[
https://issues.apache.org/jira/browse/NIFI-11437?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17711973#comment-17711973
]
ASF subversion and git services commented on NIFI-11437:
--------------------------------------------------------
Commit 61b5c1a7f517abc25b0297d6b13952560777bb59 in nifi's branch
refs/heads/support/nifi-1.x from David Handermann
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=61b5c1a7f5 ]
NIFI-11437 Switched to StreamUtils.fillBuffer() for buffer, Improved
EncryptContentPGP Content Type Detection
This closes #7166
Signed-off-by: Paul Grey <gr...@apache.org>
(cherry picked from commit bc5f00a6671c0f3fd9d6c8599d196f414ac47fd9)
> Improve EncryptContentPGP Content Type Detection
> ------------------------------------------------
>
> Key: NIFI-11437
> URL: https://issues.apache.org/jira/browse/NIFI-11437
> Project: Apache NiFi
> Issue Type: Bug
> Components: Extensions, Security
> Affects Versions: 1.15.0, 1.21.0
> Reporter: David Handermann
> Assignee: David Handermann
> Priority: Major
> Fix For: 1.latest, 2.latest
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The {{EncryptContentPGP}} Processor reads the initial bytes of incoming files
> to determine whether a file is an OpenPGP message. This initial read is
> necessary to support flows with {{SignContentPGP}} creating signed OpenPGP
> messages prior to encryption.
> The current implementation uses an InputStreamCallback with
> ProcessSession.read() to run content type detection. Instead a separate read
> callback, the StreamCallback for ProcessSession.write() can be modified to
> use a buffer with {{{}PushbackInputStream{}}}. This will avoid reading the
> initial bytes multiple times.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
