private void decodeMetaData() throws IOException {
// System.out.println("BTR.decodeMetadata mdUpto=" + metaDataUpto + " vs termCount=" +
// state.termBlockOrd + " state=" + state);
if (!seekPending) {
// TODO: cutover to random-access API
// here.... really stupid that we have to decode N
// wasted term metadata just to get to the N+1th
// that we really need...
// lazily catch up on metadata decode:
final int limit = state.termBlockOrd;
boolean absolute = metaDataUpto == 0;
// TODO: better API would be "jump straight to term=N"???
while (metaDataUpto < limit) {
// System.out.println(" decode mdUpto=" + metaDataUpto);
// TODO: we could make "tiers" of metadata, ie,
// decode docFreq/totalTF but don't decode postings
// metadata; this way caller could get
// docFreq/totalTF w/o paying decode cost for
// postings
// TODO: if docFreq were bulk decoded we could
// just skipN here:
// docFreq, totalTermFreq
state.docFreq = freqReader.readVInt();
// System.out.println(" dF=" + state.docFreq);
if (fieldInfo.getIndexOptions() != IndexOptions.DOCS) {
state.totalTermFreq = state.docFreq + freqReader.readVLong();
// System.out.println(" totTF=" + state.totalTermFreq);
}
// metadata
for (int i = 0; i < longs.length; i++) {
longs[i] = bytesReader.readVLong();
}
postingsReader.decodeTerm(longs, bytesReader, fieldInfo, state, absolute);
metaDataUpto++;
absolute = false;
}
} else {
// System.out.println(" skip! seekPending");
}
}
// Pushes a frame we seek'd to
IDVersionSegmentTermsEnumFrame pushFrame(
FST.Arc<Pair<BytesRef, Long>> arc, Pair<BytesRef, Long> frameData, int length)
throws IOException {
scratchReader.reset(
frameData.output1.bytes, frameData.output1.offset, frameData.output1.length);
final long code = scratchReader.readVLong();
final long fpSeek = code >>> VersionBlockTreeTermsWriter.OUTPUT_FLAGS_NUM_BITS;
final IDVersionSegmentTermsEnumFrame f = getFrame(1 + currentFrame.ord);
f.maxIDVersion = Long.MAX_VALUE - frameData.output2;
f.hasTerms = (code & VersionBlockTreeTermsWriter.OUTPUT_FLAG_HAS_TERMS) != 0;
f.hasTermsOrig = f.hasTerms;
f.isFloor = (code & VersionBlockTreeTermsWriter.OUTPUT_FLAG_IS_FLOOR) != 0;
if (f.isFloor) {
f.setFloorData(scratchReader, frameData.output1);
}
pushFrame(arc, fpSeek, length);
return f;
}
@SuppressWarnings("unused")
private void printSeekState(PrintStream out) throws IOException {
if (currentFrame == staticFrame) {
out.println(" no prior seek");
} else {
out.println(" prior seek state:");
int ord = 0;
boolean isSeekFrame = true;
while (true) {
IDVersionSegmentTermsEnumFrame f = getFrame(ord);
assert f != null;
final BytesRef prefix = new BytesRef(term.bytes(), 0, f.prefix);
if (f.nextEnt == -1) {
out.println(
" frame "
+ (isSeekFrame ? "(seek)" : "(next)")
+ " ord="
+ ord
+ " fp="
+ f.fp
+ (f.isFloor ? (" (fpOrig=" + f.fpOrig + ")") : "")
+ " prefixLen="
+ f.prefix
+ " prefix="
+ brToString(prefix)
+ (f.nextEnt == -1 ? "" : (" (of " + f.entCount + ")"))
+ " hasTerms="
+ f.hasTerms
+ " isFloor="
+ f.isFloor
+ " code="
+ ((f.fp << VersionBlockTreeTermsWriter.OUTPUT_FLAGS_NUM_BITS)
+ (f.hasTerms ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_HAS_TERMS : 0)
+ (f.isFloor ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_IS_FLOOR : 0))
+ " isLastInFloor="
+ f.isLastInFloor
+ " mdUpto="
+ f.metaDataUpto
+ " tbOrd="
+ f.getTermBlockOrd());
} else {
out.println(
" frame "
+ (isSeekFrame ? "(seek, loaded)" : "(next, loaded)")
+ " ord="
+ ord
+ " fp="
+ f.fp
+ (f.isFloor ? (" (fpOrig=" + f.fpOrig + ")") : "")
+ " prefixLen="
+ f.prefix
+ " prefix="
+ brToString(prefix)
+ " nextEnt="
+ f.nextEnt
+ (f.nextEnt == -1 ? "" : (" (of " + f.entCount + ")"))
+ " hasTerms="
+ f.hasTerms
+ " isFloor="
+ f.isFloor
+ " code="
+ ((f.fp << VersionBlockTreeTermsWriter.OUTPUT_FLAGS_NUM_BITS)
+ (f.hasTerms ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_HAS_TERMS : 0)
+ (f.isFloor ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_IS_FLOOR : 0))
+ " lastSubFP="
+ f.lastSubFP
+ " isLastInFloor="
+ f.isLastInFloor
+ " mdUpto="
+ f.metaDataUpto
+ " tbOrd="
+ f.getTermBlockOrd());
}
if (fr.index != null) {
assert !isSeekFrame || f.arc != null : "isSeekFrame=" + isSeekFrame + " f.arc=" + f.arc;
if (f.prefix > 0 && isSeekFrame && f.arc.label != (term.byteAt(f.prefix - 1) & 0xFF)) {
out.println(
" broken seek state: arc.label="
+ (char) f.arc.label
+ " vs term byte="
+ (char) (term.byteAt(f.prefix - 1) & 0xFF));
throw new RuntimeException("seek state is broken");
}
Pair<BytesRef, Long> output = Util.get(fr.index, prefix);
if (output == null) {
out.println(" broken seek state: prefix is not final in index");
throw new RuntimeException("seek state is broken");
} else if (isSeekFrame && !f.isFloor) {
final ByteArrayDataInput reader =
new ByteArrayDataInput(
output.output1.bytes, output.output1.offset, output.output1.length);
final long codeOrig = reader.readVLong();
final long code =
(f.fp << VersionBlockTreeTermsWriter.OUTPUT_FLAGS_NUM_BITS)
| (f.hasTerms ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_HAS_TERMS : 0)
| (f.isFloor ? VersionBlockTreeTermsWriter.OUTPUT_FLAG_IS_FLOOR : 0);
if (codeOrig != code) {
out.println(
" broken seek state: output code="
+ codeOrig
+ " doesn't match frame code="
+ code);
throw new RuntimeException("seek state is broken");
}
}
}
if (f == currentFrame) {
break;
}
if (f.prefix == validIndexPrefix) {
isSeekFrame = false;
}
ord++;
}
}
}
