Skip to content

fix(ui): emit parseable layout JSON for malformed UTF-8 #2913

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
mvanhorn wants to merge 1 commit into
base: main
Choose a base branch
from
+165 −4
Open

fix(ui): emit parseable layout JSON for malformed UTF-8 #2913

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 8 additions & 4 deletions src/ui/layout3d.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -900,12 +900,16 @@ char *cbm_layout_to_json(const cbm_layout_result_t *r) {
yyjson_mut_obj_add_val(doc, root, "edges", ea);
yyjson_mut_obj_add_int(doc, root, "total_nodes", r->total_nodes);

/* The output string is an OTHER-class memory-core block, released by
* cbm_layout_free_json; the document keeps its own allocator. */
/* ALLOW_INVALID_UNICODE alone copies malformed UTF-8 through, which
* handle_layout's strict yyjson_read rejects. ESCAPE_UNICODE makes those
* bytes U+FFFD and escapes valid non-ASCII. The output string is an
* OTHER-class memory-core block, released by cbm_layout_free_json; the
* document keeps its own allocator. */
size_t len = 0;
yyjson_write_err write_err = {0};
char *json = yyjson_mut_write_opts(doc, YYJSON_WRITE_ALLOW_INVALID_UNICODE,
cbm_json_writer_allocator(), &len, &write_err);
char *json = yyjson_mut_write_opts(
doc, YYJSON_WRITE_ALLOW_INVALID_UNICODE | YYJSON_WRITE_ESCAPE_UNICODE,
cbm_json_writer_allocator(), &len, &write_err);
yyjson_mut_doc_free(doc);
if (!json) {
char code[CBM_SZ_32];
Expand Down
157 changes: 157 additions & 0 deletions tests/test_ui.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -596,6 +596,162 @@ TEST(layout_json_memory_core_ownership) {
PASS();
}

/* handle_layout re-reads this payload with yyjson_read(..., 0). Malformed
* UTF-8 has to come back as U+FFFD (EF BF BD); valid text has to decode to
* the original string. Stack strings are not owned by the result. */
TEST(layout_json_invalid_utf8_roundtrip) {
ASSERT_NULL(cbm_layout_to_json(NULL));

cbm_layout_result_t empty = {0};
char *json = cbm_layout_to_json(&empty);
ASSERT_NOT_NULL(json);
yyjson_doc *doc = yyjson_read(json, strlen(json), 0);
ASSERT_NOT_NULL(doc);
yyjson_val *root = yyjson_doc_get_root(doc);
ASSERT_EQ((int)yyjson_arr_size(yyjson_obj_get(root, "nodes")), 0);
ASSERT_EQ((int)yyjson_arr_size(yyjson_obj_get(root, "edges")), 0);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(root, "total_nodes")), 0);
yyjson_doc_free(doc);
cbm_layout_free_json(json);

cbm_layout_node_t nodes[4] = {
{
.id = 11,
.x = 1.5f,
.y = -2.0f,
.z = 3.25f,
.label = "Section\x80", /* isolated continuation */
.name = "Head\xC3", /* truncated 2-byte lead */
.qualified_name = "doc.md::H\xE4\xB8", /* truncated 3-byte sequence */
.file_path = "docs/\xFFreadme.md", /* invalid leading byte */
.start_line = 4,
.end_line = 9,
.size = 2.5f,
.color = 0x336699,
.in_calls = 1,
.out_calls = 2,
.status = "structural",
},
{
.id = 12,
.label = "Section",
.name = "q\"\x80", /* quote, then a continuation byte */
.qualified_name = "bs\\\xFF", /* backslash, then an invalid lead */
.file_path = "nl\n\xC2", /* newline, then a truncated lead */
.size = 1.0f,
},
{
.id = 13,
.label = "Section",
.name = "caf\xC3\xA9", /* U+00E9 */
.qualified_name = "\xE7\xAB\xA0\xE8\x8A\x82", /* U+7AE0 U+8282 */
.file_path = "docs/\xF0\x9F\x98\x80.md", /* U+1F600 */
.size = 1.0f,
.status = "structural",
},
{
.id = 14,
.x = NAN,
.y = INFINITY,
.z = -INFINITY,
.size = NAN,
},
};
cbm_layout_edge_t edges[2] = {
{.source = 11, .target = 12, .type = "MENTIONS\x80"},
{.source = 12, .target = 13, .type = NULL},
};
cbm_layout_result_t layout = {
.nodes = nodes,
.node_count = 4,
.edges = edges,
.edge_count = 2,
.total_nodes = 4,
};

json = cbm_layout_to_json(&layout);
ASSERT_NOT_NULL(json);
doc = yyjson_read(json, strlen(json), 0);
ASSERT_NOT_NULL(doc);
for (size_t i = 0; json[i] != '\0'; i++) {
ASSERT((unsigned char)json[i] < 0x80);
}
ASSERT_NOT_NULL(strstr(json, "\"Section\""));
ASSERT_NOT_NULL(strstr(json, "\\u00E9"));
ASSERT_NOT_NULL(strstr(json, "\\u7AE0\\u8282"));
ASSERT_NOT_NULL(strstr(json, "\\uD83D\\uDE00"));
ASSERT_NOT_NULL(strstr(json, "\\uFFFD"));
ASSERT_NOT_NULL(strstr(json, "q\\\"\\uFFFD"));
ASSERT_NOT_NULL(strstr(json, "bs\\\\\\uFFFD"));
ASSERT_NOT_NULL(strstr(json, "nl\\n\\uFFFD"));

root = yyjson_doc_get_root(doc);
yyjson_val *nodes_json = yyjson_obj_get(root, "nodes");
yyjson_val *edges_json = yyjson_obj_get(root, "edges");
ASSERT_EQ((int)yyjson_arr_size(nodes_json), 4);
ASSERT_EQ((int)yyjson_arr_size(edges_json), 2);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(root, "total_nodes")), 4);

yyjson_val *n0 = yyjson_arr_get(nodes_json, 0);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n0, "id")), 11);
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n0, "label")), "Section\xEF\xBF\xBD");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n0, "name")), "Head\xEF\xBF\xBD");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n0, "qualified_name")),
"doc.md::H\xEF\xBF\xBD\xEF\xBF\xBD");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n0, "file_path")), "docs/\xEF\xBF\xBDreadme.md");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n0, "status")), "structural");
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n0, "start_line")), 4);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n0, "end_line")), 9);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n0, "in_calls")), 1);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n0, "out_calls")), 2);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n0, "x")), 1.5, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n0, "y")), -2.0, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n0, "z")), 3.25, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n0, "size")), 2.5, 1e-6);

yyjson_val *n1 = yyjson_arr_get(nodes_json, 1);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n1, "id")), 12);
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n1, "label")), "Section");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n1, "name")), "q\"\xEF\xBF\xBD");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n1, "qualified_name")), "bs\\\xEF\xBF\xBD");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n1, "file_path")), "nl\n\xEF\xBF\xBD");
ASSERT_NULL(yyjson_obj_get(n1, "status"));

yyjson_val *n2 = yyjson_arr_get(nodes_json, 2);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n2, "id")), 13);
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n2, "label")), "Section");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n2, "name")), "caf\xC3\xA9");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n2, "qualified_name")), "\xE7\xAB\xA0\xE8\x8A\x82");
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(n2, "file_path")), "docs/\xF0\x9F\x98\x80.md");

yyjson_val *n3 = yyjson_arr_get(nodes_json, 3);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(n3, "id")), 14);
ASSERT_NULL(yyjson_obj_get(n3, "label"));
ASSERT_NULL(yyjson_obj_get(n3, "name"));
ASSERT_NULL(yyjson_obj_get(n3, "qualified_name"));
ASSERT_NULL(yyjson_obj_get(n3, "file_path"));
ASSERT_NULL(yyjson_obj_get(n3, "status"));
ASSERT_NULL(yyjson_obj_get(n3, "start_line"));
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n3, "x")), 0.0, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n3, "y")), 0.0, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n3, "z")), 0.0, 1e-6);
ASSERT_FLOAT_EQ(yyjson_get_real(yyjson_obj_get(n3, "size")), 1.0, 1e-6);

yyjson_val *e0 = yyjson_arr_get(edges_json, 0);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(e0, "source")), 11);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(e0, "target")), 12);
ASSERT_STR_EQ(yyjson_get_str(yyjson_obj_get(e0, "type")), "MENTIONS\xEF\xBF\xBD");

yyjson_val *e1 = yyjson_arr_get(edges_json, 1);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(e1, "source")), 12);
ASSERT_EQ((int)yyjson_get_int(yyjson_obj_get(e1, "target")), 13);
ASSERT_NULL(yyjson_obj_get(e1, "type"));

yyjson_doc_free(doc);
cbm_layout_free_json(json);
PASS();
}

TEST(layout_null_inputs) {
/* NULL store → NULL result */
cbm_layout_result_t *r = cbm_layout_compute(NULL, "test", CBM_LAYOUT_OVERVIEW, NULL, 0, 100);
Expand Down Expand Up @@ -2783,6 +2939,7 @@ SUITE(ui) {
RUN_TEST(layout_deterministic);
RUN_TEST(layout_to_json);
RUN_TEST(layout_json_memory_core_ownership);
RUN_TEST(layout_json_invalid_utf8_roundtrip);
RUN_TEST(layout_null_inputs);
RUN_TEST(layout_dead_code_classification);
RUN_TEST(layout_coincident_nodes_bounded);
Expand Down
Loading