08.09.2025 - 21.09.2025
Disclaimer: Please take everything I write here with a grain of salt—I’m by no means an expert. Some of these concepts I’m encountering for the first time, and my understanding might have significant gaps—and that’s okay. It’s one of the reasons I decided to take on this project. If you see anything wrong, let me know and I’ll fix it.
Nothing different in this sprint compared to the last 3 tbh. I’ve added the remaining tests from tests1.dat - almost 50. No major issues, just some steps missing here and there.
...
else if (is_eof)
{
INCOMPLETE_IMPLEMENTATION("parse error");
dom_node_t* current = stack[stack_idx];
if (current->name == html_tag_script())
{
INCOMPLETE_IMPLEMENTATION("set already_started to true");
}
stack_pop();
consume = false;
mode = original_mode;
}
...
static void test_parser_35()
{
// #data
// <font><p>hello<b>cruel</font>world
// #errors
// (1,6): expected-doctype-but-got-start-tag
// (1,29): adoption-agency-1.3
// (1,29): adoption-agency-1.3
// (1,34): expected-closing-tag-but-got-eof
// #document
// | <html>
// | <head>
// | <body>
// | <font>
// | <p>
// | <font>
// | "hello"
// | <b>
// | "cruel"
// | <b>
// | "world"
unsigned char buffer[] = "<font><p>hello<b>cruel</font>world";
dom_node_t* expected = dom_document_new();
dom_node_t* html = dom_element_new(expected, html_tag_html());
dom_node_t* head = dom_element_new(expected, html_tag_head());
dom_node_t* body = dom_element_new(expected, html_tag_body());
dom_node_t* font1 = dom_element_new(expected, html_tag_font());
dom_node_t* p1 = dom_element_new(expected, html_tag_p());
dom_node_t* font2 = dom_element_new(expected, html_tag_font());
dom_node_t* b1 = dom_element_new(expected, html_tag_b());
dom_node_t* b2 = dom_element_new(expected, html_tag_b());
dom_node_t* t1 = dom_text_new(expected, "hello", 5);
dom_node_t* t2 = dom_text_new(expected, "cruel", 5);
dom_node_t* t3 = dom_text_new(expected, "world", 5);
APPEND_TO_TREE(expected, html);
APPEND_TO_TREE(html, head);
APPEND_TO_TREE(html, body);
APPEND_TO_TREE(body, font1);
APPEND_TO_TREE(body, p1);
APPEND_TO_TREE(p1, font2);
APPEND_TO_TREE(font2, t1);
APPEND_TO_TREE(font2, b1);
APPEND_TO_TREE(b1, t2);
APPEND_TO_TREE(p1, b2);
APPEND_TO_TREE(b2, t3);
RUN_TEST_AND_ASSERT_DOCUMENT(buffer, expected);
}
The above test was producing the wrong tree
--------------------Actual
# document
html
head
body
font
p
font
#text - hello
b
#text - cruel
#text - world
--------------------Expected
// #document
// | <html>
// | <head>
// | <body>
// | <font>
// | <p>
// | <font>
// | "hello"
// | <b>
// | "cruel"
// | <b>
// | "world"
The issue was that i was overriding elements in the formatting_elements array instead of inserting in it.
-remove_formatting_element(formatting_node);
-formatting_elements[bookmark] = new_element;
-formatting_elements_m[bookmark] = false;
-memcpy(&formatting_elements_t[bookmark], formatting_node_t, sizeof(html_token_t));
+insert_formatting_element(new_element, formatting_node_t, bookmark);
+static void insert_formatting_element(dom_node_t* node, html_token_t* token, uint32_t idx)
+{
+ for (uint32_t i = formatting_elements_size; i > idx; i--)
+ {
+ formatting_elements[i] = formatting_elements[i - 1];
+ formatting_elements_m[i] = formatting_elements_m[i - 1];
+ memcpy(&formatting_elements_t[i], &formatting_elements_t[i - 1], sizeof(html_token_t));
+ }
+
+ formatting_elements[idx] = node;
+ formatting_elements_m[idx] = false;
+ memcpy(&formatting_elements_t[idx], token, sizeof(html_token_t));
+ formatting_elements_size++;
+}
static void test_parser_49()
{
// #data
// <DIV> abc <B> def <I> ghi <P> jkl </B>
// #errors
// (1,5): expected-doctype-but-got-start-tag
// (1,38): adoption-agency-1.3
// (1,38): expected-closing-tag-but-got-eof
// #document
// | <html>
// | <head>
// | <body>
// | <div>
// | " abc "
// | <b>
// | " def "
// | <i>
// | " ghi "
// | <i>
// | <p>
// | <b>
// | " jkl "
unsigned char buffer[] = "<DIV> abc <B> def <I> ghi <P> jkl </B>";
dom_node_t* expected = dom_document_new();
dom_node_t* html = dom_element_new(expected, html_tag_html());
dom_node_t* head = dom_element_new(expected, html_tag_head());
dom_node_t* body = dom_element_new(expected, html_tag_body());
dom_node_t* div = dom_element_new(expected, html_tag_div());
dom_node_t* t1 = dom_text_new(expected, " abc ", 5);
dom_node_t* b1 = dom_element_new(expected, html_tag_b());
dom_node_t* b2 = dom_element_new(expected, html_tag_b());
dom_node_t* t2 = dom_text_new(expected, " def ", 5);
dom_node_t* i1 = dom_element_new(expected, html_tag_i());
dom_node_t* i2 = dom_element_new(expected, html_tag_i());
dom_node_t* t3 = dom_text_new(expected, " ghi ", 5);
dom_node_t* p1 = dom_element_new(expected, html_tag_p());
dom_node_t* t4 = dom_text_new(expected, " jkl ", 5);
APPEND_TO_TREE(expected, html);
APPEND_TO_TREE(html, head);
APPEND_TO_TREE(html, body);
APPEND_TO_TREE(body, div);
APPEND_TO_TREE(div, t1);
APPEND_TO_TREE(div, b1);
APPEND_TO_TREE(div, i2);
APPEND_TO_TREE(b1, t2);
APPEND_TO_TREE(b1, i1);
APPEND_TO_TREE(i1, t3);
APPEND_TO_TREE(i2, p1);
APPEND_TO_TREE(p1, b2);
APPEND_TO_TREE(b2, t4);
RUN_TEST_AND_ASSERT_DOCUMENT(buffer, expected, false);
}
The resulting trees
------------------------------- ACTUAL
# document
html
head
body
div
#text - abc
b
#text - def
i
#text - ghi
p
b
#text - jkl
i
p
b
#text - jkl
-------------------------------- EXPECTED
// #data
// <DIV> abc <B> def <I> ghi <P> jkl </B>
// #errors
// (1,5): expected-doctype-but-got-start-tag
// (1,38): adoption-agency-1.3
// (1,38): expected-closing-tag-but-got-eof
// #document
// | <html>
// | <head>
// | <body>
// | <div>
// | " abc "
// | <b>
// | " def "
// | <i>
// | " ghi "
// | <i>
// | <p>
// | <b>
// | " jkl "
Basically it looked like nodes were getting duplicated, but it was just a matter of not removing old sibling information from the structs.
dom_node_t* dom_node_remove(dom_node_t* node, dom_node_t* child)
{
..
+ child->parent = NULL;
+ child->next = NULL;
+ child->prev = NULL;
return child;
}
I also added implementations for sections that were not complete.
One thing im noticing is that adding these tests manually is a hassle. I would like to add automated test runners for both the tokenizer and the tree-construction tests. This is what I will focus on for the next couple of sprints.
Martin