2012-06-23 01:36:55 +00:00
|
|
|
/*
|
|
|
|
Open Asset Import Library (assimp)
|
|
|
|
----------------------------------------------------------------------
|
|
|
|
|
2017-05-09 17:57:36 +00:00
|
|
|
Copyright (c) 2006-2017, assimp team
|
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
All rights reserved.
|
|
|
|
|
2015-04-20 22:36:47 +00:00
|
|
|
Redistribution and use of this software in source and binary forms,
|
|
|
|
with or without modification, are permitted provided that the
|
2012-06-23 01:36:55 +00:00
|
|
|
following conditions are met:
|
|
|
|
|
|
|
|
* Redistributions of source code must retain the above
|
|
|
|
copyright notice, this list of conditions and the
|
|
|
|
following disclaimer.
|
|
|
|
|
|
|
|
* Redistributions in binary form must reproduce the above
|
|
|
|
copyright notice, this list of conditions and the
|
|
|
|
following disclaimer in the documentation and/or other
|
|
|
|
materials provided with the distribution.
|
|
|
|
|
|
|
|
* Neither the name of the assimp team, nor the names of its
|
|
|
|
contributors may be used to endorse or promote products
|
|
|
|
derived from this software without specific prior
|
|
|
|
written permission of the assimp team.
|
|
|
|
|
2015-04-20 22:36:47 +00:00
|
|
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
|
|
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
2012-06-23 01:36:55 +00:00
|
|
|
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
2015-04-20 22:36:47 +00:00
|
|
|
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
2012-06-23 01:36:55 +00:00
|
|
|
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
2015-04-20 22:36:47 +00:00
|
|
|
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
2012-06-23 01:36:55 +00:00
|
|
|
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
2015-04-20 22:36:47 +00:00
|
|
|
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
|
|
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
|
|
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
2012-06-23 01:36:55 +00:00
|
|
|
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
|
|
|
|
----------------------------------------------------------------------
|
|
|
|
*/
|
|
|
|
|
|
|
|
/** @file FBXParser.cpp
|
|
|
|
* @brief Implementation of the FBX parser and the rudimentary DOM that we use
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef ASSIMP_BUILD_NO_FBX_IMPORTER
|
|
|
|
|
2012-08-11 02:16:46 +00:00
|
|
|
|
|
|
|
#ifdef ASSIMP_BUILD_NO_OWN_ZLIB
|
2015-05-19 03:57:13 +00:00
|
|
|
# include <zlib.h>
|
2012-08-11 02:16:46 +00:00
|
|
|
#else
|
2015-05-19 03:57:13 +00:00
|
|
|
# include "../contrib/zlib/zlib.h"
|
2012-08-11 02:16:46 +00:00
|
|
|
#endif
|
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
#include "FBXTokenizer.h"
|
|
|
|
#include "FBXParser.h"
|
2012-06-25 21:03:06 +00:00
|
|
|
#include "FBXUtil.h"
|
2012-06-23 01:36:55 +00:00
|
|
|
|
2012-06-26 16:31:19 +00:00
|
|
|
#include "ParsingUtils.h"
|
|
|
|
#include "fast_atof.h"
|
2015-04-20 22:36:47 +00:00
|
|
|
#include "ByteSwapper.h"
|
2012-06-26 16:31:19 +00:00
|
|
|
|
2016-03-03 14:19:38 +00:00
|
|
|
#include <iostream>
|
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
using namespace Assimp;
|
|
|
|
using namespace Assimp::FBX;
|
|
|
|
|
2012-06-25 21:03:06 +00:00
|
|
|
namespace {
|
|
|
|
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// signal parse error, this is always unrecoverable. Throws DeadlyImportError.
|
|
|
|
AI_WONT_RETURN void ParseError(const std::string& message, const Token& token) AI_WONT_RETURN_SUFFIX;
|
|
|
|
AI_WONT_RETURN void ParseError(const std::string& message, const Token& token)
|
|
|
|
{
|
|
|
|
throw DeadlyImportError(Util::AddTokenText("FBX-Parser",message,&token));
|
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
AI_WONT_RETURN void ParseError(const std::string& message, const Element* element = NULL) AI_WONT_RETURN_SUFFIX;
|
|
|
|
AI_WONT_RETURN void ParseError(const std::string& message, const Element* element)
|
|
|
|
{
|
|
|
|
if(element) {
|
|
|
|
ParseError(message,element->KeyToken());
|
|
|
|
}
|
|
|
|
throw DeadlyImportError("FBX-Parser " + message);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
void ParseError(const std::string& message, TokenPtr token)
|
|
|
|
{
|
|
|
|
if(token) {
|
|
|
|
ParseError(message, *token);
|
|
|
|
}
|
|
|
|
ParseError(message);
|
|
|
|
}
|
|
|
|
|
|
|
|
// Initially, we did reinterpret_cast, breaking strict aliasing rules.
|
|
|
|
// This actually caused trouble on Android, so let's be safe this time.
|
|
|
|
// https://github.com/assimp/assimp/issues/24
|
|
|
|
template <typename T>
|
|
|
|
T SafeParse(const char* data, const char* end) {
|
|
|
|
// Actual size validation happens during Tokenization so
|
|
|
|
// this is valid as an assertion.
|
|
|
|
ai_assert(static_cast<size_t>(end - data) >= sizeof(T));
|
|
|
|
T result = static_cast<T>(0);
|
|
|
|
::memcpy(&result, data, sizeof(T));
|
|
|
|
return result;
|
|
|
|
}
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
2012-06-26 16:31:19 +00:00
|
|
|
namespace Assimp {
|
|
|
|
namespace FBX {
|
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-06-26 17:19:13 +00:00
|
|
|
Element::Element(const Token& key_token, Parser& parser)
|
2012-06-26 16:31:19 +00:00
|
|
|
: key_token(key_token)
|
2012-06-23 01:36:55 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
TokenPtr n = NULL;
|
|
|
|
do {
|
|
|
|
n = parser.AdvanceToNextToken();
|
|
|
|
if(!n) {
|
|
|
|
ParseError("unexpected end of file, expected closing bracket",parser.LastToken());
|
|
|
|
}
|
|
|
|
|
|
|
|
if (n->Type() == TokenType_DATA) {
|
|
|
|
tokens.push_back(n);
|
2015-12-17 10:48:44 +00:00
|
|
|
TokenPtr prev = n;
|
2015-05-19 03:57:13 +00:00
|
|
|
n = parser.AdvanceToNextToken();
|
|
|
|
if(!n) {
|
|
|
|
ParseError("unexpected end of file, expected bracket, comma or key",parser.LastToken());
|
|
|
|
}
|
|
|
|
|
2015-12-17 10:48:44 +00:00
|
|
|
const TokenType ty = n->Type();
|
|
|
|
|
|
|
|
// some exporters are missing a comma on the next line
|
|
|
|
if (ty == TokenType_DATA && prev->Type() == TokenType_DATA && (n->Line() == prev->Line() + 1)) {
|
|
|
|
tokens.push_back(n);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (ty != TokenType_OPEN_BRACKET && ty != TokenType_CLOSE_BRACKET && ty != TokenType_COMMA && ty != TokenType_KEY) {
|
|
|
|
ParseError("unexpected token; expected bracket, comma or key",n);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (n->Type() == TokenType_OPEN_BRACKET) {
|
|
|
|
compound.reset(new Scope(parser));
|
|
|
|
|
|
|
|
// current token should be a TOK_CLOSE_BRACKET
|
|
|
|
n = parser.CurrentToken();
|
|
|
|
ai_assert(n);
|
|
|
|
|
|
|
|
if (n->Type() != TokenType_CLOSE_BRACKET) {
|
|
|
|
ParseError("expected closing bracket",n);
|
|
|
|
}
|
|
|
|
|
|
|
|
parser.AdvanceToNextToken();
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
while(n->Type() != TokenType_KEY && n->Type() != TokenType_CLOSE_BRACKET);
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
Element::~Element()
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
// no need to delete tokens, they are owned by the parser
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
2012-06-26 13:40:16 +00:00
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-06-25 21:03:06 +00:00
|
|
|
Scope::Scope(Parser& parser,bool topLevel)
|
2012-06-23 01:36:55 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
if(!topLevel) {
|
|
|
|
TokenPtr t = parser.CurrentToken();
|
|
|
|
if (t->Type() != TokenType_OPEN_BRACKET) {
|
|
|
|
ParseError("expected open bracket",t);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
TokenPtr n = parser.AdvanceToNextToken();
|
|
|
|
if(n == NULL) {
|
|
|
|
ParseError("unexpected end of file");
|
|
|
|
}
|
|
|
|
|
|
|
|
// note: empty scopes are allowed
|
|
|
|
while(n->Type() != TokenType_CLOSE_BRACKET) {
|
|
|
|
if (n->Type() != TokenType_KEY) {
|
|
|
|
ParseError("unexpected token, expected TOK_KEY",n);
|
|
|
|
}
|
|
|
|
|
|
|
|
const std::string& str = n->StringContents();
|
|
|
|
elements.insert(ElementMap::value_type(str,new_Element(*n,parser)));
|
|
|
|
|
|
|
|
// Element() should stop at the next Key token (or right after a Close token)
|
|
|
|
n = parser.CurrentToken();
|
|
|
|
if(n == NULL) {
|
|
|
|
if (topLevel) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
ParseError("unexpected end of file",parser.LastToken());
|
|
|
|
}
|
|
|
|
}
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
Scope::~Scope()
|
|
|
|
{
|
2016-04-05 20:53:54 +00:00
|
|
|
for(ElementMap::value_type& v : elements) {
|
2015-05-19 03:57:13 +00:00
|
|
|
delete v.second;
|
|
|
|
}
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-08-10 19:39:47 +00:00
|
|
|
Parser::Parser (const TokenList& tokens, bool is_binary)
|
2012-06-23 01:36:55 +00:00
|
|
|
: tokens(tokens)
|
2012-06-25 21:03:06 +00:00
|
|
|
, last()
|
2012-07-08 01:28:14 +00:00
|
|
|
, current()
|
|
|
|
, cursor(tokens.begin())
|
2012-08-10 19:39:47 +00:00
|
|
|
, is_binary(is_binary)
|
2012-06-23 01:36:55 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
root.reset(new Scope(*this,true));
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
Parser::~Parser()
|
|
|
|
{
|
2017-09-12 17:07:17 +00:00
|
|
|
// empty
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-06-25 21:03:06 +00:00
|
|
|
TokenPtr Parser::AdvanceToNextToken()
|
2012-06-23 01:36:55 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
last = current;
|
|
|
|
if (cursor == tokens.end()) {
|
|
|
|
current = NULL;
|
2017-09-12 17:07:17 +00:00
|
|
|
} else {
|
2015-05-19 03:57:13 +00:00
|
|
|
current = *cursor++;
|
|
|
|
}
|
|
|
|
return current;
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-06-25 21:03:06 +00:00
|
|
|
TokenPtr Parser::CurrentToken() const
|
2012-06-23 01:36:55 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
return current;
|
2012-06-25 21:03:06 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
TokenPtr Parser::LastToken() const
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
return last;
|
2012-06-23 01:36:55 +00:00
|
|
|
}
|
|
|
|
|
2012-06-26 16:31:19 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
uint64_t ParseTokenAsID(const Token& t, const char*& err_out)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'L') {
|
|
|
|
err_out = "failed to parse ID, unexpected data type, expected L(ong) (binary)";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
BE_NCONST uint64_t id = SafeParse<uint64_t>(data+1, t.end());
|
|
|
|
AI_SWAP8(id);
|
|
|
|
return id;
|
|
|
|
}
|
|
|
|
|
|
|
|
// XXX: should use size_t here
|
|
|
|
unsigned int length = static_cast<unsigned int>(t.end() - t.begin());
|
|
|
|
ai_assert(length > 0);
|
|
|
|
|
|
|
|
const char* out;
|
|
|
|
const uint64_t id = strtoul10_64(t.begin(),&out,&length);
|
|
|
|
if (out > t.end()) {
|
|
|
|
err_out = "failed to parse ID (text)";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
return id;
|
2012-06-26 16:31:19 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
2012-06-26 18:05:24 +00:00
|
|
|
size_t ParseTokenAsDim(const Token& t, const char*& err_out)
|
2012-06-26 16:31:19 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
// same as ID parsing, except there is a trailing asterisk
|
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'L') {
|
|
|
|
err_out = "failed to parse ID, unexpected data type, expected L(ong) (binary)";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
BE_NCONST uint64_t id = SafeParse<uint64_t>(data+1, t.end());
|
|
|
|
AI_SWAP8(id);
|
|
|
|
return static_cast<size_t>(id);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(*t.begin() != '*') {
|
|
|
|
err_out = "expected asterisk before array dimension";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
// XXX: should use size_t here
|
|
|
|
unsigned int length = static_cast<unsigned int>(t.end() - t.begin());
|
|
|
|
if(length == 0) {
|
|
|
|
err_out = "expected valid integer number after asterisk";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
const char* out;
|
|
|
|
const size_t id = static_cast<size_t>(strtoul10_64(t.begin() + 1,&out,&length));
|
|
|
|
if (out > t.end()) {
|
|
|
|
err_out = "failed to parse ID";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return id;
|
2012-06-26 16:31:19 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
float ParseTokenAsFloat(const Token& t, const char*& err_out)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return 0.0f;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'F' && data[0] != 'D') {
|
|
|
|
err_out = "failed to parse F(loat) or D(ouble), unexpected data type (binary)";
|
|
|
|
return 0.0f;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (data[0] == 'F') {
|
|
|
|
return SafeParse<float>(data+1, t.end());
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
return static_cast<float>( SafeParse<double>(data+1, t.end()) );
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// need to copy the input string to a temporary buffer
|
|
|
|
// first - next in the fbx token stream comes ',',
|
|
|
|
// which fast_atof could interpret as decimal point.
|
2012-06-27 23:49:52 +00:00
|
|
|
#define MAX_FLOAT_LENGTH 31
|
2015-05-19 03:57:13 +00:00
|
|
|
char temp[MAX_FLOAT_LENGTH + 1];
|
|
|
|
const size_t length = static_cast<size_t>(t.end()-t.begin());
|
|
|
|
std::copy(t.begin(),t.end(),temp);
|
|
|
|
temp[std::min(static_cast<size_t>(MAX_FLOAT_LENGTH),length)] = '\0';
|
2012-06-26 16:31:19 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return fast_atof(temp);
|
2012-06-26 16:31:19 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
int ParseTokenAsInt(const Token& t, const char*& err_out)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'I') {
|
|
|
|
err_out = "failed to parse I(nt), unexpected data type (binary)";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
BE_NCONST int32_t ival = SafeParse<int32_t>(data+1, t.end());
|
|
|
|
AI_SWAP4(ival);
|
|
|
|
return static_cast<int>(ival);
|
|
|
|
}
|
|
|
|
|
|
|
|
ai_assert(static_cast<size_t>(t.end() - t.begin()) > 0);
|
|
|
|
|
|
|
|
const char* out;
|
|
|
|
const int intval = strtol10(t.begin(),&out);
|
|
|
|
if (out != t.end()) {
|
|
|
|
err_out = "failed to parse ID";
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return intval;
|
2012-06-26 16:31:19 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2015-03-25 22:43:41 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
int64_t ParseTokenAsInt64(const Token& t, const char*& err_out)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'L') {
|
|
|
|
err_out = "failed to parse Int64, unexpected data type";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
BE_NCONST int64_t id = SafeParse<int64_t>(data + 1, t.end());
|
|
|
|
AI_SWAP8(id);
|
|
|
|
return id;
|
|
|
|
}
|
|
|
|
|
|
|
|
// XXX: should use size_t here
|
|
|
|
unsigned int length = static_cast<unsigned int>(t.end() - t.begin());
|
|
|
|
ai_assert(length > 0);
|
|
|
|
|
|
|
|
const char* out;
|
|
|
|
const int64_t id = strtol10_64(t.begin(), &out, &length);
|
|
|
|
if (out > t.end()) {
|
|
|
|
err_out = "failed to parse Int64 (text)";
|
|
|
|
return 0L;
|
|
|
|
}
|
|
|
|
|
|
|
|
return id;
|
2015-03-25 22:43:41 +00:00
|
|
|
}
|
|
|
|
|
2012-06-26 16:31:19 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
std::string ParseTokenAsString(const Token& t, const char*& err_out)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
err_out = NULL;
|
|
|
|
|
|
|
|
if (t.Type() != TokenType_DATA) {
|
|
|
|
err_out = "expected TOK_DATA token";
|
|
|
|
return "";
|
|
|
|
}
|
|
|
|
|
|
|
|
if(t.IsBinary())
|
|
|
|
{
|
|
|
|
const char* data = t.begin();
|
|
|
|
if (data[0] != 'S') {
|
|
|
|
err_out = "failed to parse S(tring), unexpected data type (binary)";
|
|
|
|
return "";
|
|
|
|
}
|
|
|
|
|
|
|
|
// read string length
|
|
|
|
BE_NCONST int32_t len = SafeParse<int32_t>(data+1, t.end());
|
|
|
|
AI_SWAP4(len);
|
|
|
|
|
|
|
|
ai_assert(t.end() - data == 5 + len);
|
|
|
|
return std::string(data + 5, len);
|
|
|
|
}
|
|
|
|
|
|
|
|
const size_t length = static_cast<size_t>(t.end() - t.begin());
|
|
|
|
if(length < 2) {
|
|
|
|
err_out = "token is too short to hold a string";
|
|
|
|
return "";
|
|
|
|
}
|
|
|
|
|
|
|
|
const char* s = t.begin(), *e = t.end() - 1;
|
|
|
|
if (*s != '\"' || *e != '\"') {
|
|
|
|
err_out = "expected double quoted string";
|
|
|
|
return "";
|
|
|
|
}
|
|
|
|
|
|
|
|
return std::string(s+1,length-2);
|
2012-06-26 16:31:19 +00:00
|
|
|
}
|
|
|
|
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2012-08-11 02:16:46 +00:00
|
|
|
namespace {
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read the type code and element count of a binary data array and stop there
|
2015-04-20 22:36:47 +00:00
|
|
|
void ReadBinaryDataArrayHead(const char*& data, const char* end, char& type, uint32_t& count,
|
2015-05-19 03:57:13 +00:00
|
|
|
const Element& el)
|
2012-08-11 02:16:46 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
if (static_cast<size_t>(end-data) < 5) {
|
|
|
|
ParseError("binary data array is too short, need five (5) bytes for type signature and element count",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// data type
|
|
|
|
type = *data;
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// read number of elements
|
|
|
|
BE_NCONST uint32_t len = SafeParse<uint32_t>(data+1, end);
|
|
|
|
AI_SWAP4(len);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
count = len;
|
|
|
|
data += 5;
|
2012-08-11 02:16:46 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read binary data array, assume cursor points to the 'compression mode' field (i.e. behind the header)
|
2015-04-20 22:36:47 +00:00
|
|
|
void ReadBinaryDataArray(char type, uint32_t count, const char*& data, const char* end,
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<char>& buff,
|
|
|
|
const Element& /*el*/)
|
2012-08-11 02:16:46 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
BE_NCONST uint32_t encmode = SafeParse<uint32_t>(data, end);
|
|
|
|
AI_SWAP4(encmode);
|
|
|
|
data += 4;
|
|
|
|
|
|
|
|
// next comes the compressed length
|
|
|
|
BE_NCONST uint32_t comp_len = SafeParse<uint32_t>(data, end);
|
|
|
|
AI_SWAP4(comp_len);
|
|
|
|
data += 4;
|
|
|
|
|
|
|
|
ai_assert(data + comp_len == end);
|
|
|
|
|
|
|
|
// determine the length of the uncompressed data by looking at the type signature
|
|
|
|
uint32_t stride = 0;
|
|
|
|
switch(type)
|
|
|
|
{
|
|
|
|
case 'f':
|
|
|
|
case 'i':
|
|
|
|
stride = 4;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 'd':
|
|
|
|
case 'l':
|
|
|
|
stride = 8;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
ai_assert(false);
|
|
|
|
};
|
|
|
|
|
|
|
|
const uint32_t full_length = stride * count;
|
|
|
|
buff.resize(full_length);
|
|
|
|
|
|
|
|
if(encmode == 0) {
|
|
|
|
ai_assert(full_length == comp_len);
|
|
|
|
|
|
|
|
// plain data, no compression
|
|
|
|
std::copy(data, end, buff.begin());
|
|
|
|
}
|
|
|
|
else if(encmode == 1) {
|
|
|
|
// zlib/deflate, next comes ZIP head (0x78 0x01)
|
|
|
|
// see http://www.ietf.org/rfc/rfc1950.txt
|
|
|
|
|
|
|
|
z_stream zstream;
|
|
|
|
zstream.opaque = Z_NULL;
|
|
|
|
zstream.zalloc = Z_NULL;
|
|
|
|
zstream.zfree = Z_NULL;
|
|
|
|
zstream.data_type = Z_BINARY;
|
|
|
|
|
|
|
|
// http://hewgill.com/journal/entries/349-how-to-decompress-gzip-stream-with-zlib
|
|
|
|
if(Z_OK != inflateInit(&zstream)) {
|
|
|
|
ParseError("failure initializing zlib");
|
|
|
|
}
|
|
|
|
|
|
|
|
zstream.next_in = reinterpret_cast<Bytef*>( const_cast<char*>(data) );
|
|
|
|
zstream.avail_in = comp_len;
|
|
|
|
|
2016-11-19 12:50:03 +00:00
|
|
|
zstream.avail_out = static_cast<uInt>(buff.size());
|
2015-05-19 03:57:13 +00:00
|
|
|
zstream.next_out = reinterpret_cast<Bytef*>(&*buff.begin());
|
|
|
|
const int ret = inflate(&zstream, Z_FINISH);
|
|
|
|
|
|
|
|
if (ret != Z_STREAM_END && ret != Z_OK) {
|
|
|
|
ParseError("failure decompressing compressed data section");
|
|
|
|
}
|
|
|
|
|
|
|
|
// terminate zlib
|
|
|
|
inflateEnd(&zstream);
|
|
|
|
}
|
2013-11-21 12:05:48 +00:00
|
|
|
#ifdef ASSIMP_BUILD_DEBUG
|
2015-05-19 03:57:13 +00:00
|
|
|
else {
|
|
|
|
// runtime check for this happens at tokenization stage
|
|
|
|
ai_assert(false);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
#endif
|
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
data += comp_len;
|
|
|
|
ai_assert(data == end);
|
2012-08-11 02:16:46 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
} // !anon
|
|
|
|
|
|
|
|
|
2012-08-10 21:24:58 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of float3 tuples
|
|
|
|
void ParseVectorDataArray(std::vector<aiVector3D>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
|
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
|
|
|
|
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
|
|
|
|
|
|
|
if(count % 3 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of three (3) (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (type != 'd' && type != 'f') {
|
|
|
|
ParseError("expected float or double array (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
|
|
|
|
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * (type == 'd' ? 8 : 4));
|
|
|
|
|
|
|
|
const uint32_t count3 = count / 3;
|
|
|
|
out.reserve(count3);
|
|
|
|
|
|
|
|
if (type == 'd') {
|
|
|
|
const double* d = reinterpret_cast<const double*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count3; ++i, d += 3) {
|
|
|
|
out.push_back(aiVector3D(static_cast<float>(d[0]),
|
|
|
|
static_cast<float>(d[1]),
|
|
|
|
static_cast<float>(d[2])));
|
|
|
|
}
|
2016-03-22 19:25:29 +00:00
|
|
|
// for debugging
|
|
|
|
/*for ( size_t i = 0; i < out.size(); i++ ) {
|
2016-03-03 14:19:38 +00:00
|
|
|
aiVector3D vec3( out[ i ] );
|
|
|
|
std::stringstream stream;
|
|
|
|
stream << " vec3.x = " << vec3.x << " vec3.y = " << vec3.y << " vec3.z = " << vec3.z << std::endl;
|
|
|
|
DefaultLogger::get()->info( stream.str() );
|
2016-03-22 19:25:29 +00:00
|
|
|
}*/
|
2015-05-19 03:57:13 +00:00
|
|
|
}
|
|
|
|
else if (type == 'f') {
|
|
|
|
const float* f = reinterpret_cast<const float*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count3; ++i, f += 3) {
|
|
|
|
out.push_back(aiVector3D(f[0],f[1],f[2]));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
|
|
|
|
|
|
|
// may throw bad_alloc if the input is rubbish, but this need
|
|
|
|
// not to be prevented - importing would fail but we wouldn't
|
|
|
|
// crash since assimp handles this case properly.
|
|
|
|
out.reserve(dim);
|
|
|
|
|
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
|
|
|
|
|
|
|
if (a.Tokens().size() % 3 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of three (3)",&el);
|
|
|
|
}
|
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
aiVector3D v;
|
|
|
|
v.x = ParseTokenAsFloat(**it++);
|
|
|
|
v.y = ParseTokenAsFloat(**it++);
|
|
|
|
v.z = ParseTokenAsFloat(**it++);
|
|
|
|
|
|
|
|
out.push_back(v);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of color4 tuples
|
|
|
|
void ParseVectorDataArray(std::vector<aiColor4D>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
|
|
|
|
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
|
|
|
|
|
|
|
if(count % 4 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of four (4) (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (type != 'd' && type != 'f') {
|
|
|
|
ParseError("expected float or double array (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
|
|
|
|
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * (type == 'd' ? 8 : 4));
|
|
|
|
|
|
|
|
const uint32_t count4 = count / 4;
|
|
|
|
out.reserve(count4);
|
|
|
|
|
|
|
|
if (type == 'd') {
|
|
|
|
const double* d = reinterpret_cast<const double*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count4; ++i, d += 4) {
|
|
|
|
out.push_back(aiColor4D(static_cast<float>(d[0]),
|
|
|
|
static_cast<float>(d[1]),
|
|
|
|
static_cast<float>(d[2]),
|
|
|
|
static_cast<float>(d[3])));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (type == 'f') {
|
|
|
|
const float* f = reinterpret_cast<const float*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count4; ++i, f += 4) {
|
|
|
|
out.push_back(aiColor4D(f[0],f[1],f[2],f[3]));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
|
|
|
|
|
|
|
// see notes in ParseVectorDataArray() above
|
|
|
|
out.reserve(dim);
|
|
|
|
|
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
|
|
|
|
|
|
|
if (a.Tokens().size() % 4 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of four (4)",&el);
|
|
|
|
}
|
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
aiColor4D v;
|
|
|
|
v.r = ParseTokenAsFloat(**it++);
|
|
|
|
v.g = ParseTokenAsFloat(**it++);
|
|
|
|
v.b = ParseTokenAsFloat(**it++);
|
|
|
|
v.a = ParseTokenAsFloat(**it++);
|
|
|
|
|
|
|
|
out.push_back(v);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of float2 tuples
|
|
|
|
void ParseVectorDataArray(std::vector<aiVector2D>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
|
|
|
|
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
|
|
|
|
|
|
|
if(count % 2 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of two (2) (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (type != 'd' && type != 'f') {
|
|
|
|
ParseError("expected float or double array (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
|
|
|
|
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * (type == 'd' ? 8 : 4));
|
|
|
|
|
|
|
|
const uint32_t count2 = count / 2;
|
|
|
|
out.reserve(count2);
|
|
|
|
|
|
|
|
if (type == 'd') {
|
|
|
|
const double* d = reinterpret_cast<const double*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count2; ++i, d += 2) {
|
|
|
|
out.push_back(aiVector2D(static_cast<float>(d[0]),
|
|
|
|
static_cast<float>(d[1])));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (type == 'f') {
|
|
|
|
const float* f = reinterpret_cast<const float*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count2; ++i, f += 2) {
|
|
|
|
out.push_back(aiVector2D(f[0],f[1]));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
|
|
|
|
|
|
|
// see notes in ParseVectorDataArray() above
|
|
|
|
out.reserve(dim);
|
|
|
|
|
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
|
|
|
|
|
|
|
if (a.Tokens().size() % 2 != 0) {
|
|
|
|
ParseError("number of floats is not a multiple of two (2)",&el);
|
|
|
|
}
|
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
aiVector2D v;
|
|
|
|
v.x = ParseTokenAsFloat(**it++);
|
|
|
|
v.y = ParseTokenAsFloat(**it++);
|
|
|
|
|
|
|
|
out.push_back(v);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of ints
|
|
|
|
void ParseVectorDataArray(std::vector<int>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (type != 'i') {
|
|
|
|
ParseError("expected int array (binary)",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * 4);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.reserve(count);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const int32_t* ip = reinterpret_cast<const int32_t*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++ip) {
|
|
|
|
BE_NCONST int32_t val = *ip;
|
|
|
|
AI_SWAP4(val);
|
|
|
|
out.push_back(val);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// see notes in ParseVectorDataArray()
|
|
|
|
out.reserve(dim);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
const int ival = ParseTokenAsInt(**it++);
|
|
|
|
out.push_back(ival);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of floats
|
|
|
|
void ParseVectorDataArray(std::vector<float>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
|
|
|
|
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
|
|
|
|
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (type != 'd' && type != 'f') {
|
|
|
|
ParseError("expected float or double array (binary)",&el);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
|
|
|
|
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * (type == 'd' ? 8 : 4));
|
|
|
|
|
|
|
|
if (type == 'd') {
|
|
|
|
const double* d = reinterpret_cast<const double*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++d) {
|
|
|
|
out.push_back(static_cast<float>(*d));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (type == 'f') {
|
|
|
|
const float* f = reinterpret_cast<const float*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++f) {
|
|
|
|
out.push_back(*f);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
|
|
|
|
|
|
|
// see notes in ParseVectorDataArray()
|
|
|
|
out.reserve(dim);
|
|
|
|
|
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
|
|
|
|
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
const float ival = ParseTokenAsFloat(**it++);
|
|
|
|
out.push_back(ival);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of uints
|
|
|
|
void ParseVectorDataArray(std::vector<unsigned int>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (type != 'i') {
|
|
|
|
ParseError("expected (u)int array (binary)",&el);
|
|
|
|
}
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * 4);
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.reserve(count);
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const int32_t* ip = reinterpret_cast<const int32_t*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++ip) {
|
|
|
|
BE_NCONST int32_t val = *ip;
|
|
|
|
if(val < 0) {
|
|
|
|
ParseError("encountered negative integer index (binary)");
|
|
|
|
}
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
AI_SWAP4(val);
|
|
|
|
out.push_back(val);
|
|
|
|
}
|
2012-08-11 02:48:08 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// see notes in ParseVectorDataArray()
|
|
|
|
out.reserve(dim);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
const int ival = ParseTokenAsInt(**it++);
|
|
|
|
if(ival < 0) {
|
|
|
|
ParseError("encountered negative integer index");
|
|
|
|
}
|
|
|
|
out.push_back(static_cast<unsigned int>(ival));
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of uint64_ts
|
|
|
|
void ParseVectorDataArray(std::vector<uint64_t>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if(tok.empty()) {
|
|
|
|
ParseError("unexpected empty element",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(!count) {
|
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (type != 'l') {
|
|
|
|
ParseError("expected long array (binary)",&el);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * 8);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.reserve(count);
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const uint64_t* ip = reinterpret_cast<const uint64_t*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++ip) {
|
|
|
|
BE_NCONST uint64_t val = *ip;
|
|
|
|
AI_SWAP8(val);
|
|
|
|
out.push_back(val);
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return;
|
|
|
|
}
|
2012-08-11 02:16:46 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// see notes in ParseVectorDataArray()
|
|
|
|
out.reserve(dim);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope,"a",&el);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end; ) {
|
|
|
|
const uint64_t ival = ParseTokenAsID(**it++);
|
2015-04-20 22:36:47 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.push_back(ival);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
2015-03-25 22:43:41 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// read an array of int64_ts
|
|
|
|
void ParseVectorDataArray(std::vector<int64_t>& out, const Element& el)
|
|
|
|
{
|
2016-03-03 14:19:38 +00:00
|
|
|
out.resize( 0 );
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& tok = el.Tokens();
|
|
|
|
if (tok.empty()) {
|
|
|
|
ParseError("unexpected empty element", &el);
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (tok[0]->IsBinary()) {
|
|
|
|
const char* data = tok[0]->begin(), *end = tok[0]->end();
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
char type;
|
|
|
|
uint32_t count;
|
|
|
|
ReadBinaryDataArrayHead(data, end, type, count, el);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (!count) {
|
|
|
|
return;
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if (type != 'l') {
|
|
|
|
ParseError("expected long array (binary)", &el);
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<char> buff;
|
|
|
|
ReadBinaryDataArray(type, count, data, end, buff, el);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
ai_assert(data == end);
|
|
|
|
ai_assert(buff.size() == count * 8);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.reserve(count);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const int64_t* ip = reinterpret_cast<const int64_t*>(&buff[0]);
|
|
|
|
for (unsigned int i = 0; i < count; ++i, ++ip) {
|
|
|
|
BE_NCONST int64_t val = *ip;
|
|
|
|
AI_SWAP8(val);
|
|
|
|
out.push_back(val);
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return;
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const size_t dim = ParseTokenAsDim(*tok[0]);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
// see notes in ParseVectorDataArray()
|
|
|
|
out.reserve(dim);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
const Scope& scope = GetRequiredScope(el);
|
|
|
|
const Element& a = GetRequiredElement(scope, "a", &el);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
for (TokenList::const_iterator it = a.Tokens().begin(), end = a.Tokens().end(); it != end;) {
|
|
|
|
const int64_t ival = ParseTokenAsInt64(**it++);
|
2015-03-25 22:43:41 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
out.push_back(ival);
|
|
|
|
}
|
2015-03-25 22:43:41 +00:00
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
aiMatrix4x4 ReadMatrix(const Element& element)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
std::vector<float> values;
|
|
|
|
ParseVectorDataArray(values,element);
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
if(values.size() != 16) {
|
|
|
|
ParseError("expected 16 matrix elements");
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
aiMatrix4x4 result;
|
2012-08-10 21:24:58 +00:00
|
|
|
|
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
result.a1 = values[0];
|
|
|
|
result.a2 = values[1];
|
|
|
|
result.a3 = values[2];
|
|
|
|
result.a4 = values[3];
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
result.b1 = values[4];
|
|
|
|
result.b2 = values[5];
|
|
|
|
result.b3 = values[6];
|
|
|
|
result.b4 = values[7];
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
result.c1 = values[8];
|
|
|
|
result.c2 = values[9];
|
|
|
|
result.c3 = values[10];
|
|
|
|
result.c4 = values[11];
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
result.d1 = values[12];
|
|
|
|
result.d2 = values[13];
|
|
|
|
result.d3 = values[14];
|
|
|
|
result.d4 = values[15];
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
result.Transpose();
|
|
|
|
return result;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsString() with ParseError handling
|
|
|
|
std::string ParseTokenAsString(const Token& t)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const std::string& i = ParseTokenAsString(t,err);
|
|
|
|
if(err) {
|
|
|
|
ParseError(err,t);
|
|
|
|
}
|
|
|
|
return i;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// extract a required element from a scope, abort if the element cannot be found
|
2015-04-20 22:36:47 +00:00
|
|
|
const Element& GetRequiredElement(const Scope& sc, const std::string& index, const Element* element /*= NULL*/)
|
2012-08-10 21:24:58 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const Element* el = sc[index];
|
|
|
|
if(!el) {
|
|
|
|
ParseError("did not find required element \"" + index + "\"",element);
|
|
|
|
}
|
|
|
|
return *el;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// extract required compound scope
|
|
|
|
const Scope& GetRequiredScope(const Element& el)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const Scope* const s = el.Compound();
|
|
|
|
if(!s) {
|
|
|
|
ParseError("expected compound scope",&el);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return *s;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// get token at a particular index
|
|
|
|
const Token& GetRequiredToken(const Element& el, unsigned int index)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const TokenList& t = el.Tokens();
|
|
|
|
if(index >= t.size()) {
|
|
|
|
ParseError(Formatter::format( "missing token at index " ) << index,&el);
|
|
|
|
}
|
2012-08-10 21:24:58 +00:00
|
|
|
|
2015-05-19 03:57:13 +00:00
|
|
|
return *t[index];
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsID() with ParseError handling
|
2015-04-20 22:36:47 +00:00
|
|
|
uint64_t ParseTokenAsID(const Token& t)
|
2012-08-10 21:24:58 +00:00
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const uint64_t i = ParseTokenAsID(t,err);
|
|
|
|
if(err) {
|
|
|
|
ParseError(err,t);
|
|
|
|
}
|
|
|
|
return i;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsDim() with ParseError handling
|
|
|
|
size_t ParseTokenAsDim(const Token& t)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const size_t i = ParseTokenAsDim(t,err);
|
|
|
|
if(err) {
|
|
|
|
ParseError(err,t);
|
|
|
|
}
|
|
|
|
return i;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsFloat() with ParseError handling
|
|
|
|
float ParseTokenAsFloat(const Token& t)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const float i = ParseTokenAsFloat(t,err);
|
|
|
|
if(err) {
|
|
|
|
ParseError(err,t);
|
|
|
|
}
|
|
|
|
return i;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsInt() with ParseError handling
|
|
|
|
int ParseTokenAsInt(const Token& t)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const int i = ParseTokenAsInt(t,err);
|
|
|
|
if(err) {
|
|
|
|
ParseError(err,t);
|
|
|
|
}
|
|
|
|
return i;
|
2012-08-10 21:24:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
2015-03-25 22:43:41 +00:00
|
|
|
// ------------------------------------------------------------------------------------------------
|
|
|
|
// wrapper around ParseTokenAsInt64() with ParseError handling
|
|
|
|
int64_t ParseTokenAsInt64(const Token& t)
|
|
|
|
{
|
2015-05-19 03:57:13 +00:00
|
|
|
const char* err;
|
|
|
|
const int64_t i = ParseTokenAsInt64(t, err);
|
|
|
|
if (err) {
|
|
|
|
ParseError(err, t);
|
|
|
|
}
|
|
|
|
return i;
|
2015-03-25 22:43:41 +00:00
|
|
|
}
|
|
|
|
|
2012-06-26 16:31:19 +00:00
|
|
|
} // !FBX
|
|
|
|
} // !Assimp
|
|
|
|
|
2012-06-23 01:36:55 +00:00
|
|
|
#endif
|