Message140311
| Author |
pitrou |
| Recipients |
Jajcus, jcea, pitrou |
| Date |
2011年07月13日.22:01:23 |
| SpamBayes Score |
3.3154304e-08 |
| Marked as misclassified |
No |
| Message-id |
<1310594484.14.0.0389267839547.issue12551@psf.upfronthosting.co.za> |
| In-reply-to |
| Content |
Thank you, this looks mostly good.
A couple of nits:
+#if OPENSSL_VERSION_NUMBER >= 0x0090500fL
+# define HAVE_OPENSSL_FINISHED 1
+#else
+# undef HAVE_OPENSSL_FINNISHED
+#endif
you have a typo in the #undef, also it would be more logical to have
# define HAVE_OPENSSL_FINISHED 0
instead.
_ssl.c will not compile if OpenSSL is too old, because you lack some #if's (or #ifdef's) around PySSL_tls_unique_cb.
Also, it would be nice to expose the availability of tls-unique as a public constant, as we already do for "ssl.HAS_SNI". ssl.HAS_TLS_UNIQUE?
Similarly, you need to skip some of the tests when the functionality isn't available.
And I think get_channel_binding() should raise NotImplementedError in that case. |
|
History
|
|---|
| Date |
User |
Action |
Args |
| 2011年07月13日 22:01:24 | pitrou | set | recipients:
+ pitrou, jcea, Jajcus |
| 2011年07月13日 22:01:24 | pitrou | set | messageid: <1310594484.14.0.0389267839547.issue12551@psf.upfronthosting.co.za> |
| 2011年07月13日 22:01:23 | pitrou | link | issue12551 messages |
| 2011年07月13日 22:01:23 | pitrou | create |
|