mirdata icon indicating copy to clipboard operation
mirdata copied to clipboard

Option to ignore md5 checksum inconsistencies during download

Open rabitt opened this issue 5 years ago • 2 comments

In some exceptional cases, the md5 checksum of a downloaded file may not match what we expect (e.g. see #494 ), but after decompression, the internal files have the expected checksums.

To avoid this, we could add a allow_invalid_checksum boolean option to download_utils.downloader (which defaults to False), which prints a warning when a checksum is invalid but does not raise an exception, and prompts the user to run validation afterward to ensure the integrity of the downloaded data.

(Thanks @migperfer for pointing out this issue and for the idea!)

rabitt avatar Apr 14 '21 12:04 rabitt

I like this idea!

What if we make cleanup dependent on that boolean, and if we're not checking checksums by default when we download we delete the compressed files?

magdalenafuentes avatar Apr 18 '21 01:04 magdalenafuentes

I guess this one is done then? :)

migperfer avatar Nov 09 '21 12:11 migperfer

solved in #530

nkundiushuti avatar Feb 01 '23 07:02 nkundiushuti