 0;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
Added a new dynamic tree shaking model to upb, with the intention of removing the old model once YouTube has migrated.
The `kUpb_DecodeOption_ExperimentalAllowUnlinked` flag to the decoder will enable the new behavior. When that flag is not passed, tree shaking with the old model will still be possible.
"Dynamic tree shaking" in upb is a feature that allows messages to be parsed even if the MiniTables have not been fully linked. Unlinked sub-message fields can be parsed by preserving their data in the unknown fields. If the application later discovers that the message field is actually needed, the MiniTable can be patched to properly link that field, and existing message instances can "promote" the data from the unknown fields to an actual message of the correct type.
Before this change, dynamic tree shaking stored unparsed message data in the unknown fields of the *parent*. In effect, we were treating the field as if it did not exist at all. This meant that parsing an unlinked field did not affect the hasbits or oneof cases of the parent, nor did it create a `upb_Array` or `upb_Map` for array/map fields. Only when a message was linked and promoted did any of these things occur.
While this model had some amount of conceptual simplicity, it caused significant problems with oneofs. When multiple fields inside a single oneof are parsed from the wire, order matters, because later oneof fields must overwrite earlier ones. Dynamic tree shaking can mean that some fields in a oneof are linked while others are not. It is essential that we preserve this ordering semantic even when dynamic tree shaking is being used, but it is difficult to do if the oneof's data can be split between linked fields (which have been reified into parsed field data) and unlinked fields (whose data lives in the unknown fields of the parent).
To solve this problem, this CL changes the representation for unlinked fields. Instead of being placed in the parent's unknown fields, we create an actual message instance for each unlinked message we parse, but we use a placeholder "empty message" MiniTable as the message's type. All of the message's data will therefore be placed into the "empty message's" unknown fields. But unlike before, this "empty message" is actually present according to the hasbits, oneof case, and `upb_Array`/`upb_Map` of the parent. This means that all of the oneof presence logic works as normal.
Since the MiniTable can be patched at any time, we need a bit in the message instance itself to signal whether a pointer to a sub-message is an "empty message" or not. When dynamic tree shaking is in use, all users must be capable of recognizing an empty message and acting accordingly (promoting, etc) even if the MiniTable itself says that the field is linked.
Because dynamic tree shaking imposes this extra requirement on users, we require that users pass an extra option to the decoder to allow parsing of unlinked sub-messages. Many existing users of upb (Ruby, PHP, Python, etc) will always have fully-linked MiniTables, so there is no reason for them to add extra logic to handle empty messages. By omitting the `kUpb_DecodeOption_ExperimentalAllowUnlinked` option, they will be relieved of the duty to check the tagged pointer that would indicate an empty, unlinked message.
For existing users of dynamic tree shaking, there are three main changes:
1. The APIs in message/promote.h have changed, and users will need to update to the new interfaces.
2. The model for maps has changed slightly. Before, we required that map entries always had their values linked; for dynamic tree shaking to apply to maps, we required that the *entry* was left unlinked, not the entry's value. In the new model, that is reversed: map entries must always be linked, but a map entry's value can be unlinked.
3. The presence model for unlinked fields has changed. Unlinked fields will now register as "present" from the perspective of hasbits, oneof cases, and array/map entries. Users must test the tagged pointer to know if a message is of the correct, linked type or whether it is a placeholder "empty" message. There is a new function `upb_Message_GetTaggedMessagePtr()`, as well as a new accessor `upb_MessageValue.tagged_msg_val` that can be used to read and test the tagged pointer directly.
PiperOrigin-RevId: 535288031
2 years ago
|
|
|
static upb_DecodeStatus upb_Message_PromoteOne(upb_TaggedMessagePtr* tagged,
|
|
|
|
const upb_MiniTable* mini_table,
|
|
|
|
int decode_options,
|
|
|
|
upb_Arena* arena) {
|
|
|
|
upb_Message* empty =
|
|
|
|
UPB_PRIVATE(_upb_TaggedMessagePtr_GetEmptyMessage)(*tagged);
|
Added a new dynamic tree shaking model to upb, with the intention of removing the old model once YouTube has migrated.
The `kUpb_DecodeOption_ExperimentalAllowUnlinked` flag to the decoder will enable the new behavior. When that flag is not passed, tree shaking with the old model will still be possible.
"Dynamic tree shaking" in upb is a feature that allows messages to be parsed even if the MiniTables have not been fully linked. Unlinked sub-message fields can be parsed by preserving their data in the unknown fields. If the application later discovers that the message field is actually needed, the MiniTable can be patched to properly link that field, and existing message instances can "promote" the data from the unknown fields to an actual message of the correct type.
Before this change, dynamic tree shaking stored unparsed message data in the unknown fields of the *parent*. In effect, we were treating the field as if it did not exist at all. This meant that parsing an unlinked field did not affect the hasbits or oneof cases of the parent, nor did it create a `upb_Array` or `upb_Map` for array/map fields. Only when a message was linked and promoted did any of these things occur.
While this model had some amount of conceptual simplicity, it caused significant problems with oneofs. When multiple fields inside a single oneof are parsed from the wire, order matters, because later oneof fields must overwrite earlier ones. Dynamic tree shaking can mean that some fields in a oneof are linked while others are not. It is essential that we preserve this ordering semantic even when dynamic tree shaking is being used, but it is difficult to do if the oneof's data can be split between linked fields (which have been reified into parsed field data) and unlinked fields (whose data lives in the unknown fields of the parent).
To solve this problem, this CL changes the representation for unlinked fields. Instead of being placed in the parent's unknown fields, we create an actual message instance for each unlinked message we parse, but we use a placeholder "empty message" MiniTable as the message's type. All of the message's data will therefore be placed into the "empty message's" unknown fields. But unlike before, this "empty message" is actually present according to the hasbits, oneof case, and `upb_Array`/`upb_Map` of the parent. This means that all of the oneof presence logic works as normal.
Since the MiniTable can be patched at any time, we need a bit in the message instance itself to signal whether a pointer to a sub-message is an "empty message" or not. When dynamic tree shaking is in use, all users must be capable of recognizing an empty message and acting accordingly (promoting, etc) even if the MiniTable itself says that the field is linked.
Because dynamic tree shaking imposes this extra requirement on users, we require that users pass an extra option to the decoder to allow parsing of unlinked sub-messages. Many existing users of upb (Ruby, PHP, Python, etc) will always have fully-linked MiniTables, so there is no reason for them to add extra logic to handle empty messages. By omitting the `kUpb_DecodeOption_ExperimentalAllowUnlinked` option, they will be relieved of the duty to check the tagged pointer that would indicate an empty, unlinked message.
For existing users of dynamic tree shaking, there are three main changes:
1. The APIs in message/promote.h have changed, and users will need to update to the new interfaces.
2. The model for maps has changed slightly. Before, we required that map entries always had their values linked; for dynamic tree shaking to apply to maps, we required that the *entry* was left unlinked, not the entry's value. In the new model, that is reversed: map entries must always be linked, but a map entry's value can be unlinked.
3. The presence model for unlinked fields has changed. Unlinked fields will now register as "present" from the perspective of hasbits, oneof cases, and array/map entries. Users must test the tagged pointer to know if a message is of the correct, linked type or whether it is a placeholder "empty" message. There is a new function `upb_Message_GetTaggedMessagePtr()`, as well as a new accessor `upb_MessageValue.tagged_msg_val` that can be used to read and test the tagged pointer directly.
PiperOrigin-RevId: 535288031
2 years ago
|
|
|
size_t unknown_size;
|
|
|
|
const char* unknown_data = upb_Message_GetUnknown(empty, &unknown_size);
|
|
|
|
upb_Message* promoted = upb_Message_New(mini_table, arena);
|
|
|
|
if (!promoted) return kUpb_DecodeStatus_OutOfMemory;
|
|
|
|
upb_DecodeStatus status = upb_Decode(unknown_data, unknown_size, promoted,
|
|
|
|
mini_table, NULL, decode_options, arena);
|
|
|
|
if (status == kUpb_DecodeStatus_Ok) {
|
|
|
|
*tagged = UPB_PRIVATE(_upb_TaggedMessagePtr_Pack)(promoted, false);
|
Added a new dynamic tree shaking model to upb, with the intention of removing the old model once YouTube has migrated.
The `kUpb_DecodeOption_ExperimentalAllowUnlinked` flag to the decoder will enable the new behavior. When that flag is not passed, tree shaking with the old model will still be possible.
"Dynamic tree shaking" in upb is a feature that allows messages to be parsed even if the MiniTables have not been fully linked. Unlinked sub-message fields can be parsed by preserving their data in the unknown fields. If the application later discovers that the message field is actually needed, the MiniTable can be patched to properly link that field, and existing message instances can "promote" the data from the unknown fields to an actual message of the correct type.
Before this change, dynamic tree shaking stored unparsed message data in the unknown fields of the *parent*. In effect, we were treating the field as if it did not exist at all. This meant that parsing an unlinked field did not affect the hasbits or oneof cases of the parent, nor did it create a `upb_Array` or `upb_Map` for array/map fields. Only when a message was linked and promoted did any of these things occur.
While this model had some amount of conceptual simplicity, it caused significant problems with oneofs. When multiple fields inside a single oneof are parsed from the wire, order matters, because later oneof fields must overwrite earlier ones. Dynamic tree shaking can mean that some fields in a oneof are linked while others are not. It is essential that we preserve this ordering semantic even when dynamic tree shaking is being used, but it is difficult to do if the oneof's data can be split between linked fields (which have been reified into parsed field data) and unlinked fields (whose data lives in the unknown fields of the parent).
To solve this problem, this CL changes the representation for unlinked fields. Instead of being placed in the parent's unknown fields, we create an actual message instance for each unlinked message we parse, but we use a placeholder "empty message" MiniTable as the message's type. All of the message's data will therefore be placed into the "empty message's" unknown fields. But unlike before, this "empty message" is actually present according to the hasbits, oneof case, and `upb_Array`/`upb_Map` of the parent. This means that all of the oneof presence logic works as normal.
Since the MiniTable can be patched at any time, we need a bit in the message instance itself to signal whether a pointer to a sub-message is an "empty message" or not. When dynamic tree shaking is in use, all users must be capable of recognizing an empty message and acting accordingly (promoting, etc) even if the MiniTable itself says that the field is linked.
Because dynamic tree shaking imposes this extra requirement on users, we require that users pass an extra option to the decoder to allow parsing of unlinked sub-messages. Many existing users of upb (Ruby, PHP, Python, etc) will always have fully-linked MiniTables, so there is no reason for them to add extra logic to handle empty messages. By omitting the `kUpb_DecodeOption_ExperimentalAllowUnlinked` option, they will be relieved of the duty to check the tagged pointer that would indicate an empty, unlinked message.
For existing users of dynamic tree shaking, there are three main changes:
1. The APIs in message/promote.h have changed, and users will need to update to the new interfaces.
2. The model for maps has changed slightly. Before, we required that map entries always had their values linked; for dynamic tree shaking to apply to maps, we required that the *entry* was left unlinked, not the entry's value. In the new model, that is reversed: map entries must always be linked, but a map entry's value can be unlinked.
3. The presence model for unlinked fields has changed. Unlinked fields will now register as "present" from the perspective of hasbits, oneof cases, and array/map entries. Users must test the tagged pointer to know if a message is of the correct, linked type or whether it is a placeholder "empty" message. There is a new function `upb_Message_GetTaggedMessagePtr()`, as well as a new accessor `upb_MessageValue.tagged_msg_val` that can be used to read and test the tagged pointer directly.
PiperOrigin-RevId: 535288031
2 years ago
|
|
|
}
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
upb_DecodeStatus upb_Message_PromoteMessage(upb_Message* parent,
|
|
|
|
const upb_MiniTable* mini_table,
|
|
|
|
const upb_MiniTableField* field,
|
|
|
|
int decode_options,
|
|
|
|
upb_Arena* arena,
|
|
|
|
upb_Message** promoted) {
|
|
|
|
const upb_MiniTable* sub_table =
|
|
|
|
upb_MiniTable_GetSubMessageTable(mini_table, field);
|
|
|
|
UPB_ASSERT(sub_table);
|
|
|
|
upb_TaggedMessagePtr tagged =
|
|
|
|
upb_Message_GetTaggedMessagePtr(parent, field, NULL);
|
|
|
|
upb_DecodeStatus ret =
|
|
|
|
upb_Message_PromoteOne(&tagged, sub_table, decode_options, arena);
|
|
|
|
if (ret == kUpb_DecodeStatus_Ok) {
|
|
|
|
*promoted = upb_TaggedMessagePtr_GetNonEmptyMessage(tagged);
|
|
|
|
upb_Message_SetMessage(parent, mini_table, field, *promoted);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
upb_DecodeStatus upb_Array_PromoteMessages(upb_Array* arr,
|
|
|
|
const upb_MiniTable* mini_table,
|
|
|
|
int decode_options,
|
|
|
|
upb_Arena* arena) {
|
|
|
|
void** data = upb_Array_MutableDataPtr(arr);
|
|
|
|
size_t size = arr->UPB_PRIVATE(size);
|
Added a new dynamic tree shaking model to upb, with the intention of removing the old model once YouTube has migrated.
The `kUpb_DecodeOption_ExperimentalAllowUnlinked` flag to the decoder will enable the new behavior. When that flag is not passed, tree shaking with the old model will still be possible.
"Dynamic tree shaking" in upb is a feature that allows messages to be parsed even if the MiniTables have not been fully linked. Unlinked sub-message fields can be parsed by preserving their data in the unknown fields. If the application later discovers that the message field is actually needed, the MiniTable can be patched to properly link that field, and existing message instances can "promote" the data from the unknown fields to an actual message of the correct type.
Before this change, dynamic tree shaking stored unparsed message data in the unknown fields of the *parent*. In effect, we were treating the field as if it did not exist at all. This meant that parsing an unlinked field did not affect the hasbits or oneof cases of the parent, nor did it create a `upb_Array` or `upb_Map` for array/map fields. Only when a message was linked and promoted did any of these things occur.
While this model had some amount of conceptual simplicity, it caused significant problems with oneofs. When multiple fields inside a single oneof are parsed from the wire, order matters, because later oneof fields must overwrite earlier ones. Dynamic tree shaking can mean that some fields in a oneof are linked while others are not. It is essential that we preserve this ordering semantic even when dynamic tree shaking is being used, but it is difficult to do if the oneof's data can be split between linked fields (which have been reified into parsed field data) and unlinked fields (whose data lives in the unknown fields of the parent).
To solve this problem, this CL changes the representation for unlinked fields. Instead of being placed in the parent's unknown fields, we create an actual message instance for each unlinked message we parse, but we use a placeholder "empty message" MiniTable as the message's type. All of the message's data will therefore be placed into the "empty message's" unknown fields. But unlike before, this "empty message" is actually present according to the hasbits, oneof case, and `upb_Array`/`upb_Map` of the parent. This means that all of the oneof presence logic works as normal.
Since the MiniTable can be patched at any time, we need a bit in the message instance itself to signal whether a pointer to a sub-message is an "empty message" or not. When dynamic tree shaking is in use, all users must be capable of recognizing an empty message and acting accordingly (promoting, etc) even if the MiniTable itself says that the field is linked.
Because dynamic tree shaking imposes this extra requirement on users, we require that users pass an extra option to the decoder to allow parsing of unlinked sub-messages. Many existing users of upb (Ruby, PHP, Python, etc) will always have fully-linked MiniTables, so there is no reason for them to add extra logic to handle empty messages. By omitting the `kUpb_DecodeOption_ExperimentalAllowUnlinked` option, they will be relieved of the duty to check the tagged pointer that would indicate an empty, unlinked message.
For existing users of dynamic tree shaking, there are three main changes:
1. The APIs in message/promote.h have changed, and users will need to update to the new interfaces.
2. The model for maps has changed slightly. Before, we required that map entries always had their values linked; for dynamic tree shaking to apply to maps, we required that the *entry* was left unlinked, not the entry's value. In the new model, that is reversed: map entries must always be linked, but a map entry's value can be unlinked.
3. The presence model for unlinked fields has changed. Unlinked fields will now register as "present" from the perspective of hasbits, oneof cases, and array/map entries. Users must test the tagged pointer to know if a message is of the correct, linked type or whether it is a placeholder "empty" message. There is a new function `upb_Message_GetTaggedMessagePtr()`, as well as a new accessor `upb_MessageValue.tagged_msg_val` that can be used to read and test the tagged pointer directly.
PiperOrigin-RevId: 535288031
2 years ago
|
|
|
for (size_t i = 0; i < size; i++) {
|
|
|
|
upb_TaggedMessagePtr tagged;
|
|
|
|
memcpy(&tagged, &data[i], sizeof(tagged));
|
|
|
|
if (!upb_TaggedMessagePtr_IsEmpty(tagged)) continue;
|
|
|
|
upb_DecodeStatus status =
|
|
|
|
upb_Message_PromoteOne(&tagged, mini_table, decode_options, arena);
|
|
|
|
if (status != kUpb_DecodeStatus_Ok) return status;
|
|
|
|
memcpy(&data[i], &tagged, sizeof(tagged));
|
|
|
|
}
|
|
|
|
return kUpb_DecodeStatus_Ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
upb_DecodeStatus upb_Map_PromoteMessages(upb_Map* map,
|
|
|
|
const upb_MiniTable* mini_table,
|
|
|
|
int decode_options, upb_Arena* arena) {
|
|
|
|
size_t iter = kUpb_Map_Begin;
|
|
|
|
upb_MessageValue key, val;
|
|
|
|
while (upb_Map_Next(map, &key, &val, &iter)) {
|
|
|
|
if (!upb_TaggedMessagePtr_IsEmpty(val.tagged_msg_val)) continue;
|
|
|
|
upb_DecodeStatus status = upb_Message_PromoteOne(
|
|
|
|
&val.tagged_msg_val, mini_table, decode_options, arena);
|
|
|
|
if (status != kUpb_DecodeStatus_Ok) return status;
|
|
|
|
upb_Map_SetEntryValue(map, iter, val);
|
|
|
|
}
|
|
|
|
return kUpb_DecodeStatus_Ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
////////////////////////////////////////////////////////////////////////////////
|
|
|
|
// OLD promotion functions, will be removed!
|
|
|
|
////////////////////////////////////////////////////////////////////////////////
|
|
|
|
|
|
|
|
// Warning: See TODO
|
|
|
|
upb_UnknownToMessageRet upb_MiniTable_PromoteUnknownToMessage(
|
|
|
|
upb_Message* msg, const upb_MiniTable* mini_table,
|
|
|
|
const upb_MiniTableField* field, const upb_MiniTable* sub_mini_table,
|
|
|
|
int decode_options, upb_Arena* arena) {
|
|
|
|
upb_FindUnknownRet unknown;
|
|
|
|
// We need to loop and merge unknowns that have matching tag field->number.
|
|
|
|
upb_Message* message = NULL;
|
|
|
|
// Callers should check that message is not set first before calling
|
|
|
|
// PromotoUnknownToMessage.
|
|
|
|
UPB_ASSERT(upb_MiniTable_GetSubMessageTable(mini_table, field) ==
|
|
|
|
sub_mini_table);
|
|
|
|
bool is_oneof = upb_MiniTableField_IsInOneof(field);
|
|
|
|
if (!is_oneof || UPB_PRIVATE(_upb_Message_GetOneofCase)(msg, field) ==
|
|
|
|
upb_MiniTableField_Number(field)) {
|
|
|
|
UPB_ASSERT(upb_Message_GetMessage(msg, field, NULL) == NULL);
|
|
|
|
}
|
|
|
|
upb_UnknownToMessageRet ret;
|
|
|
|
ret.status = kUpb_UnknownToMessage_Ok;
|
|
|
|
do {
|
|
|
|
unknown =
|
|
|
|
upb_Message_FindUnknown(msg, upb_MiniTableField_Number(field),
|
|
|
|
upb_DecodeOptions_GetMaxDepth(decode_options));
|
|
|
|
switch (unknown.status) {
|
|
|
|
case kUpb_FindUnknown_Ok: {
|
|
|
|
const char* unknown_data = unknown.ptr;
|
|
|
|
size_t unknown_size = unknown.len;
|
|
|
|
ret = upb_MiniTable_ParseUnknownMessage(unknown_data, unknown_size,
|
|
|
|
sub_mini_table, message,
|
|
|
|
decode_options, arena);
|
|
|
|
if (ret.status == kUpb_UnknownToMessage_Ok) {
|
|
|
|
message = ret.message;
|
|
|
|
upb_Message_DeleteUnknown(msg, unknown_data, unknown_size);
|
|
|
|
}
|
|
|
|
} break;
|
|
|
|
case kUpb_FindUnknown_ParseError:
|
|
|
|
ret.status = kUpb_UnknownToMessage_ParseError;
|
|
|
|
break;
|
|
|
|
case kUpb_FindUnknown_NotPresent:
|
|
|
|
// If we parsed at least one unknown, we are done.
|
|
|
|
ret.status =
|
|
|
|
message ? kUpb_UnknownToMessage_Ok : kUpb_UnknownToMessage_NotFound;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
} while (unknown.status == kUpb_FindUnknown_Ok);
|
|
|
|
if (message) {
|
|
|
|
if (is_oneof) {
|
|
|
|
UPB_PRIVATE(_upb_Message_SetOneofCase)(msg, field);
|
|
|
|
}
|
|
|
|
upb_Message_SetMessage(msg, mini_table, field, message);
|
|
|
|
ret.message = message;
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Moves repeated messages in unknowns to a upb_Array.
|
|
|
|
//
|
|
|
|
// Since the repeated field is not a scalar type we don't check for
|
|
|
|
// kUpb_LabelFlags_IsPacked.
|
|
|
|
// TODO: Optimize. Instead of converting messages one at a time,
|
|
|
|
// scan all unknown data once and compact.
|
|
|
|
upb_UnknownToMessage_Status upb_MiniTable_PromoteUnknownToMessageArray(
|
|
|
|
upb_Message* msg, const upb_MiniTableField* field,
|
|
|
|
const upb_MiniTable* mini_table, int decode_options, upb_Arena* arena) {
|
|
|
|
upb_Array* repeated_messages = upb_Message_GetMutableArray(msg, field);
|
|
|
|
// Find all unknowns with given field number and parse.
|
|
|
|
upb_FindUnknownRet unknown;
|
|
|
|
do {
|
|
|
|
unknown =
|
|
|
|
upb_Message_FindUnknown(msg, upb_MiniTableField_Number(field),
|
|
|
|
upb_DecodeOptions_GetMaxDepth(decode_options));
|
|
|
|
if (unknown.status == kUpb_FindUnknown_Ok) {
|
|
|
|
upb_UnknownToMessageRet ret = upb_MiniTable_ParseUnknownMessage(
|
|
|
|
unknown.ptr, unknown.len, mini_table,
|
|
|
|
/* base_message= */ NULL, decode_options, arena);
|
|
|
|
if (ret.status == kUpb_UnknownToMessage_Ok) {
|
|
|
|
upb_MessageValue value;
|
|
|
|
value.msg_val = ret.message;
|
|
|
|
// Allocate array on demand before append.
|
|
|
|
if (!repeated_messages) {
|
|
|
|
upb_Message_ResizeArrayUninitialized(msg, field, 0, arena);
|
|
|
|
repeated_messages = upb_Message_GetMutableArray(msg, field);
|
|
|
|
}
|
|
|
|
if (!upb_Array_Append(repeated_messages, value, arena)) {
|
|
|
|
return kUpb_UnknownToMessage_OutOfMemory;
|
|
|
|
}
|
|
|
|
upb_Message_DeleteUnknown(msg, unknown.ptr, unknown.len);
|
|
|
|
} else {
|
|
|
|
return ret.status;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
} while (unknown.status == kUpb_FindUnknown_Ok);
|
|
|
|
return kUpb_UnknownToMessage_Ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Moves repeated messages in unknowns to a upb_Map.
|
|
|
|
upb_UnknownToMessage_Status upb_MiniTable_PromoteUnknownToMap(
|
|
|
|
upb_Message* msg, const upb_MiniTable* mini_table,
|
|
|
|
const upb_MiniTableField* field, int decode_options, upb_Arena* arena) {
|
|
|
|
// TODO: use a variant of upb_MiniTable_GetSubMessageTable() here.
|
|
|
|
const upb_MiniTable* map_entry_mini_table = upb_MiniTableSub_Message(
|
|
|
|
mini_table->UPB_PRIVATE(subs)[field->UPB_PRIVATE(submsg_index)]);
|
|
|
|
UPB_ASSERT(map_entry_mini_table);
|
|
|
|
UPB_ASSERT(map_entry_mini_table);
|
|
|
|
UPB_ASSERT(map_entry_mini_table->UPB_PRIVATE(field_count) == 2);
|
|
|
|
UPB_ASSERT(upb_MiniTableField_IsMap(field));
|
|
|
|
// Find all unknowns with given field number and parse.
|
|
|
|
upb_FindUnknownRet unknown;
|
|
|
|
while (1) {
|
|
|
|
unknown =
|
|
|
|
upb_Message_FindUnknown(msg, upb_MiniTableField_Number(field),
|
|
|
|
upb_DecodeOptions_GetMaxDepth(decode_options));
|
|
|
|
if (unknown.status != kUpb_FindUnknown_Ok) break;
|
|
|
|
upb_UnknownToMessageRet ret = upb_MiniTable_ParseUnknownMessage(
|
|
|
|
unknown.ptr, unknown.len, map_entry_mini_table,
|
|
|
|
/* base_message= */ NULL, decode_options, arena);
|
|
|
|
if (ret.status != kUpb_UnknownToMessage_Ok) return ret.status;
|
|
|
|
// Allocate map on demand before append.
|
|
|
|
upb_Map* map = upb_Message_GetOrCreateMutableMap(msg, map_entry_mini_table,
|
|
|
|
field, arena);
|
|
|
|
upb_Message* map_entry_message = ret.message;
|
|
|
|
bool insert_success = upb_Message_SetMapEntry(map, mini_table, field,
|
|
|
|
map_entry_message, arena);
|
|
|
|
if (!insert_success) return kUpb_UnknownToMessage_OutOfMemory;
|
|
|
|
upb_Message_DeleteUnknown(msg, unknown.ptr, unknown.len);
|
|
|
|
}
|
|
|
|
return kUpb_UnknownToMessage_Ok;
|
|
|
|
}
|