From #14497 (comment) (second last paragraph)

(Check Documentation)

Goal of this Issue:

• Conclude whether whitespace between generic and record header should be considered as a violation or not.
• Based on the conclusion necessary changes can be done.

How it works Now:
Currently whitespace between generic and record header is not considered a violation.

Test.java

record Test<T> () {}
//            ^
//            +--- whitespace in question

config.xml

<?xml version="1.0"?>
<!DOCTYPE module PUBLIC
        "-//Checkstyle//DTD Checkstyle Configuration 1.3//EN"
        "https://checkstyle.org/dtds/configuration_1_3.dtd">
<module name="Checker">
  <module name="TreeWalker">
    <module name="GenericWhitespace"/>
  </module>
</module>

Current CLI Execution:

java -jar checkstyle-10.13.0-all.jar -c config.xml Test.java
Starting audit...
Audit done.

Expected Working: To be decided
[ERROR] Test.java:1:14: '>' is followed by whitespace. [GenericWhitespace]

Findings that may be useful:

• The syntax is kind of similar to Issue #14344
• Horizontal Whitespace from google style, which includes the only places where occurrence of whitespace is valid does not includes this usage.
• In oracle docs for record classes [see first point of Features of Record Classes], a whitespace is present between generic and record header in the example.
• This check adheres to the "typical convention".

Typical Usages of Generic Records (Anecdotal):

• Generic records in Checkstyle, default projects for regression ,and those I have encountered so far, do not contain whitespace between generic and record header.
  This can be verified by running the following two grep cmds:
  1. grep -P -r "record [a-zA-Z0-9\s$_]+\<.*\>\(.*\)" dir/ | tee >(wc -l)
  2. grep -P -r "record [a-zA-Z0-9\s$_]+\<.*\>\s+\(.*\)" dir/ | tee >(wc -l)
• The pattern will match for the generic records, first one matching no-whitespace between generic and record header, while the second one matching atleast a single-whitespace between generic and record header.
• Output for the cmd on Checkstyle repo: i. 32; ii: 0
• Output for the cmd on regression repos: i. 60; ii. 0
  Side note: Arguably better way to verify that there are no such cases would be implement changes which detects space between generic and record header as a violation, and then check for differences in the report. (lmk if that is needed)