Read-only mirrors. Development, issues and pull requests are on GitHub.

summaryrefslogtreecommitdiffhomepage
path: root/tests/TestScripts/DiscoverTests/register-tests.cpp
diff options
context:
space:
mode:
authorMartin Hořeňovský2026-07-26 13:15:10 +0200
committerMartin Hořeňovský2026-07-26 19:51:58 +0200
commit1079da4c5f5fea2a31d795bba95eacaa2f622849 (patch)
tree4f708b68ab9038d75a50e15467b3ea8182e939f1 /tests/TestScripts/DiscoverTests/register-tests.cpp
parent60c8b878292ec4afc625a807cdc3902397a408a3 (diff)
downloadcatch3-1079da4c5f5fea2a31d795bba95eacaa2f622849.tar.gz
Avoid quadratic JSON array parse behaviour in `catch_discover_tests`
Using CMake's `string(JSON` to parse JSON array leads to quadratic running time in number of tests, see https://gitlab.kitware.com/cmake/cmake/-/work_items/27985 This leads to _terrible_ runtime for `catch_discover_tests` when called on binaries with lot of tests (1k+). To get reasonable runtimes, we have to avoid using `string(JSON` to parse out the individual test objects from the array with all tests. This commit replaces the sane approach of using real JSON parser with a set of terrible hacks, where we use CMake's string APIs to split the JSON array on what looks like object boundary (`}<ws>*,<ws>*{`), and then checking whether the resulting thing can be parsed as JSON object. If not, we append the next piece and check again. And again, and again, until we get a proper JSON object. This is all around a hilariously terrible idea, however: 1) It works in practice for all tested inputs. 2) It improves the time it takes to run `catch_discover_tests` on binary with 1k tests from 4.2s to 1.1s and 2k tests from 16s to 3.9s.
Diffstat (limited to 'tests/TestScripts/DiscoverTests/register-tests.cpp')
-rw-r--r--tests/TestScripts/DiscoverTests/register-tests.cpp21
1 files changed, 21 insertions, 0 deletions
diff --git a/tests/TestScripts/DiscoverTests/register-tests.cpp b/tests/TestScripts/DiscoverTests/register-tests.cpp
index 5aec0f19..b833dba2 100644
--- a/tests/TestScripts/DiscoverTests/register-tests.cpp
+++ b/tests/TestScripts/DiscoverTests/register-tests.cpp
@@ -41,3 +41,24 @@ public:
TEST_CASE_METHOD(TestCaseFixture, "A test case as method", "[tagstagstags]") {}
TEST_CASE( "Newlines\nAnd\rOther\n\tWhitespace", "[whitespace-going-wild]" ) {}
+
+// Some JSON-like and JSON-adjacent characters and substrings in the test names/tags
+// This serves to test that the parse-json-via-string-splitting hack in
+// catch_discover_tests works properly.
+TEST_CASE( "We split on variants of },{ in name" ) {}
+TEST_CASE( "Then }\t, { we }\t,\t{ concatenate } , { them } ,{back},{" ) {}
+TEST_CASE( "}},{{" ) {}
+TEST_CASE( "Arrays [{},{}] wheee", "[also},{tags]" ) {}
+TEST_CASE( "Let's add semicolon into the mix ;},{;},{};,{}" ) {}
+TEST_CASE( "[", "[unmatched-square-bracket]" ) {}
+TEST_CASE( "]", "[unmatched-square-bracket]" ) {}
+
+// Some CMake-like special strings ($ as dereference) strings in test names.
+// This serves to test that the names of test cases are not evaluated
+// inside the catch_discover_tests.
+TEST_CASE( "Plain ${NOT_A_VAR} variable" ) {}
+TEST_CASE( "Env variable access $ENV{HOME}" ) {}
+TEST_CASE( "Cache check $CACHE{FOO}" ) {}
+TEST_CASE( "Also some generator exprs $<CONFIG> in $<1:yes> name" ) {}
+TEST_CASE( "Mess of bare $ $$ $$$ and unterminated $} ${ exprs" ) {}
+TEST_CASE( "$ENV{X};[weird]{},{ mix $<0:no> ${VAR} };,{ }},{{" ) {}