feat(tvix/eval): implement builtins.split r/5244

This implements builtins.split, and passes eval-okay-regex-split.nix
(which is moved out of notyetpassing).

Signed-off-by: Adam Joseph <adam@westernsemico.com>
Change-Id: Ieb0975da2058966c697ee0e2f5b3f26ccabfae57
Reviewed-on: https://cl.tvl.fyi/c/depot/+/7143
Tested-by: BuildkiteCI
Reviewed-by: grfn <grfn@gws.fyi>

1 files changed, 48 insertions, 0 deletions

diff --git a/tvix/eval/src/tests/nix_tests/eval-okay-regex-split.nix b/tvix/eval/src/tests/nix_tests/eval-okay-regex-split.nix
new file mode 100644
index 000000000000..0073e057787d
--- /dev/null
+++ b/tvix/eval/src/tests/nix_tests/eval-okay-regex-split.nix
@@ -0,0 +1,48 @@
+with builtins;
+
+# Non capturing regex returns empty lists
+assert  split "foobar" "foobar"  == ["" [] ""];
+assert  split "fo*" "f"          == ["" [] ""];
+assert  split "fo+" "f"          == ["f"];
+assert  split "fo*" "fo"         == ["" [] ""];
+assert  split "fo*" "foo"        == ["" [] ""];
+assert  split "fo+" "foo"        == ["" [] ""];
+assert  split "fo{1,2}" "foo"    == ["" [] ""];
+assert  split "fo{1,2}" "fooo"   == ["" [] "o"];
+assert  split "fo*" "foobar"     == ["" [] "bar"];
+
+# Capturing regex returns a list of sub-matches
+assert  split "(fo*)" "f"        == ["" ["f"] ""];
+assert  split "(fo+)" "f"        == ["f"];
+assert  split "(fo*)" "fo"       == ["" ["fo"] ""];
+assert  split "(f)(o*)" "f"      == ["" ["f" ""] ""];
+assert  split "(f)(o*)" "foo"    == ["" ["f" "oo"] ""];
+assert  split "(fo+)" "foo"      == ["" ["foo"] ""];
+assert  split "(fo{1,2})" "foo"  == ["" ["foo"] ""];
+assert  split "(fo{1,2})" "fooo" == ["" ["foo"] "o"];
+assert  split "(fo*)" "foobar"   == ["" ["foo"] "bar"];
+
+# Matches are greedy.
+assert  split "(o+)" "oooofoooo" == ["" ["oooo"] "f" ["oooo"] ""];
+
+# Matches multiple times.
+assert  split "(b)" "foobarbaz"  == ["foo" ["b"] "ar" ["b"] "az"];
+
+# Split large strings containing newlines. null are inserted when a
+# pattern within the current did not match anything.
+assert  split "[[:space:]]+|([',.!?])" ''
+  Nix Rocks!
+  That's why I use it.
+''  == [
+  "Nix" [ null ] "Rocks" ["!"] "" [ null ]
+  "That" ["'"] "s" [ null ] "why" [ null ] "I" [ null ] "use" [ null ] "it" ["."] "" [ null ]
+  ""
+];
+
+# Documentation examples
+assert  split  "(a)b" "abc"      == [ "" [ "a" ] "c" ];
+assert  split  "([ac])" "abc"    == [ "" [ "a" ] "b" [ "c" ] "" ];
+assert  split  "(a)|(c)" "abc"   == [ "" [ "a" null ] "b" [ null "c" ] "" ];
+assert  split  "([[:upper:]]+)" "  FOO   " == [ "  " [ "FOO" ] "   " ];
+
+true