PtexReader.cpp

Go to the documentation of this file.

/*
PTEX SOFTWARE
Copyright 2014 Disney Enterprises, Inc.  All rights reserved
 
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the 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.
 
  * The names "Disney", "Walt Disney Pictures", "Walt Disney Animation
    Studios" or the names of its contributors may NOT be used to
    endorse or promote products derived from this software without
    specific prior written permission from Walt Disney Pictures.
 
Disclaimer: THIS SOFTWARE IS PROVIDED BY WALT DISNEY PICTURES AND
CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS
FOR A PARTICULAR PURPOSE, NONINFRINGEMENT AND TITLE ARE DISCLAIMED.
IN NO EVENT SHALL WALT DISNEY PICTURES, THE COPYRIGHT HOLDER OR
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND BASED ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
*/
 
#include "PtexPlatform.h"
#include <iostream>
#include <sstream>
#include <stdio.h>
 
#include <libdeflate.h>
 
#include "Ptexture.h"
#include "PtexUtils.h"
#include "PtexReader.h"
 
namespace {
    class TempErrorHandler : public PtexErrorHandler
    {
        std::string _error;
    public:
        virtual void reportError(const char* error) {
            _error += error;
        }
        const std::string& getErrorString() const { return _error; }
    };
}
 
PTEX_NAMESPACE_BEGIN
 
PtexTexture* PtexTexture::open(const char* path, Ptex::String& error, bool premultiply)
{
    PtexReader* reader = new PtexReader(premultiply, (PtexInputHandler*) 0, (PtexErrorHandler*) 0);
    bool ok = reader->open(path, error);
    if (!ok) {
        reader->release();
        return 0;
    }
    return reader;
}
 
 
PtexReader::PtexReader(bool premultiply, PtexInputHandler* io, PtexErrorHandler* err)
    : _io(io ? io : &_defaultIo),
      _err(err),
      _premultiply(premultiply),
      _ok(true),
      _needToOpen(true),
      _pendingPurge(false),
      _fp(0),
      _pos(0),
      _pixelsize(0),
      _constdata(0),
      _metadata(0),
      _baseMemUsed(sizeof(*this)),
      _memUsed(_baseMemUsed),
      _opens(0),
      _blockReads(0)
{
    _decompressor = libdeflate_alloc_decompressor();
}
 
 
PtexReader::~PtexReader()
{
    closeFP();
    if (_constdata) delete [] _constdata;
    if (_metadata) delete _metadata;
 
    for (std::vector<Level*>::iterator i = _levels.begin(); i != _levels.end(); ++i) {
        if (*i) delete *i;
    }
    libdeflate_free_decompressor(_decompressor);
}
 
void PtexReader::prune()
{
    if (_metadata) { delete _metadata; _metadata = 0; }
    for (std::vector<Level*>::iterator i = _levels.begin(); i != _levels.end(); ++i) {
        if (*i) { delete *i; *i = 0; }
    }
    _reductions.clear();
    _memUsed = _baseMemUsed;
}
 
 
void PtexReader::purge()
{
    // free all dynamic data
    prune();
    if (_constdata) {delete [] _constdata; _constdata = 0; }
    std::vector<FaceInfo>().swap(_faceinfo);
    std::vector<uint32_t>().swap(_rfaceids);
    std::vector<LevelInfo>().swap(_levelinfo);
    std::vector<FilePos>().swap(_levelpos);
    std::vector<Level*>().swap(_levels);
    closeFP();
 
    // reset initial state
    _ok = true;
    _needToOpen = true;
    _pendingPurge = false;
    _memUsed = _baseMemUsed = sizeof(*this);
}
 
 
bool PtexReader::open(const char* pathArg, Ptex::String& error)
{
    AutoMutex locker(readlock);
    if (!needToOpen()) return false;
 
    if (!LittleEndian()) {
        error = "Ptex library doesn't currently support big-endian cpu's";
        return 0;
    }
    _path = pathArg;
    _fp = _io->open(pathArg);
    if (!_fp) {
        std::string errstr = "Can't open ptex file: ";
        errstr += pathArg; errstr += "\n"; errstr += _io->lastError();
        error = errstr.c_str();
        _ok = 0;
        return 0;
    }
    memset(&_header, 0, sizeof(_header));
    readBlock(&_header, HeaderSize);
    if (_header.magic != Magic) {
        std::string errstr = "Not a ptex file: "; errstr += pathArg;
        error = errstr.c_str();
        _ok = 0;
        closeFP();
        return 0;
    }
    if (_header.version != 1) {
        std::stringstream s;
        s << "Unsupported ptex file version ("<< _header.version << "): " << pathArg;
        error = s.str();
        _ok = 0;
        closeFP();
        return 0;
    }
    if (!(_header.meshtype == mt_triangle || _header.meshtype == mt_quad)) {
        std::stringstream s;
        s << "Invalid mesh type (" << _header.meshtype << "): " << pathArg;
        error = s.str();
        _ok = 0;
        closeFP();
        return 0;
    }
    _pixelsize = _header.pixelSize();
    _errorPixel.resize(_pixelsize);
 
    // install temp error handler to capture error (to return in error param)
    TempErrorHandler tempErr;
    PtexErrorHandler* prevErr = _err;
    _err = &tempErr;
 
    // read extended header
    memset(&_extheader, 0, sizeof(_extheader));
    readBlock(&_extheader, std::min(uint32_t(ExtHeaderSize), _header.extheadersize));
 
    // compute offsets of various blocks
    FilePos pos = HeaderSize + _header.extheadersize;
    _faceinfopos = pos;   pos += _header.faceinfosize;
    _constdatapos = pos;  pos += _header.constdatasize;
    _levelinfopos = pos;  pos += _header.levelinfosize;
    _leveldatapos = pos;  pos += _header.leveldatasize;
    _metadatapos = pos;   pos += _header.metadatazipsize;
                          pos += sizeof(uint64_t); // compatibility barrier
    _lmdheaderpos = pos;  pos += _extheader.lmdheaderzipsize;
    _lmddatapos = pos;    pos += _extheader.lmddatasize;
 
    // read basic file info
    readFaceInfo();
    readConstData();
    readLevelInfo();
    _baseMemUsed = _memUsed;
 
    // restore error handler
    _err = prevErr;
 
    // an error occurred while reading the file
    if (!_ok) {
        error = tempErr.getErrorString();
        closeFP();
        return 0;
    }
    AtomicStore(&_needToOpen, false);
    return true;
}
 
bool PtexReader::tryClose()
{
    if (_fp) {
        if (!readlock.trylock()) return false;
        closeFP();
        readlock.unlock();
    }
    return true;
}
 
 
void PtexReader::closeFP()
{
    if (_fp) {
        _io->close(_fp);
        _fp = 0;
    }
}
 
 
bool PtexReader::reopenFP()
{
    if (_fp) return true;
 
    // we assume this is called lazily in a scope where readlock is already held
    _fp = _io->open(_path.c_str());
    if (!_fp) {
        setIOError("Can't reopen");
        return false;
    }
    _pos = 0;
    Header headerval;
    ExtHeader extheaderval;
    readBlock(&headerval, HeaderSize);
    memset(&extheaderval, 0, sizeof(extheaderval));
    readBlock(&extheaderval, std::min(uint32_t(ExtHeaderSize), headerval.extheadersize));
    if (0 != memcmp(&headerval, &_header, sizeof(headerval)) ||
        0 != memcmp(&extheaderval, &_extheader, sizeof(extheaderval)))
    {
        setError("Header mismatch on reopen of");
        return false;
    }
    logOpen();
    return true;
}
 
 
const Ptex::FaceInfo& PtexReader::getFaceInfo(int faceid)
{
    if (faceid >= 0 && uint32_t(faceid) < _faceinfo.size())
        return _faceinfo[faceid];
 
    static Ptex::FaceInfo dummy;
    return dummy;
}
 
 
void PtexReader::readFaceInfo()
{
    if (_faceinfo.empty()) {
        // read compressed face info block
        seek(_faceinfopos);
        int nfaces = _header.nfaces;
        _faceinfo.resize(nfaces);
        readZipBlock(&_faceinfo[0], _header.faceinfosize,
                     (int)(sizeof(FaceInfo)*nfaces));
 
        // generate rfaceids
        _rfaceids.resize(nfaces);
        std::vector<uint32_t> faceids_r(nfaces);
        PtexUtils::genRfaceids(&_faceinfo[0], nfaces,
                               &_rfaceids[0], &faceids_r[0]);
        increaseMemUsed(nfaces * (sizeof(_faceinfo[0]) + sizeof(_rfaceids[0])));
    }
}
 
 
 
void PtexReader::readLevelInfo()
{
    if (_levelinfo.empty()) {
        // read level info block
        seek(_levelinfopos);
        _levelinfo.resize(_header.nlevels);
        readBlock(&_levelinfo[0], LevelInfoSize*_header.nlevels);
 
        // initialize related data
        _levels.resize(_header.nlevels);
        _levelpos.resize(_header.nlevels);
        FilePos pos = _leveldatapos;
        for (int i = 0; i < _header.nlevels; i++) {
            _levelpos[i] = pos;
            pos += _levelinfo[i].leveldatasize;
        }
        increaseMemUsed(_header.nlevels * sizeof(_levelinfo[0]) + sizeof(_levels[0]) + sizeof(_levelpos[0]));
    }
}
 
 
void PtexReader::readConstData()
{
    if (!_constdata) {
        // read compressed constant data block
        seek(_constdatapos);
        int size = _pixelsize * _header.nfaces;
        _constdata = new uint8_t[size];
        readZipBlock(_constdata, _header.constdatasize, size);
        if (_premultiply && _header.hasAlpha())
            PtexUtils::multalpha(_constdata, _header.nfaces, datatype(),
                                 _header.nchannels, _header.alphachan);
        increaseMemUsed(size);
    }
}
 
 
PtexMetaData* PtexReader::getMetaData()
{
    if (!_metadata) readMetaData();
    return _metadata;
}
 
 
PtexReader::MetaData::Entry*
PtexReader::MetaData::getEntry(int index)
{
    if (index < 0 || index >= int(_entries.size())) {
        return 0;
    }
 
    Entry* e = _entries[index];
    if (!e->isLmd) {
        // normal (small) meta data - just return directly
        return e;
    }
 
    // large meta data - may not be read in yet
    if (e->lmdData) {
        // already in memory
        return e;
    }
    else {
        // not present, must read from file
 
        // get read lock and make sure we still need to read
        AutoMutex locker(_reader->readlock);
        if (e->lmdData) {
            return e;
        }
        // go ahead and read, keep local until finished
        LargeMetaData* lmdData = new LargeMetaData(e->datasize);
        e->data = (char*) lmdData->data();
        _reader->increaseMemUsed(sizeof(LargeMetaData) + e->datasize);
        _reader->seek(e->lmdPos);
        _reader->readZipBlock(e->data, e->lmdZipSize, e->datasize);
        // update entry
        e->lmdData = lmdData;
        return e;
    }
}
 
 
void PtexReader::readMetaData()
{
    // get read lock and make sure we still need to read
    AutoMutex locker(readlock);
    if (_metadata) {
        return;
    }
 
    // allocate new meta data (keep local until fully initialized)
    MetaData* newmeta = new MetaData(this);
    size_t metaDataMemUsed = sizeof(MetaData);
 
    // read primary meta data blocks
    if (_header.metadatamemsize)
        readMetaDataBlock(newmeta, _metadatapos,
                          _header.metadatazipsize, _header.metadatamemsize, metaDataMemUsed);
 
    // read large meta data headers
    if (_extheader.lmdheadermemsize)
        readLargeMetaDataHeaders(newmeta, _lmdheaderpos,
                                 _extheader.lmdheaderzipsize, _extheader.lmdheadermemsize, metaDataMemUsed);
 
    // store meta data
    AtomicStore(&_metadata, newmeta);
    increaseMemUsed(newmeta->selfDataSize() + metaDataMemUsed);
}
 
 
void PtexReader::readMetaDataBlock(MetaData* metadata, FilePos pos, int zipsize, int memsize, size_t& metaDataMemUsed)
{
    seek(pos);
    // read from file
    bool useNew = memsize > AllocaMax;
    char* buff = useNew ? new char[memsize] : (char*)alloca(memsize);
 
    if (readZipBlock(buff, zipsize, memsize)) {
        // unpack data entries
        char* ptr = buff;
        char* end = ptr + memsize;
        while (ptr < end) {
            uint8_t keysize = *ptr++;
            char* key = (char*)ptr; ptr += keysize;
            key[keysize-1] = '\0';
            uint8_t datatypeval = *ptr++;
            uint32_t datasize; memcpy(&datasize, ptr, sizeof(datasize));
            ptr += sizeof(datasize);
            char* data = ptr; ptr += datasize;
            metadata->addEntry((uint8_t)(keysize-1), key, datatypeval, datasize, data, metaDataMemUsed);
        }
    }
    if (useNew) delete [] buff;
}
 
 
void PtexReader::readLargeMetaDataHeaders(MetaData* metadata, FilePos pos, int zipsize, int memsize, size_t& metaDataMemUsed)
{
    seek(pos);
    // read from file
    bool useNew = memsize > AllocaMax;
    char* buff = useNew ? new char [memsize] : (char*)alloca(memsize);
 
    if (readZipBlock(buff, zipsize, memsize)) {
        pos += zipsize;
 
        // unpack data entries
        char* ptr = buff;
        char* end = ptr + memsize;
        while (ptr < end) {
            uint8_t keysize = *ptr++;
            char* key = (char*)ptr; ptr += keysize;
            uint8_t datatypeval = *ptr++;
            uint32_t datasize; memcpy(&datasize, ptr, sizeof(datasize));
            ptr += sizeof(datasize);
            uint32_t zipsizeval; memcpy(&zipsizeval, ptr, sizeof(zipsizeval));
            ptr += sizeof(zipsizeval);
            metadata->addLmdEntry((uint8_t)(keysize-1), key, datatypeval, datasize, pos, zipsizeval, metaDataMemUsed);
            pos += zipsizeval;
        }
    }
    if (useNew) delete [] buff;
}
 
 
bool PtexReader::readBlock(void* data, int size)
{
    assert(_fp && size >= 0);
    if (!_fp || size < 0) return false;
    int result = (int)_io->read(data, size, _fp);
    if (result == size) {
        _pos += size;
        return true;
    }
    setIOError("PtexReader error: read failed");
    return false;
}
 
 
bool PtexReader::readZipBlock(void* data, int zipsize, int unzipsize)
{
    if (!_ok || zipsize < 0 || unzipsize < 0) return false;
    std::vector<std::byte> compressedBuffer(zipsize);
    if (!readBlock(compressedBuffer.data(), compressedBuffer.size())) {
        return false;
    }
    size_t bytesDecompressed{0};
    if (libdeflate_zlib_decompress(_decompressor, compressedBuffer.data(), compressedBuffer.size(),
                                   data, unzipsize, &bytesDecompressed) != 0 ||
        bytesDecompressed != size_t(unzipsize))
    {
        setError("PtexReader error: unzip failed, file corrupt");
        return false;
    }
    return true;
}
 
 
void PtexReader::readLevel(int levelid, Level*& level)
{
    // get read lock and make sure we still need to read
    AutoMutex locker(readlock);
    if (level) {
        return;
    }
 
    // go ahead and read the level
    LevelInfo& l = _levelinfo[levelid];
 
    // keep new level local until finished
    Level* newlevel = new Level(l.nfaces);
 
    // read level header
    seek(_levelpos[levelid]);
    readZipBlock(&newlevel->fdh[0], l.levelheadersize, FaceDataHeaderSize * l.nfaces);
 
    // compute face offsets
    std::vector<uint32_t> largeFaces;
    FilePos offset = tell();
    for (uint32_t f = 0; f < l.nfaces; f++) {
        newlevel->offsets[f] = offset;
        if (!newlevel->fdh[f].isLargeFace()) {
            offset += newlevel->fdh[f].blocksize();
        } else {
            // large faces have a 64-bit size, stored after the level header
            largeFaces.push_back(f);
        }
    }
 
    // update offsets to account for large faces
    if (!largeFaces.empty()) {
        int nlarge = int(largeFaces.size());
        // read large face header (64-bit sizes of large faces)
        std::vector<size_t> largeFaceHeader(nlarge);
        size_t largeFaceHeaderSize = sizeof(size_t) * nlarge;
        readBlock(largeFaceHeader.data(), largeFaceHeaderSize);
 
        // update offsets
        size_t extraOffset = largeFaceHeaderSize;
        uint32_t f = 0;
        for (int i = 0; i < nlarge; i++) {
            uint32_t lf = largeFaces[i];
            while (f <= lf) {
                newlevel->offsets[f++] += extraOffset;
            }
            extraOffset += largeFaceHeader[i];
        }
        while (f < l.nfaces) {
            newlevel->offsets[f++] += extraOffset;
        }
    }
 
    // don't assign to result until level data is fully initialized
    AtomicStore(&level, newlevel);
    increaseMemUsed(level->memUsed());
}
 
 
void PtexReader::readFace(int levelid, Level* level, int faceid, Ptex::Res res)
{
    FaceData*& face = level->faces[faceid];
    FaceDataHeader fdh = level->fdh[faceid];
    readFaceData(level->offsets[faceid], fdh, res, levelid, face);
}
 
 
void PtexReader::TiledFace::readTile(int tile, FaceData*& data)
{
    _reader->readFaceData(_offsets[tile], _fdh[tile], _tileres, _levelid, data);
}
 
 
void PtexReader::readFaceData(FilePos pos, FaceDataHeader fdh, Res res, int levelid,
                              FaceData*& face)
{
    AutoMutex locker(readlock);
    if (face) {
        return;
    }
 
    // keep new face local until fully initialized
    FaceData* newface = 0;
    size_t newMemUsed = 0;
 
    seek(pos);
    switch (fdh.encoding()) {
    case enc_constant:
        {
            ConstantFace* cf = new ConstantFace(_pixelsize);
            newface = cf;
            newMemUsed = sizeof(ConstantFace) + _pixelsize;
            readBlock(cf->data(), _pixelsize);
            if (levelid==0 && _premultiply && _header.hasAlpha())
                PtexUtils::multalpha(cf->data(), 1, datatype(),
                                     _header.nchannels, _header.alphachan);
        }
        break;
    case enc_tiled:
        {
            Res tileres;
            readBlock(&tileres, sizeof(tileres));
            uint32_t tileheadersize;
            readBlock(&tileheadersize, sizeof(tileheadersize));
            TiledFace* tf = new TiledFace(this, res, tileres, levelid);
            newface = tf;
            newMemUsed = tf->memUsed();
            readZipBlock(&tf->_fdh[0], tileheadersize, FaceDataHeaderSize * tf->_ntiles);
            computeFaceTileOffsets(tell(), tf->_ntiles, &tf->_fdh[0], &tf->_offsets[0]);
        }
        break;
    case enc_zipped:
    case enc_diffzipped:
        {
            int uw = res.u(), vw = res.v();
            int npixels = uw * vw;
            int unpackedSize = _pixelsize * npixels;
            PackedFace* pf = new PackedFace(res, _pixelsize, unpackedSize);
            newface = pf;
            newMemUsed = sizeof(PackedFace) + unpackedSize;
            bool useNew = unpackedSize > AllocaMax;
            char* tmp = useNew ? new char [unpackedSize] : (char*) alloca(unpackedSize);
            readZipBlock(tmp, fdh.blocksize(), unpackedSize);
            if (fdh.encoding() == enc_diffzipped)
                PtexUtils::decodeDifference(tmp, unpackedSize, datatype());
            PtexUtils::interleave(tmp, uw * DataSize(datatype()), uw, vw,
                                  pf->data(), uw * _pixelsize,
                                  datatype(), _header.nchannels);
            if (levelid==0 && _premultiply && _header.hasAlpha())
                PtexUtils::multalpha(pf->data(), npixels, datatype(),
                                     _header.nchannels, _header.alphachan);
            if (useNew) delete [] tmp;
        }
        break;
    }
 
    if (!newface) newface = errorData();
 
    AtomicStore(&face, newface);
    increaseMemUsed(newMemUsed);
}
 
 
void PtexReader::getData(int faceid, void* buffer, int stride)
{
    const FaceInfo& f = getFaceInfo(faceid);
    getData(faceid, buffer, stride, f.res);
}
 
 
void PtexReader::getData(int faceid, void* buffer, int stride, Res res)
{
    if (!_ok || faceid < 0 || size_t(faceid) >= _header.nfaces) {
        PtexUtils::fill(&_errorPixel[0], buffer, stride, res.u(), res.v(), _pixelsize);
        return;
    }
 
    // note - all locking is handled in called getData methods
    int resu = res.u(), resv = res.v();
    int rowlen = _pixelsize * resu;
    if (stride == 0) stride = rowlen;
 
    PtexPtr<PtexFaceData> d ( getData(faceid, res) );
    if (d->isConstant()) {
        // fill dest buffer with pixel value
        PtexUtils::fill(d->getData(), buffer, stride,
                        resu, resv, _pixelsize);
    }
    else if (d->isTiled()) {
        // loop over tiles
        Res tileres = d->tileRes();
        int ntilesu = res.ntilesu(tileres);
        int ntilesv = res.ntilesv(tileres);
        int tileures = tileres.u();
        int tilevres = tileres.v();
        int tilerowlen = _pixelsize * tileures;
        int tile = 0;
        char* dsttilerow = (char*) buffer;
        for (int i = 0; i < ntilesv; i++) {
            char* dsttile = dsttilerow;
            for (int j = 0; j < ntilesu; j++) {
                PtexPtr<PtexFaceData> t ( d->getTile(tile++) );
                if (t->isConstant())
                    PtexUtils::fill(t->getData(), dsttile, stride,
                                    tileures, tilevres, _pixelsize);
                else
                    PtexUtils::copy(t->getData(), tilerowlen, dsttile, stride,
                                    tilevres, tilerowlen);
                dsttile += tilerowlen;
            }
            dsttilerow += stride * tilevres;
        }
    }
    else {
        PtexUtils::copy(d->getData(), rowlen, buffer, stride, resv, rowlen);
    }
}
 
 
PtexFaceData* PtexReader::getData(int faceid)
{
    if (!_ok || faceid < 0 || size_t(faceid) >= _header.nfaces) {
        return errorData(/*deleteOnRelease*/ true);
    }
 
    FaceInfo& fi = _faceinfo[faceid];
    if (fi.isConstant() || fi.res == 0) {
        return new ConstDataPtr(getConstantData(faceid), _pixelsize);
    }
 
    // get level zero (full) res face
    Level* level = getLevel(0);
    FaceData* face = getFace(0, level, faceid, fi.res);
    return face;
}
 
 
PtexFaceData* PtexReader::getData(int faceid, Res res)
{
    if (!_ok || faceid < 0 || size_t(faceid) >= _header.nfaces) {
        return errorData(/*deleteOnRelease*/ true);
    }
 
    FaceInfo& fi = _faceinfo[faceid];
    if (fi.isConstant() || res == 0) {
        return new ConstDataPtr(getConstantData(faceid), _pixelsize);
    }
 
    // determine how many reduction levels are needed
    int redu = fi.res.ulog2 - res.ulog2, redv = fi.res.vlog2 - res.vlog2;
 
    if (redu == 0 && redv == 0) {
        // no reduction - get level zero (full) res face
        Level* level = getLevel(0);
        FaceData* face = getFace(0, level, faceid, res);
        return face;
    }
 
    if (redu == redv) {
        // reduction is symmetric and non-negative
        // => access data from reduction level (if present)
        int levelid = redu;
        if (size_t(levelid) < _levels.size()) {
            Level* level = getLevel(levelid);
 
            // get reduction face id
            int rfaceid = _rfaceids[faceid];
 
            // get the face data (if present)
            FaceData* face = 0;
            if (size_t(rfaceid) < level->faces.size()) {
                face = getFace(levelid, level, rfaceid, res);
            }
            if (face) {
                return face;
            }
        }
    }
 
    // dynamic reduction required - look in dynamic reduction cache
    ReductionKey key(faceid, res);
    FaceData* face = _reductions.get(key);
    if (face) {
        return face;
    }
 
    // not found,  generate new reduction
    FaceData *newface = 0;
    size_t newMemUsed = 0;
 
    if (res.ulog2 < 0 || res.vlog2 < 0) {
        std::cerr << "PtexReader::getData - reductions below 1 pixel not supported" << std::endl;
        newface = errorData();
    }
    else if (redu < 0 || redv < 0) {
        std::cerr << "PtexReader::getData - enlargements not supported" << std::endl;
        newface = errorData();
    }
    else if (_header.meshtype == mt_triangle)
    {
        if (redu != redv) {
            std::cerr << "PtexReader::getData - anisotropic reductions not supported for triangle mesh" << std::endl;
            newface = errorData();
        }
        else {
            PtexPtr<PtexFaceData> psrc ( getData(faceid, Res((int8_t)(res.ulog2+1), (int8_t)(res.vlog2+1))) );
            FaceData* src = static_cast<FaceData*>(psrc.get());
            newface = src->reduce(this, res, PtexUtils::reduceTri, newMemUsed);
        }
    }
    else {
        // determine which direction to blend
        bool blendu;
        if (redu == redv) {
            // for symmetric face blends, alternate u and v blending
            blendu = (res.ulog2 & 1);
        }
        else blendu = redu > redv;
 
        if (blendu) {
            // get next-higher u-res and reduce in u
            PtexPtr<PtexFaceData> psrc ( getData(faceid, Res((int8_t)(res.ulog2+1), (int8_t)res.vlog2)) );
            FaceData* src = static_cast<FaceData*>(psrc.get());
            newface = src->reduce(this, res, PtexUtils::reduceu, newMemUsed);
        }
        else {
            // get next-higher v-res and reduce in v
            PtexPtr<PtexFaceData> psrc ( getData(faceid, Res((int8_t)res.ulog2, (int8_t)(res.vlog2+1))) );
            FaceData* src = static_cast<FaceData*>(psrc.get());
            newface = src->reduce(this, res, PtexUtils::reducev, newMemUsed);
        }
    }
 
    size_t tableNewMemUsed = 0;
    face = _reductions.tryInsert(key, newface, tableNewMemUsed);
    if (face != newface) {
        delete newface;
    }
    else {
        increaseMemUsed(newMemUsed + tableNewMemUsed);
    }
    return face;
}
 
 
void PtexReader::getPixel(int faceid, int u, int v,
                          float* result, int firstchan, int nchannelsArg)
{
    memset(result, 0, sizeof(*result)*nchannelsArg);
 
    // clip nchannels against actual number available
    nchannelsArg = std::min(nchannelsArg, _header.nchannels-firstchan);
    if (nchannelsArg <= 0) return;
 
    // get raw pixel data
    void* pixel;
    if (faceid >= 0 && size_t(faceid) < _header.nfaces && _faceinfo[faceid].isConstant()) {
        pixel = getConstantData(faceid);
    }
    else {
        PtexPtr<PtexFaceData> data ( getData(faceid) );
        pixel = alloca(_pixelsize);
        data->getPixel(u, v, pixel);
    }
 
    // adjust for firstchan offset
    int datasize = DataSize(datatype());
    if (firstchan)
        pixel = (char*) pixel + datasize * firstchan;
 
    // convert/copy to result as needed
    if (datatype() == dt_float)
        memcpy(result, pixel, datasize * nchannelsArg);
    else
        ConvertToFloat(result, pixel, datatype(), nchannelsArg);
}
 
 
void PtexReader::getPixel(int faceid, int u, int v,
                          float* result, int firstchan, int nchannelsArg,
                          Ptex::Res res)
{
    memset(result, 0, sizeof(*result)*nchannelsArg);
 
    // clip nchannels against actual number available
    nchannelsArg = std::min(nchannelsArg, _header.nchannels-firstchan);
    if (nchannelsArg <= 0) return;
 
    // get raw pixel data
    void* pixel;
    if (faceid >= 0 && size_t(faceid) < _header.nfaces && _faceinfo[faceid].isConstant()) {
        pixel = getConstantData(faceid);
    }
    else {
        PtexPtr<PtexFaceData> data ( getData(faceid, res) );
        pixel = alloca(_pixelsize);
        data->getPixel(u, v, pixel);
    }
 
    // adjust for firstchan offset
    int datasize = DataSize(datatype());
    if (firstchan)
        pixel = (char*) pixel + datasize * firstchan;
 
    // convert/copy to result as needed
    if (datatype() == dt_float)
        memcpy(result, pixel, datasize * nchannelsArg);
    else
        ConvertToFloat(result, pixel, datatype(), nchannelsArg);
}
 
 
void PtexReader::getCompressedData(int faceid, int levelid, FaceDataHeader& fdh, std::vector<std::byte>& data)
{
    if (!_ok || faceid < 0 || size_t(faceid) >= _header.nfaces || size_t(levelid) >= _levels.size()) {
        return;
    }
 
    Level* level = getLevel(levelid);
    int rfaceid = levelid == 0 ? faceid : _rfaceids[faceid];
    fdh = level->fdh[rfaceid];
    FilePos offset = level->offsets[rfaceid];
    size_t size{0};
    if (!fdh.isLargeFace()) {
        size = fdh.blocksize();
    } else if (fdh.encoding() == enc_tiled) {
        // read tiled face header
        seek(offset);
        FaceInfo& fi = _faceinfo[faceid];
        Res level_res(fi.res.ulog2 - levelid, fi.res.vlog2 - levelid);
        Res tileres;
        readBlock(&tileres, sizeof(tileres));
        uint32_t tileheadersize;
        readBlock(&tileheadersize, sizeof(tileheadersize));
        int ntiles = level_res.ntiles(tileres);
        std::vector<FaceDataHeader> tiledFace_fdh(ntiles);
        readZipBlock(&tiledFace_fdh[0], tileheadersize, FaceDataHeaderSize * ntiles);
 
        // sum total size of header and tiles
        size = sizeof(tileres) + sizeof(tileheadersize) + tileheadersize;
        for (auto fdh : tiledFace_fdh) {
            size += fdh.blocksize();
        }
    }
 
    // read compressed face data
    seek(offset);
    data.resize(size);
    readBlock(data.data(), size);
}
 
 
PtexReader::FaceData*
PtexReader::PackedFace::reduce(PtexReader* r, Res newres, PtexUtils::ReduceFn reducefn,
                               size_t& newMemUsed)
{
    // allocate a new face and reduce image
    DataType dt = r->datatype();
    int nchan = r->nchannels();
    int memsize = _pixelsize * newres.size64();
    PackedFace* pf = new PackedFace(newres, _pixelsize, memsize);
    newMemUsed = sizeof(PackedFace) + memsize;
    // reduce and copy into new face
    reducefn(_data, _pixelsize * _res.u(), _res.u(), _res.v(),
             pf->_data, _pixelsize * newres.u(), dt, nchan);
    return pf;
}
 
 
 
PtexReader::FaceData* PtexReader::ConstantFace::reduce(PtexReader*, Res, PtexUtils::ReduceFn, size_t& newMemUsed)
{
    // must make a new constant face (even though it's identical to this one)
    // because it will be owned by a different reduction level
    // and will therefore have a different parent
    ConstantFace* pf = new ConstantFace(_pixelsize);
    newMemUsed = sizeof(ConstantFace) + _pixelsize;
    memcpy(pf->_data, _data, _pixelsize);
    return pf;
}
 
 
PtexReader::FaceData*
PtexReader::TiledFaceBase::reduce(PtexReader* r, Res newres, PtexUtils::ReduceFn reducefn,
                                  size_t& newMemUsed)
{
    /* Tiled reductions should generally only be anisotropic (just u
       or v, not both) since isotropic reductions are precomputed and
       stored on disk.  (This function should still work for isotropic
       reductions though.)
 
       In the anisotropic case, the number of tiles should be kept the
       same along the direction not being reduced in order to preserve
       the laziness of the file access.  In contrast, if reductions
       were not tiled, then any reduction would read all the tiles and
       defeat the purpose of tiling.
    */
 
    // keep new face local until fully initialized
    FaceData* newface = 0;
 
    // don't tile triangle reductions (too complicated)
    Res newtileres;
    bool isTriangle = r->_header.meshtype == mt_triangle;
    if (isTriangle) {
        newtileres = newres;
    }
    else {
        // propagate the tile res to the reduction
        newtileres = _tileres;
        // but make sure tile isn't larger than the new face!
        if (newtileres.ulog2 > newres.ulog2) newtileres.ulog2 = newres.ulog2;
        if (newtileres.vlog2 > newres.vlog2) newtileres.vlog2 = newres.vlog2;
    }
 
 
    // determine how many tiles we will have on the reduction
    int newntiles = newres.ntiles(newtileres);
 
    if (newntiles == 1) {
        // no need to keep tiling, reduce tiles into a single face
        // first, get all tiles and check if they are constant (with the same value)
        PtexFaceData** tiles = (PtexFaceData**) alloca(_ntiles * sizeof(PtexFaceData*));
        bool allConstant = true;
        for (int i = 0; i < _ntiles; i++) {
            PtexFaceData* tile = tiles[i] = getTile(i);
            allConstant = (allConstant && tile->isConstant() &&
                           (i == 0 || (0 == memcmp(tiles[0]->getData(), tile->getData(),
                                                   _pixelsize))));
        }
        if (allConstant) {
            // allocate a new constant face
            newface = new ConstantFace(_pixelsize);
            memcpy(newface->getData(), tiles[0]->getData(), _pixelsize);
            newMemUsed = sizeof(ConstantFace) + _pixelsize;
        }
        else if (isTriangle) {
            // reassemble all tiles into temporary contiguous image
            // (triangle reduction doesn't work on tiles)
            int tileures = _tileres.u();
            int tilevres = _tileres.v();
            int sstride = _pixelsize * tileures;
            int dstride = sstride * _ntilesu;
            int dstepv = dstride * tilevres - sstride*(_ntilesu-1);
 
            char* tmp = new char [_ntiles * _tileres.size64() * _pixelsize];
            char* tmpptr = tmp;
            for (int i = 0; i < _ntiles;) {
                PtexFaceData* tile = tiles[i];
                if (tile->isConstant())
                    PtexUtils::fill(tile->getData(), tmpptr, dstride,
                                    tileures, tilevres, _pixelsize);
                else
                    PtexUtils::copy(tile->getData(), sstride, tmpptr, dstride, tilevres, sstride);
                i++;
                tmpptr += (i%_ntilesu) ? sstride : dstepv;
            }
 
            // allocate a new packed face
            int memsize = _pixelsize * newres.size64();
            newface = new PackedFace(newres, _pixelsize, memsize);
            newMemUsed = sizeof(PackedFace) + memsize;
            // reduce and copy into new face
            reducefn(tmp, _pixelsize * _res.u(), _res.u(), _res.v(),
                     newface->getData(), _pixelsize * newres.u(), _dt, _nchan);
 
            delete [] tmp;
        }
        else {
            // allocate a new packed face
            int memsize = _pixelsize * newres.size64();
            newface = new PackedFace(newres, _pixelsize, memsize);
            newMemUsed = sizeof(PackedFace) + memsize;
 
            int tileures = _tileres.u();
            int tilevres = _tileres.v();
            int sstride = _pixelsize * tileures;
            int dstride = _pixelsize * newres.u();
            int dstepu = dstride/_ntilesu;
            int dstepv = dstride*newres.v()/_ntilesv - dstepu*(_ntilesu-1);
 
            char* dst = (char*) newface->getData();
            for (int i = 0; i < _ntiles;) {
                PtexFaceData* tile = tiles[i];
                if (tile->isConstant())
                    PtexUtils::fill(tile->getData(), dst, dstride,
                                    newres.u()/_ntilesu, newres.v()/_ntilesv,
                                    _pixelsize);
                else
                    reducefn(tile->getData(), sstride, tileures, tilevres,
                             dst, dstride, _dt, _nchan);
                i++;
                dst += (i%_ntilesu) ? dstepu : dstepv;
            }
        }
        // release the tiles
        for (int i = 0; i < _ntiles; i++) tiles[i]->release();
    }
    else {
        // otherwise, tile the reduced face
        TiledReducedFace* tf = new TiledReducedFace(_reader, newres, newtileres, this, reducefn);
        newface = tf;
        newMemUsed = tf->memUsed();
    }
    return newface;
}
 
 
void PtexReader::TiledFaceBase::getPixel(int ui, int vi, void* result)
{
    int tileu = ui >> _tileres.ulog2;
    int tilev = vi >> _tileres.vlog2;
    PtexPtr<PtexFaceData> tile ( getTile(tilev * _ntilesu + tileu) );
    tile->getPixel(ui - (tileu<<_tileres.ulog2),
                   vi - (tilev<<_tileres.vlog2), result);
}
 
 
 
PtexFaceData* PtexReader::TiledReducedFace::getTile(int tile)
{
    FaceData*& face = _tiles[tile];
    if (face) {
        return face;
    }
 
    // first, get all parent tiles for this tile
    // and check if they are constant (with the same value)
    int pntilesu = _parentface->ntilesu();
    int pntilesv = _parentface->ntilesv();
    int nu = pntilesu / _ntilesu; // num parent tiles for this tile in u dir
    int nv = pntilesv / _ntilesv; // num parent tiles for this tile in v dir
 
    int ntilesval = nu*nv; // num parent tiles for this tile
    PtexFaceData** tiles = (PtexFaceData**) alloca(ntilesval * sizeof(PtexFaceData*));
    bool allConstant = true;
    int ptile = (tile/_ntilesu) * nv * pntilesu + (tile%_ntilesu) * nu;
    for (int i = 0; i < ntilesval;) {
        PtexFaceData* tileval = tiles[i] = _parentface->getTile(ptile);
        allConstant = (allConstant && tileval->isConstant() &&
                       (i==0 || (0 == memcmp(tiles[0]->getData(), tileval->getData(),
                                             _pixelsize))));
        i++;
        ptile += (i%nu)? 1 : pntilesu - nu + 1;
    }
 
    FaceData* newface = 0;
    size_t newMemUsed = 0;
    if (allConstant) {
        // allocate a new constant face
        newface = new ConstantFace(_pixelsize);
        newMemUsed = sizeof(ConstantFace) + _pixelsize;
        memcpy(newface->getData(), tiles[0]->getData(), _pixelsize);
    }
    else {
        // allocate a new packed face for the tile
        int memsize = _pixelsize*_tileres.size64();
        newface = new PackedFace(_tileres, _pixelsize, memsize);
        newMemUsed = sizeof(PackedFace) + memsize;
 
        // generate reduction from parent tiles
        int ptileures = _parentface->tileres().u();
        int ptilevres = _parentface->tileres().v();
        int sstride = ptileures * _pixelsize;
        int dstride = _tileres.u() * _pixelsize;
        int dstepu = dstride/nu;
        int dstepv = dstride*_tileres.v()/nv - dstepu*(nu-1);
 
        char* dst = (char*) newface->getData();
        for (int i = 0; i < ntilesval;) {
            PtexFaceData* tileval = tiles[i];
            if (tileval->isConstant())
                PtexUtils::fill(tileval->getData(), dst, dstride,
                                _tileres.u()/nu, _tileres.v()/nv,
                                _pixelsize);
            else
                _reducefn(tileval->getData(), sstride, ptileures, ptilevres,
                          dst, dstride, _dt, _nchan);
            i++;
            dst += (i%nu) ? dstepu : dstepv;
        }
    }
 
    if (!AtomicCompareAndSwap(&face, (FaceData*)0, newface)) {
        delete newface;
    }
    else {
        _reader->increaseMemUsed(newMemUsed);
    }
 
    return face;
}
 
PTEX_NAMESPACE_END