Merge pull request #492 from biojppm/fix/emit_empty_key

Fix emit of explicit keys
biojppm · Jan 20, 2025 · c7b792d · c7b792d
2 parents d4457b9 + 6291900
commit c7b792d
Show file tree

Hide file tree

Showing 10 changed files with 1,098 additions and 686 deletions.
diff --git a/.github/workflows/infra.yml b/.github/workflows/infra.yml
@@ -18,12 +18,6 @@ env:
   CMAKE_FLAGS: -DRYML_TEST_SUITE=ON
   NUM_JOBS_BUILD: null
 jobs:
-  log-github-event:
-    name: log github event
-    runs-on: ubuntu-24.04
-    steps:
-    - name: log github event
-      run: echo "${{toJSON(github.event)}}" || echo >/dev/null
   check_workflows:
     if: always()
     continue-on-error: false

diff --git a/changelog/current.md b/changelog/current.md
@@ -10,6 +10,16 @@
   - add workarounds for problems with codegen of gcc 11,12,13.
   - improve CI coverage of gcc and clang optimization levels.
 - [BREAKING] Fix [#477](https://github.com/biojppm/rapidyaml/issues/477) ([PR#479](https://github.com/biojppm/rapidyaml/pull/479)): changed `read<std::map>()` to overwrite existing entries. The provided implementations had an inconsistency between `std::map` (which wasn't overwriting) and `std::vector` (which *was* overwriting).
+- [PR#492](https://github.com/biojppm/rapidyaml/pull/492): fix emit of explicit keys when indented:
+  ```yaml
+  fixed:
+      ? explicit key
+      : value
+  previously:
+      ? explicit key
+  : value   # this was not indented
+  ```
+- [PR#492](https://github.com/biojppm/rapidyaml/pull/492):  fix parser reset for full reuse (`m_doc_empty` was not resetted), which would cause problems under specific scenarios in subsequent reuse.
 
 
 ## Thanks

diff --git a/ext/c4core b/ext/c4core
diff --git a/src/c4/yml/emit.def.hpp b/src/c4/yml/emit.def.hpp
@@ -778,7 +778,10 @@ void Emitter<Writer>::_write_scalar_literal(csubstr s, id_type ilevel, bool expl
     for(size_t i = !is_newline_only; i < numnewlines_at_end; ++i)
         this->Writer::_do_write('\n');
     if(explicit_key)
+    {
         this->Writer::_do_write('\n');
+        this->_indent(ilevel);
+    }
 }
 
 template<class Writer>
@@ -856,7 +859,10 @@ void Emitter<Writer>::_write_scalar_folded(csubstr s, id_type ilevel, bool expli
     for(size_t i = !is_newline_only; i < numnewlines_at_end; ++i)
         this->Writer::_do_write('\n');
     if(explicit_key)
+    {
         this->Writer::_do_write('\n');
+        this->_indent(ilevel);
+    }
 }
 
 template<class Writer>

diff --git a/src/c4/yml/parse_engine.def.hpp b/src/c4/yml/parse_engine.def.hpp
@@ -384,6 +384,7 @@ void ParseEngine<EventHandler>::_reset()
 {
     m_pending_anchors = {};
     m_pending_tags = {};
+    m_doc_empty = true;
     if(m_options.locations())
     {
         _prepare_locations();

diff --git a/test/CMakeLists.txt b/test/CMakeLists.txt
@@ -106,7 +106,6 @@ ryml_add_test_case_group(map_generic)
 ryml_add_test_case_group(map_set)
 ryml_add_test_case_group(seq_of_map)
 ryml_add_test_case_group(map_of_seq)
-ryml_add_test_case_group(scalar_empty)
 ryml_add_test_case_group(scalar_null)
 ryml_add_test_case_group(scalar_squoted)
 ryml_add_test_case_group(scalar_dquoted)

diff --git a/test/test_emit.cpp b/test/test_emit.cpp
@@ -359,6 +359,316 @@ TEST(emit, no_node)
     test_emits(t, NONE, expected, expected_json);
 }
 
+TEST(emit, empty_key_squo)
+{
+    {
+        SCOPED_TRACE("one only");
+        const Tree t = parse_in_arena(R"(
+? ''
+: literal
+)");
+        std::string expected = "'': literal\n";
+        std::string expected_json = R"({"": "literal"})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("nested");
+        const Tree t = parse_in_arena(R"(
+level1:
+  ? ''
+  : literal
+)");
+        std::string expected = "level1:\n  '': literal\n";
+        std::string expected_json = R"({"level1": {"": "literal"}})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    const Tree t = parse_in_arena(R"(
+? ''
+: literal
+level1:
+  ? ''
+  : literal
+  level2:
+    ? ''
+    : literal
+    level3:
+      ? ''
+      : literal
+)");
+    {
+        SCOPED_TRACE("level3");
+        std::string expected = "level3:\n  '': literal\n";
+        std::string expected_json = R"("level3": {"": "literal"})";
+        test_emits(t, t["level1"]["level2"]["level3"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level2");
+        std::string expected = "level2:\n  '': literal\n  level3:\n    '': literal\n";
+        std::string expected_json = "\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}";
+        test_emits(t, t["level1"]["level2"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level1");
+        std::string expected = "level1:\n  '': literal\n  level2:\n    '': literal\n    level3:\n      '': literal\n";
+        std::string expected_json = "\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}";
+        test_emits(t, t["level1"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level0");
+        std::string expected = "'': literal\nlevel1:\n  '': literal\n  level2:\n    '': literal\n    level3:\n      '': literal\n";
+        std::string expected_json = "{\"\": \"literal\",\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}}";
+        test_emits(t, NONE, expected, expected_json);
+    }
+}
+
+TEST(emit, empty_key_dquo)
+{
+    {
+        SCOPED_TRACE("one only");
+        const Tree t = parse_in_arena(R"(
+? ""
+: literal
+)");
+        std::string expected = "\"\": literal\n";
+        std::string expected_json = R"({"": "literal"})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("nested");
+        const Tree t = parse_in_arena(R"(
+level1:
+  ? ""
+  : literal
+)");
+        std::string expected = "level1:\n  \"\": literal\n";
+        std::string expected_json = R"({"level1": {"": "literal"}})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    const Tree t = parse_in_arena(R"(
+? ""
+: literal
+level1:
+  ? ""
+  : literal
+  level2:
+    ? ""
+    : literal
+    level3:
+      ? ""
+      : literal
+)");
+    {
+        SCOPED_TRACE("level3");
+        std::string expected = "level3:\n  \"\": literal\n";
+        std::string expected_json = R"("level3": {"": "literal"})";
+        test_emits(t, t["level1"]["level2"]["level3"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level2");
+        std::string expected = "level2:\n  \"\": literal\n  level3:\n    \"\": literal\n";
+        std::string expected_json = "\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}";
+        test_emits(t, t["level1"]["level2"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level1");
+        std::string expected = "level1:\n  \"\": literal\n  level2:\n    \"\": literal\n    level3:\n      \"\": literal\n";
+        std::string expected_json = "\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}";
+        test_emits(t, t["level1"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level0");
+        std::string expected = "\"\": literal\nlevel1:\n  \"\": literal\n  level2:\n    \"\": literal\n    level3:\n      \"\": literal\n";
+        std::string expected_json = "{\"\": \"literal\",\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}}";
+        test_emits(t, NONE, expected, expected_json);
+    }
+}
+
+TEST(emit, empty_key_plain)
+{
+    {
+        SCOPED_TRACE("one only");
+        const Tree t = parse_in_arena(R"(
+?
+: literal
+)");
+        std::string expected = ": literal\n";
+        std::string expected_json = R"({"": "literal"})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("nested");
+        const Tree t = parse_in_arena(R"(
+level1:
+  ?
+  : literal
+)");
+        std::string expected = "level1:\n  : literal\n";
+        std::string expected_json = R"({"level1": {"": "literal"}})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    const Tree t = parse_in_arena(R"(
+?
+: literal
+level1:
+  ?
+  : literal
+  level2:
+    ?
+    : literal
+    level3:
+      ?
+      : literal
+)");
+    {
+        SCOPED_TRACE("level3");
+        std::string expected = "level3:\n  : literal\n";
+        std::string expected_json = R"("level3": {"": "literal"})";
+        test_emits(t, t["level1"]["level2"]["level3"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level2");
+        std::string expected = "level2:\n  : literal\n  level3:\n    : literal\n";
+        std::string expected_json = "\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}";
+        test_emits(t, t["level1"]["level2"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level1");
+        std::string expected = "level1:\n  : literal\n  level2:\n    : literal\n    level3:\n      : literal\n";
+        std::string expected_json = "\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}";
+        test_emits(t, t["level1"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level0");
+        std::string expected = ": literal\nlevel1:\n  : literal\n  level2:\n    : literal\n    level3:\n      : literal\n";
+        std::string expected_json = "{\"\": \"literal\",\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}}";
+        test_emits(t, NONE, expected, expected_json);
+    }
+}
+
+TEST(emit, empty_key_literal)
+{
+    {
+        SCOPED_TRACE("one only");
+        const Tree t = parse_in_arena(R"(
+? |-
+: literal
+)");
+        std::string expected = "? |-\n: literal\n";
+        std::string expected_json = R"({"": "literal"})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("nested");
+        const Tree t = parse_in_arena(R"(
+level1:
+  ? |-
+  : literal
+)");
+        std::string expected = "level1:\n  ? |-\n  : literal\n";
+        std::string expected_json = R"({"level1": {"": "literal"}})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    const Tree t = parse_in_arena(R"(
+? |-
+: literal
+level1:
+  ? |-
+  : literal
+  level2:
+    ? |-
+    : literal
+    level3:
+      ? |-
+      : literal
+)");
+    {
+        SCOPED_TRACE("level3");
+        std::string expected = "level3:\n  ? |-\n  : literal\n";
+        std::string expected_json = R"("level3": {"": "literal"})";
+        test_emits(t, t["level1"]["level2"]["level3"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level2");
+        std::string expected = "level2:\n  ? |-\n  : literal\n  level3:\n    ? |-\n    : literal\n";
+        std::string expected_json = "\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}";
+        test_emits(t, t["level1"]["level2"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level1");
+        std::string expected = "level1:\n  ? |-\n  : literal\n  level2:\n    ? |-\n    : literal\n    level3:\n      ? |-\n      : literal\n";
+        std::string expected_json = "\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}";
+        test_emits(t, t["level1"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level0");
+        std::string expected = "? |-\n: literal\nlevel1:\n  ? |-\n  : literal\n  level2:\n    ? |-\n    : literal\n    level3:\n      ? |-\n      : literal\n";
+        std::string expected_json = "{\"\": \"literal\",\"level1\": {\"\": \"literal\",\"level2\": {\"\": \"literal\",\"level3\": {\"\": \"literal\"}}}}";
+        test_emits(t, NONE, expected, expected_json);
+    }
+}
+
+TEST(emit, empty_key_folded)
+{
+    {
+        SCOPED_TRACE("one only");
+        const Tree t = parse_in_arena(R"(
+? >-
+: folded
+)");
+        std::string expected = "? >-\n: folded\n";
+        std::string expected_json = R"({"": "folded"})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("nested");
+        const Tree t = parse_in_arena(R"(
+level1:
+  ? >-
+  : folded
+)");
+        std::string expected = "level1:\n  ? >-\n  : folded\n";
+        std::string expected_json = R"({"level1": {"": "folded"}})";
+        test_emits(t, t.root_id(), expected, expected_json);
+    }
+    const Tree t = parse_in_arena(R"(
+? >-
+: folded
+level1:
+  ? >-
+  : folded
+  level2:
+    ? >-
+    : folded
+    level3:
+      ? >-
+      : folded
+)");
+    {
+        SCOPED_TRACE("level3");
+        std::string expected = "level3:\n  ? >-\n  : folded\n";
+        std::string expected_json = R"("level3": {"": "folded"})";
+        test_emits(t, t["level1"]["level2"]["level3"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level2");
+        std::string expected = "level2:\n  ? >-\n  : folded\n  level3:\n    ? >-\n    : folded\n";
+        std::string expected_json = "\"level2\": {\"\": \"folded\",\"level3\": {\"\": \"folded\"}}";
+        test_emits(t, t["level1"]["level2"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level1");
+        std::string expected = "level1:\n  ? >-\n  : folded\n  level2:\n    ? >-\n    : folded\n    level3:\n      ? >-\n      : folded\n";
+        std::string expected_json = "\"level1\": {\"\": \"folded\",\"level2\": {\"\": \"folded\",\"level3\": {\"\": \"folded\"}}}";
+        test_emits(t, t["level1"].id(), expected, expected_json);
+    }
+    {
+        SCOPED_TRACE("level0");
+        std::string expected = "? >-\n: folded\nlevel1:\n  ? >-\n  : folded\n  level2:\n    ? >-\n    : folded\n    level3:\n      ? >-\n      : folded\n";
+        std::string expected_json = "{\"\": \"folded\",\"level1\": {\"\": \"folded\",\"level2\": {\"\": \"folded\",\"level3\": {\"\": \"folded\"}}}}";
+        test_emits(t, NONE, expected, expected_json);
+    }
+}
+
 TEST(emit, existing_seq_node_flow)
 {
     Tree nct = parse_in_arena("[foo, bar, [nested, seq], {nested: map}]");
+3 −3		src/c4/charconv.hpp
+0 −2		src/c4/format.hpp
+0 −3		src/c4/substr.hpp
+53 −0		src/c4/utf.cpp
+16 −0		src/c4/utf.hpp
+102 −0		test/test_utf.cpp