RFR: 8297211: Expensive fillInStackTrace operation in HttpURLConnection.getOutputStream0 when no content-length in response [v2]

[Hannes Greule]

On Mon, 21 Nov 2022 11:17:27 GMT, Jaikiran Pai <jpai at openjdk.org> wrote:

>> Can I please get a review for this small change which addresses https://bugs.openjdk.org/browse/JDK-8297211? 
>> The change checks if the `content-length` header value is present before parsing it into a `Long` value. The commit also now catches the more specific `NumberFormatException` instead of a general `Exception`. Given the context of the code, I believe this change in the exception type in the catch block should be fine.
>> 
>> No new test has been introduced given the nature of this change. I have triggered existing tests to verify no unexpected regressions show up.
>
> Jaikiran Pai has updated the pull request incrementally with one additional commit since the last revision:
> 
>   review suggestion - do a similar change in URLConnection

src/java.base/share/classes/java/net/URLConnection.java line 680:

> 678:             try {
> 679:                 return Date.parse(value);
> 680:             } catch (NumberFormatException e) { }

(not a reviewer) I think NumberFormatException isn't sufficient here. `Date.parse` might throw an `IllegalArgumentException`.

-------------

PR: https://git.openjdk.org/jdk/pull/11258