Fix S3 PartialContentInputStream to be compliant to InputStream specification #2217
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
paulcwarren
reviewed
Dec 4, 2024
.../src/test/java/internal/org/springframework/content/s3/io/PartialContentInputStreamTest.java
Outdated
Show resolved
Hide resolved
|
Looks good. I just noticed that the tests are focused which might prevent tests from being run (now or later). Let's unfit and we can merge this. Side note; yes we probably should have something in CI that warns when tests are fitted but I have never gotten around to that. |
vierbergenlars
force-pushed
the
fix-partial-input-stream
branch
from
ac57657 to
06ceaba
Compare
The PartialContentInputStream was a minimal implementation that worked well when used directly with the spring byte-range support. However, using the InputStream variant in any other place (like wrapping it for decryption), the minimal implementation breaks down, requiring workarounds when handling the encryption. Using an implementation that is fully compliant with the InputStream specifications makes implementation of a wrapping InputStream easier, as no special provisions need to be made for a read method that immediately reads the byte-range (instead of empty space before the start)
Because an actual compliant InputStream is now returned from the backing store, encryption can be handled simpler by just running the cipher for the whole range. Since encryption is not authenticated, and CTR mode is byte-addressable, this just works. The zero-bytes before/after the actual data are decrypted as garbage, but they will be cut out when streaming the response back to the user.
Although the testcase works when aligned on a block boundary, the test is more comprehensive when it is not aligned
Skipping decrypting blocks that do not have data and will be skipped anyways is wasted work. Reintroduce the technique to calculate IV for a certain block offset to save decryption work for larger offsets
Using an int limits the offset to about 2GB, which might not be sufficient for very large files. Using a long results in an offset of 9EB, which should be sufficient.
vierbergenlars
force-pushed
the
fix-partial-input-stream
branch
from
8654230 to
fadb30c
Compare
Hello @paulcwarren I believe this issue has been addressed. Is there anything left here that blocks you from merging? kind regards Thijs |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The PartialContentInputStream was a minimal implementation that worked
well when used directly with the spring byte-range support.
However, using the InputStream variant in any other place (like wrapping it for decryption),
the minimal implementation breaks down, requiring workarounds when
handling the encryption.
Using an implementation that is fully compliant with the InputStream
specifications makes implementation of a wrapping InputStream easier,
as no special provisions need to be made for a read method that
immediately reads the byte-range (instead of empty space before the start of the byte-range)