Skip to content
This repository has been archived by the owner on Jun 14, 2024. It is now read-only.

Commit

Permalink
fix notation used for keywords
Browse files Browse the repository at this point in the history
  • Loading branch information
paryoja committed Nov 4, 2021
1 parent 0c1c022 commit 046f5cc
Showing 1 changed file with 11 additions and 11 deletions.
22 changes: 11 additions & 11 deletions SPEC.md
Original file line number Diff line number Diff line change
Expand Up @@ -46,12 +46,12 @@ possible experience, with relevant helper APIs, documentation and tutorials.

Indexes are managed by `IndexLogEntry` which consists of

* name: Name of the index.
* derivedDataset: Data that has been derived from one or more datasets and may be optionally used by
* `name`: Name of the index.
* `derivedDataset`: Data that has been derived from one or more datasets and may be optionally used by
an arbitrary query optimizer to improve the speed of data retrieval.
* content: File contents used by the index.
* source: Data source.
* properties: Hash map for managing properties of the index.
* `content`: File contents used by the index.
* `source`: Data source.
* `properties`: Hash map for managing properties of the index.

Indexes can have the following states:
* Stable states
Expand Down Expand Up @@ -110,7 +110,7 @@ Available modes:

A user can drop an existing index by using the deleteIndex API and providing the index name.
Index deletion is a soft-delete operation i.e., only the index's status in the Hyperspace metadata from is changed
from "ACTIVE" to "DELETED". This will exclude the deleted index from any future query optimization and Hyperspace
from `ACTIVE` to `DELETED`. This will exclude the deleted index from any future query optimization and Hyperspace
no longer picks that index for any query. However, index files for a deleted index still remain available
(since it is a soft-delete), so if you accidentally deleted the index, you could still restore it.

Expand All @@ -131,20 +131,20 @@ This will bring back the latest version of index into ACTIVE status and makes it
### Cancel

Cancel API to bring back index from an inconsistent state to the last known stable state.
E.g. if index fails during creation, in "CREATING" state.
E.g. if index fails during creation, in `CREATING` state.
The index will not allow any index modifying operations unless a cancel is called.

> Note: Cancel from "VACUUMING" state will move it forward to "DOESNOTEXIST" state.
> Note: Cancel from `VACUUMING` state will move it forward to `DOESNOTEXIST` state.
> Note: If no previous stable state exists, cancel will move it to "DOESNOTEXIST" state.
> Note: If no previous stable state exists, cancel will move it to `DOESNOTEXIST` state.

# Index Type

Hyperspace provides several index types.

* Covering Index
* Roughly speaking, index data for CoveringIndex is just a vertical
* Roughly speaking, index data for `CoveringIndex` is just a vertical
slice of the source data, including only the indexed and included columns,
bucketed and sorted by the indexed columns for efficient access.
* Data Skipping Index
Expand Down Expand Up @@ -182,7 +182,7 @@ automatically if it is applicable.
## Explain

Explains how indexes will be applied to the given dataframe.
Explain API from Hyperspace is very similar to Spark's df.explain API but allows users to compare their original plan
Explain API from Hyperspace is very similar to Spark's `df.explain` API but allows users to compare their original plan
vs the updated index-dependent plan before running their query.
You have an option to choose from html/plaintext/console mode to display the command output.

Expand Down

0 comments on commit 046f5cc

Please sign in to comment.