deps: update cel-cpp and abseil

[kyessenov]

Signed-off-by: Kuat Yessenov kuat@google.com

Description: updates google/cel-cpp and abseil
Risk Level: low
Testing: backwards compatible
Docs Changes: none
Release Notes: none

[asraa]

Thanks! Looks like there's a legitimate failure, due to how absl now throws when calling front() on an empty string_view here:

envoy/source/common/http/utility.cc

Line 785 in 0957bfe

if (potential_ip_address.front() == '[' && potential_ip_address.back() == ']') {

But also we only passed in an empty string because there are 15 patterns but only 14 specified.

envoy/test/common/http/utility_test.cc

Line 767 in 0957bfe

std::array<std::tuple<bool, std::string, std::string, absl::optional<uint32_t>>, 15> patterns{

We probably should guard against an empty host in parseAuthority or at least document that the host should be nonempty anyway.

[kyessenov]

Thanks @asraa , nice find. I added a defensive check since changing the pre-condition would require changing all the clients (and some might live outside of this repo). Let me know what you think.

[kyessenov]

Coverage failed in some QUIC test. I don't think this change could possibly cause that.

[asraa]

/retest

[repokitteh-read-only]

🔨 rebuilding ci/circleci: coverage (failed build)

🐱

Caused by: a #10243 (comment) was created by @asraa.

see: more, trace.

[kyessenov]

/retest

[repokitteh-read-only]

🔨 rebuilding ci/circleci: coverage (failed build)

🐱

Caused by: a #10243 (comment) was created by @kyessenov.

see: more, trace.

[kyessenov]

@lizan Coverage is again acting up. Let me know if you have any ideas how to resolve it this time.

[mattklein123]

Please merge master and it will probably be (mostly) fixed.