Details
-
Improvement
-
Status: Open
-
Minor
-
Resolution: Unresolved
-
1.0-alpha2
-
None
-
None
Description
Currently getInputStream(start) is marked as final and it reads through the stream to skip bytes.
For specific sources such as File or URL with byte ranges it'd be more efficient to seek straight to the specified offset.
Also it'd be useful to cache the length of a file. Especially if a ByteSourceUrl was implemented.
I'm looking to see if commons-imaging could be improved to support efficient access to https://www.cogeo.org (Cloud optimized GeoTIFF) served from a web server.