Comparing v0.6.0-rc.0...v0.6.0-rc.1 · Thalhammer/jwt-cpp
Commits on Sep 14, 2021
-
🗺️ Add support for Boost.JSON traits (#165)
* wip boost test * wip hardcoded just to test * wip more * working POC * trying boost json with c++17 * working boost json standalone with cxx17 * Delete conanfile.txt * revert some changes to see if they are required * adding a workflow * cleanup * naming convention * testing for substr on string_type * trying to test for operator+ * fix check for operator+ * losen std::string requirement (experimental) * Update .gitignore Co-authored-by: gerardoprada <gerardoprada@users.noreply.github.com> Co-authored-by: gerardoprada <gerardoprada@users.noreply.github.com>
-
* Renamed variables being shadowed by lambda arguments Co-authored-by: Tobias Kux <XyFreak@users.noreply.github.com> * Renamed lambda parameter name shadowing parameter of parent function Co-authored-by: Tobias Kux <XyFreak@users.noreply.github.com> * Update include/jwt-cpp/jwt.h Co-authored-by: Tobias Kux <XyFreak@users.noreply.github.com> Co-authored-by: Tobias Kux <XyFreak@users.noreply.github.com>
Commits on Sep 17, 2021
-
🧰 Adding clang-tidy, cmake-format, and ASAN tests (#168)
* adding clang-tidy and asan tests Playing with clang-tidy, we already had the rule removed in #166... running the static analyzer from CMake reproduced the report in many places... I suspect the macro and `new`/`delete` separation clang-tidy is just unable to understand. Running the problematic examples + tests with the ASAN enabled reported no problems * Apply suggestions from code review * Update lint.yml * applying linting rules * fix bad rename * formating * clang-tidy on jsoncons tests * and cmake-format to lint * cmake format improve change detection * Update lint.yml * apply linting
Commits on Sep 18, 2021
-
* Update coverage.yml * adding a fuzz test * adding more fuzzing TBA how to handle input requirement of decode * base64 encoding input to avoid "obvious" exceptions trying out the EVP_ interface from #89 * fixing decode fuzz * accepting exceptions are normal After comparing with https://github.com/nlohmann/json/blob/v3.9.0/test/src/fuzzer-parse_json.cpp I must agree data can be random so it should be accepted * decoding twice should produce the same result again based on https://github.com/nlohmann/json/blob/v3.9.0/test/src/fuzzer-parse_json.cpp * fixing token decode fuzzer * adding corpus for fuzz tests + adding them to ci * removing numbers with more meaning descriptions * Update BaseEncodeFuzz.cpp * Update coverage.yml * Update coverage.yml * shrink interations * cleaning cmake * Update and rename coverage.yml to jwt.yml * Update lint.yml * Update jwt.yml * Update jwt.yml
Commits on Sep 22, 2021
-
* outline for wolfssl support lots of missing symbols * rm old libressl dockerfile * ass wolfssl test + dry up workflows with improved composite action having `use` keyword * autoreconf * brute force zap openssl headers * Update action.yml * try to access nested headers * trying to log include headers * more debug logs * fix bad variable * tweaks * more features * cleanup * wolfssl is special like libressl * test header order * revert order * trying wolfssl master * adding wolfssl compatiblity macros trying to get access to https://github.com/wolfSSL/wolfssl/blob/1662b0115766ee80dd7b102d4d08ef160e2c287f/wolfssl/ssl.h#L3960 * trying no newline to fix tests * typo * bad copy paste pt2 * not a new line problem * typo * workaround failing helper test * update cmake test list(TRANSFORM raised min * test cmake install with wolfSSL * add missing EOF * more generic wolfssl install since there is no release that works yet * typo * debug print * add more checks * force HAVE_ECC * install master * hunter does not support wolfssl * pretty job name * clean up * Update README.md * bump to latest master * Update cmake.yml * bump coveralls action * switch badge name to sha * Update jwt.yml they published a bad tag =)
Commits on Nov 17, 2021
-
⬆️ Bump WolfSSL to v5.0.0 (#184)
* Update ssl.yml * Update README.md * Update ssl.yml * Update README.md * update test data to 5.0 impl * remove older abi version the data generated changed so the test code no longer supports the old way something ~4.8 should work