fix(parquet store gateway): obey query sharding matchers

[npazosmendez]

Inefficient, but at least correct query sharding. The new test on sharding fails on the base branch.

It's not trivial to add caching to the hashes like the main path does, because we don't have a SeriesRef to use as a cache key at the block level (to match what the main path does). We could in theory use something like the row number in the parquet file, but we don't have easy access to that in this part of the code. In any case, the priority right now is correctness, we'll work on optimizing later as appropriate.

For referece, see how query sharding is handled on the main path:

mimir/pkg/storegateway/series_refs.go

Lines 1021 to 1047 in 604775d

	// filterSeries filters out series that don't belong to this shard (if sharding is configured) or that don't have any
	// chunk refs. Empty chunks ranges indicates that the series doesn't have any chunk ranges in the
	// requested time range. filterSeries expects that the number of series matches the number of postings.
	func (s *loadingSeriesChunkRefsSetIterator) filterSeries(set seriesChunkRefsSet, postings []storage.SeriesRef, stats *queryStats) seriesChunkRefsSet {
	writeIdx := 0
	for sIdx, series := range set.series {
	// We skip this series under three conditions:
	// 1. The series doesn't have any chunks in this block OR the series didn't have any chunks in the requested time range,
	// but also the request didn't require the chunks (i.e. s.strategy.isNoChunkRefs()). This is signified by an empty label set.
	if series.lset.IsEmpty() {
	continue
	}
	// 2. The series doesn't have any chunks in the requested time range but the request required the chunks (i.e. !s.strategy.isNoChunkRefs()).
	if !s.strategy.isNoChunkRefs() && len(series.refs) == 0 {
	continue
	}
	// 3. The series doesn't belong to this shard.
	if !shardOwned(s.shard, s.seriesHasher, postings[sIdx], series.lset, stats) {
	continue
	}
	set.series[writeIdx] = set.series[sIdx]
	writeIdx++
	}
	set.series = set.series[:writeIdx]
	return set
	}

[jesusvazquez]

Nice find Nico. I can't find of ways to improve this at the time.