Incapsulate Session.getRawInput() to prevent close Socket#InputStream

[valenpo]

All InputStream's that decorate Socket.inputStream could be used in try block safe.
Close of Socket InputStream happens on Session class.

[codecov]

Codecov Report

Attention: Patch coverage is 96.29630% with 1 lines in your changes are missing coverage. Please review.

Project coverage is 71.02%. Comparing base (e4c24f4) to head (8673475).
Report is 19 commits behind head on master.

❗ Current head 8673475 differs from pull request most recent head 87c94b8

Please upload reports for the commit 87c94b8 to get more accurate results.

Files	Patch %	Lines
...subethamail/smtp/internal/command/DataCommand.java	88.88%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master     #113      +/-   ##
============================================
+ Coverage     70.92%   71.02%   +0.09%     
- Complexity      463      466       +3     
============================================
  Files            72       72              
  Lines          2373     2381       +8     
  Branches        248      248              
============================================
+ Hits           1683     1691       +8     
  Misses          572      572              
  Partials        118      118              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[davidmoten]

excuse the delay, what do we gain from this change? Do we have a resource leak? I don't believe we do so I don't see any benefit (reduces false positives from code quality checkers perhaps?).

[valenpo]

excuse the delay, what do we gain from this change? Do we have a resource leak? I don't believe we do so I don't see any benefit (reduces false positives from code quality checkers perhaps?).

we encapsulate the stream in the place where it should be closed. Otherwise, any other user who uses the recommended try catch block will fall into the trap that I fell into.

We are also correcting the TODO, which says that the stream should be closed in the place from where we open it. In the current implementation, it is necessary to keep in mind where it is necessary and where it is not necessary to close the streams, this is not very nice and correct. And an open door for future mistakes.

[valenpo]

@davidmoten Hi. Same way tika incapsulate stream that should not be closed. Any update?

[davidmoten]

Sorry Valentin, working two jobs at the moment and am short on headspace. What I plan to do is look at your suggestion in the IDE and give you some feedback then.

[valenpo]

Thanks!

On 18 Oct 2023, at 10:38, Dave Moten ***@***.***> wrote: Sorry Valentin, working two jobs at the moment and am short on headspace. What I plan to do is look at your suggestion in the IDE and give you some feedback then. — Reply to this email directly, view it on GitHub <#113 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIMLQ77HVSIFFWBOYCDGH3X76BO5AVCNFSM6AAAAAA5QTYA7SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONRXHA2TENZVGM>. You are receiving this because you authored the thread.

Regards, Valentin Popov

[davidmoten]

Can you drop the changes to the pom please (jacoco and guava-mini)? And rename convert to read?

[valenpo]

Done, sorry for delay.

On 18 Oct 2023, at 10:42, Dave Moten ***@***.***> wrote: Can you drop the changes to the pom please (jacoco and guava-mini)? And rename convert to read? — Reply to this email directly, view it on GitHub <#113 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIMLQYBMGQT7IDPIJYL7UTX76B5BAVCNFSM6AAAAAA5QTYA7SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONRXHA3DENZVG4>. You are receiving this because you authored the thread.

Regards, Valentin Popov

[valenpo]

@davidmoten hi. Any chance to look?