If `advanceExact(docid)` returns false it means that the field does not have a value for the given doc, or it could mean that you attempted to iterate backwards. `advance(docid)` is there to assist with iterating over values in forward order - it's more efficient than callig advanceExact multiple times just to find out there are no values
On Fri, Feb 28, 2020 at 1:42 PM Sergio Bilello <lapostadiser...@gmail.com> wrote: > > Given a docId if I call advanceExact(docId) and if this will return true. > > Can I call > https://github.com/apache/lucene-solr/blob/releases/lucene-solr/8.4.1/lucene/core/src/java/org/apache/lucene/index/NumericDocValues.java#L32-L35 > to retrieve the value? > > If YES I am fine with it if NO what should I do to retrieve the value for the > docId specified in case it exists? > I didn't get the meaning of the comment on top of the method advanceExact. > > Separate question: > What's the purpose to call advance(doc) and eventually get a docId greater > than doc? > > Thanks, > Sergio > > On 2020/02/28 11:04:45, Alan Woodward <romseyg...@gmail.com> wrote: > > You’re dealing with cases where the document in question doesn’t have a > > value in the iterator. So `advance(doc)` will return a docId greater than > > doc, as it moves to the next doc that does have a value, and > > `advanceExact(doc)` returns `false` because you can’t do an exact advance > > to a value that doesn’t exist. > > > > > On 28 Feb 2020, at 01:05, Sergio Bilello <lapostadiser...@gmail.com> > > > wrote: > > > > > > I changed the code to return a default value instead of an exception. In > > > case the doc will not be found. > > > I would like really to understand better why we have such comment in that > > > iterator > > > > > > * Returns the numeric value for the current document ID. * It is illegal > > >> to call this method after {@link #advanceExact(int)} * returned {@code > > >> false}. * @return numeric value* > > > > > > and if I use advance method instead of advanceExact the subsequent call > > > to retrieve the value would be legit? > > > > > > Thanks > > > Sergio > > > On 2020/02/27 22:27:09, Sergio <lapostadiser...@gmail.com> wrote: > > >> Hi guys! > > >> > > >> I am a newbie and I am trying to upgrade from Lucene 6.2.1 to 8.4.1 > > >> The previous code was leveraging this > > >> https://github.com/apache/lucene-solr/blob/releases/lucene-solr/6.2.1/lucene/core/src/java/org/apache/lucene/index/NumericDocValues.java#L34 > > >> > > >> now > > >> > > >> https://github.com/apache/lucene-solr/blob/releases/lucene-solr/8.4.1/lucene/core/src/java/org/apache/lucene/index/NumericDocValues.java#L32-L35 > > >> > > >> Previously there was no need to deal with the iterator and I found in the > > >> code a comment regarding > > >> > > >> > > >> > > >> ** Returns the numeric value for the current document ID. * It is > > >> illegal > > >> to call this method after {@link #advanceExact(int)} * returned {@code > > >> false}. * @return numeric value* > > >> If the below snippet is not correct how should I proceed? > > >> > > >> public static float > > >> getFloatValueFromNumericDocValuesAndDocId(NumericDocValues > > >> numericDocValues, int docId) throws IOException { > > >> if (numericDocValues.advanceExact(docId)) { > > >> return Float.intBitsToFloat((int)numericDocValues.longValue()); > > >> } else { > > >> throw new RuntimeException("It has not been possible to > > >> advanceExact for " + docId) ; > > >> } > > >> } > > >> > > >> public static int > > >> getIntValueFromNumericDocValuesAndDocId(NumericDocValues > > >> numericDocValues, int docId) throws IOException { > > >> if (numericDocValues.advanceExact(docId)) { > > >> return (int) numericDocValues.longValue(); > > >> } else { > > >> throw new RuntimeException("It has not been possible to > > >> advanceExact for " + docId) ; > > >> } > > >> } > > >> > > >> public static long > > >> getLongValueFromNumericDocValuesAndDocId(NumericDocValues > > >> numericDocValues, int docId) throws IOException { > > >> if (numericDocValues.advanceExact(docId)) { > > >> return numericDocValues.longValue(); > > >> } else { > > >> throw new RuntimeException("It has not been possible to > > >> advanceExact for " + docId) ; > > >> } > > >> } > > >> > > >> public static BytesRef > > >> getBytesRefValueFromSortedDocValuesAndDocId(SortedDocValues > > >> numericDocValues, int docId) throws IOException { > > >> if (numericDocValues.advanceExact(docId)) { > > >> return numericDocValues.binaryValue(); > > >> } else { > > >> throw new RuntimeException("It has not been possible to > > >> advanceExact for " + docId) ; > > >> } > > >> } > > >> > > >> > > >> > > >> > > >> Thanks to everyone! > > >> > > >> Sergio > > >> > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org > > > For additional commands, e-mail: dev-h...@lucene.apache.org > > > > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org > > For additional commands, e-mail: dev-h...@lucene.apache.org > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org > For additional commands, e-mail: dev-h...@lucene.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
