:hedgehog: PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP.
Python ab-testingai-analyticsanalyticscdpdata-warehouseexperimentsfeature-flagsjavascriptproduct-analyticspythonreactsession-replaysurveystypescriptweb-analytics
import time
from django.test import SimpleTestCase
from parameterized import parameterized
from posthog.schema import MarkdownBlock, SessionReplayBlock
from posthog.comment.tiptap import _parse_inline, markdown_to_tiptap_nodes
from ee.hogai.artifacts.types import StoredBlock, VisualizationRefBlock
from ee.hogai.tools.create_notebook.tiptap import blocks_to_tiptap_doc, tiptap_doc_to_text
class TestParseInline(SimpleTestCase):
@parameterized.expand(
[
("plain text", "hello world", [{"type": "text", "text": "hello world"}]),
(
"bold",
"**bold text**",
[{"type": "text", "text": "bold text", "marks": [{"type": "bold"}]}],
),
(
"italic",
"*italic text*",
[{"type": "text", "text": "italic text", "marks": [{"type": "italic"}]}],
),
(
"inline code",
"`some code`",
[{"type": "text", "text": "some code", "marks": [{"type": "code"}]}],
),
(
"link",
"[click here](https://example.com)",
[
{
"type": "text",
"text": "click here",
"marks": [{"type": "link", "attrs": {"href": "https://example.com"}}],
}
],
),
(
"bold link label",
"[**Set up**](https://example.com)",
[
{
"type": "text",
"text": "Set up",
"marks": [
{"type": "bold"},
{"type": "link", "attrs": {"href": "https://example.com"}},
],
}
],
),
(
"unsafe link stays text",
"[click](javascript:alert(1))",
[{"type": "text", "text": "[click](javascript:alert(1))"}],
),
(
"mixed inline",
"hello **bold** and *italic*",
[
{"type": "text", "text": "hello "},
{"type": "text", "text": "bold", "marks": [{"type": "bold"}]},
{"type": "text", "text": " and "},
{"type": "text", "text": "italic", "marks": [{"type": "italic"}]},
],
),
("empty", "", []),
]
)
def test_parse_inline(self, _name, text, expected):
assert _parse_inline(text) == expected
class TestMarkdownToTiptapNodes(SimpleTestCase):
@parameterized.expand(
[
(
"h1",
"# Hello",
[{"type": "heading", "attrs": {"level": 1}, "content": [{"type": "text", "text": "Hello"}]}],
),
(
"h3",
"### Sub heading",
[
{
"type": "heading",
"attrs": {"level": 3},
"content": [{"type": "text", "text": "Sub heading"}],
}
],
),
(
"paragraph",
"Just a paragraph.",
[{"type": "paragraph", "content": [{"type": "text", "text": "Just a paragraph."}]}],
),
(
"code block",
"```python\nprint('hi')\n```",
[
{
"type": "codeBlock",
"attrs": {"language": "python"},
"content": [{"type": "text", "text": "print('hi')"}],
}
],
),
(
"code block no lang",
"```\nsome code\n```",
[{"type": "codeBlock", "content": [{"type": "text", "text": "some code"}]}],
),
(
"bullet list",
"- item one\n- item two",
[
{
"type": "bulletList",
"content": [
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "item one"}],
}
],
},
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "item two"}],
}
],
},
],
}
],
),
(
"ordered list",
"1. first\n2. second",
[
{
"type": "orderedList",
"content": [
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "first"}],
}
],
},
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "second"}],
}
],
},
],
}
],
),
(
"ordered list keeps its numbers across a blank line",
"1. first\n\n2. second",
[
{
"type": "orderedList",
"content": [
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "first"}],
}
],
},
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "second"}],
}
],
},
],
}
],
),
(
"ordered item keeps the next line",
"1. first\nSee more",
[
{
"type": "orderedList",
"content": [
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "first See more"}],
}
],
}
],
}
],
),
(
"nested bullets stay inside the numbered step",
"1. step\n - bullet\n2. next",
[
{
"type": "orderedList",
"content": [
{
"type": "listItem",
"content": [
{"type": "paragraph", "content": [{"type": "text", "text": "step"}]},
{
"type": "bulletList",
"content": [
{
"type": "listItem",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "bullet"}],
}
],
}
],
},
],
},
{
"type": "listItem",
"content": [
{"type": "paragraph", "content": [{"type": "text", "text": "next"}]},
],
},
],
}
],
),
(
"empty heading does not stall the next line",
"## \nAfter",
[
{"type": "paragraph", "content": [{"type": "text", "text": "## "}]},
{"type": "paragraph", "content": [{"type": "text", "text": "After"}]},
],
),
("empty string", "", []),
("whitespace only", " \n \n ", []),
]
)
def test_markdown_to_tiptap_nodes(self, _name, markdown, expected):
assert markdown_to_tiptap_nodes(markdown) == expected
def test_unclosed_brackets_stay_text(self):
text = "[" * 20000
started = time.monotonic()
nodes = _parse_inline(text)
assert time.monotonic() - started < 1
assert nodes == [{"type": "text", "text": text}]
def test_failed_link_destination_is_not_rescanned(self):
text = "[" * 4000 + "](https://" + "a" * 4000
started = time.monotonic()
nodes = _parse_inline(text)
assert time.monotonic() - started < 1
assert "".join(node["text"] for node in nodes) == text
blocked = "[" * 2000 + "](javascript:" + "a" * 2000 + ")"
started = time.monotonic()
nodes = _parse_inline(blocked)
assert time.monotonic() - started < 1
assert "".join(node["text"] for node in nodes) == blocked
def test_multi_block_markdown(self):
md = "# Title\n\nSome text.\n\n- a\n- b\n\n```\ncode\n```"
nodes = markdown_to_tiptap_nodes(md)
assert len(nodes) == 4
assert nodes[0]["type"] == "heading"
assert nodes[1]["type"] == "paragraph"
assert nodes[2]["type"] == "bulletList"
assert nodes[3]["type"] == "codeBlock"
class TestBlocksToTiptapDoc(SimpleTestCase):
def test_empty_blocks_with_title(self):
doc = blocks_to_tiptap_doc([], title="My Notebook")
assert doc["type"] == "doc"
assert len(doc["content"]) == 1
assert doc["content"][0]["type"] == "heading"
assert doc["content"][0]["attrs"]["level"] == 1
def test_empty_blocks_no_title(self):
doc = blocks_to_tiptap_doc([])
assert doc == {"type": "doc", "content": [{"type": "paragraph"}]}
def test_markdown_block(self):
blocks = [MarkdownBlock(content="## Section\n\nHello world.")]
doc = blocks_to_tiptap_doc(blocks)
content = doc["content"]
assert content[0]["type"] == "heading"
assert content[1]["type"] == "paragraph"
def test_session_replay_block(self):
blocks = [SessionReplayBlock(session_id="abc123", timestamp_ms=0)]
doc = blocks_to_tiptap_doc(blocks)
node = doc["content"][0]
assert node["type"] == "ph-recording"
assert node["attrs"]["id"] == "abc123"
def test_visualization_ref_without_resolver(self):
blocks = [VisualizationRefBlock(artifact_id="xyz")]
doc = blocks_to_tiptap_doc(blocks)
node = doc["content"][0]
assert node["type"] == "paragraph"
assert "Visualization: xyz" in node["content"][0]["text"]
def test_visualization_ref_with_resolver(self):
def resolver(artifact_id):
return {
"query": {"kind": "InsightVizNode", "source": {"kind": "TrendsQuery"}},
"name": "My Chart",
}
blocks = [VisualizationRefBlock(artifact_id="xyz")]
doc = blocks_to_tiptap_doc(blocks, resolve_visualization=resolver)
node = doc["content"][0]
assert node["type"] == "ph-query"
assert node["attrs"]["title"] == "My Chart"
assert node["attrs"]["query"]["kind"] == "InsightVizNode"
def test_visualization_ref_resolver_returns_none(self):
blocks = [VisualizationRefBlock(artifact_id="xyz")]
doc = blocks_to_tiptap_doc(blocks, resolve_visualization=lambda _: None)
node = doc["content"][0]
assert node["type"] == "paragraph"
assert "not found" in node["content"][0]["text"]
def test_mixed_blocks(self):
blocks: list[StoredBlock] = [
MarkdownBlock(content="# Intro\n\nSome text."),
VisualizationRefBlock(artifact_id="v1"),
SessionReplayBlock(session_id="s1", timestamp_ms=0),
MarkdownBlock(content="## Conclusion"),
]
doc = blocks_to_tiptap_doc(blocks, title="Report")
content = doc["content"]
# Title heading + heading from markdown + paragraph + viz fallback + recording + conclusion heading
types = [n["type"] for n in content]
assert types[0] == "heading" # title
assert "ph-recording" in types
class TestTiptapDocToText(SimpleTestCase):
@parameterized.expand(
[
("none_doc", None, ""),
("empty_dict", {}, ""),
("empty_content", {"type": "doc", "content": []}, ""),
(
"heading",
{
"type": "doc",
"content": [
{"type": "heading", "attrs": {"level": 2}, "content": [{"type": "text", "text": "Section"}]}
],
},
"## Section",
),
(
"paragraph",
{"type": "doc", "content": [{"type": "paragraph", "content": [{"type": "text", "text": "Hello"}]}]},
"Hello",
),
(
"code_block_with_lang",
{
"type": "doc",
"content": [
{
"type": "codeBlock",
"attrs": {"language": "sql"},
"content": [{"type": "text", "text": "SELECT 1"}],
}
],
},
"```sql\nSELECT 1\n```",
),
(
"bullet_list",
{
"type": "doc",
"content": [
{
"type": "bulletList",
"content": [
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "one"}]}],
},
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "two"}]}],
},
],
}
],
},
"- one\n- two",
),
(
"ordered_list",
{
"type": "doc",
"content": [
{
"type": "orderedList",
"content": [
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "first"}]}],
},
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "second"}]}],
},
],
}
],
},
"1. first\n2. second",
),
(
"ph_recording",
{
"type": "doc",
"content": [{"type": "ph-recording", "attrs": {"id": "sess-abc"}}],
},
'<session_replay id="sess-abc" />',
),
(
"inline_bold",
{
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [{"type": "text", "text": "important", "marks": [{"type": "bold"}]}],
}
],
},
"**important**",
),
(
"inline_link",
{
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "click",
"marks": [{"type": "link", "attrs": {"href": "https://example.com"}}],
}
],
}
],
},
"[click](https://example.com)",
),
]
)
def test_tiptap_doc_to_text(self, _name, doc, expected):
assert tiptap_doc_to_text(doc) == expected
def test_ph_query_node(self):
doc = {
"type": "doc",
"content": [
{
"type": "ph-query",
"attrs": {
"title": "My Trend",
"query": {
"kind": "InsightVizNode",
"source": {"kind": "TrendsQuery"},
},
},
}
],
}
result = tiptap_doc_to_text(doc)
assert '<insight title="My Trend"' in result
assert 'query_kind="InsightVizNode"' in result
assert 'source_kind="TrendsQuery"' in result
assert "</insight>" in result
def test_markdown_notebook_v2_node(self):
doc = {
"type": "doc",
"content": [
{
"type": "ph-markdown-notebook",
"attrs": {
"markdown": '# V2 notebook\n\n<Query query={{"kind":"SavedInsightNode","shortId":"abc"}} />'
},
}
],
}
assert (
tiptap_doc_to_text(doc) == '# V2 notebook\n\n<Query query={{"kind":"SavedInsightNode","shortId":"abc"}} />'
)
def test_multi_node_doc(self):
doc = {
"type": "doc",
"content": [
{"type": "heading", "attrs": {"level": 1}, "content": [{"type": "text", "text": "Report"}]},
{"type": "paragraph", "content": [{"type": "text", "text": "Summary below."}]},
{
"type": "bulletList",
"content": [
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "item"}]}],
}
],
},
],
}
result = tiptap_doc_to_text(doc)
assert "# Report" in result
assert "Summary below." in result
assert "- item" in result
def test_ph_query_node_with_stringified_query(self):
# Some persisted notebooks store `attrs.query` as a JSON-encoded string rather than a dict.
# The formatter must parse it instead of crashing with AttributeError.
doc = {
"type": "doc",
"content": [
{
"type": "ph-query",
"attrs": {
"title": "Pages les plus touchées",
"query": '{"kind": "InsightVizNode", "source": {"kind": "TrendsQuery"}}',
},
}
],
}
result = tiptap_doc_to_text(doc)
assert '<insight title="Pages les plus touchées"' in result
assert 'query_kind="InsightVizNode"' in result
assert 'source_kind="TrendsQuery"' in result
def test_ph_query_node_with_unparseable_query_string(self):
doc = {
"type": "doc",
"content": [
{
"type": "ph-query",
"attrs": {"title": "Broken", "query": "not-json"},
}
],
}
result = tiptap_doc_to_text(doc)
assert '<insight title="Broken"' in result
assert 'query_kind="unknown"' in result
assert "source_kind=" not in result
def test_ph_query_node_with_string_source(self):
doc = {
"type": "doc",
"content": [
{
"type": "ph-query",
"attrs": {
"title": "Edgy",
"query": {"kind": "InsightVizNode", "source": "TrendsQuery"},
},
}
],
}
result = tiptap_doc_to_text(doc)
assert 'query_kind="InsightVizNode"' in result
assert "source_kind=" not in result
def test_ph_query_node_missing_attrs(self):
doc = {"type": "doc", "content": [{"type": "ph-query"}]}
result = tiptap_doc_to_text(doc)
assert '<insight title="Untitled"' in result
assert 'query_kind="unknown"' in result
@parameterized.expand(
[
("string_node", "stray string"),
("none_node", None),
("number_node", 42),
("list_node", ["unexpected"]),
]
)
def test_non_dict_top_level_node_is_skipped(self, _name, bad_node):
doc = {
"type": "doc",
"content": [
{"type": "paragraph", "content": [{"type": "text", "text": "before"}]},
bad_node,
{"type": "paragraph", "content": [{"type": "text", "text": "after"}]},
],
}
result = tiptap_doc_to_text(doc)
assert "before" in result
assert "after" in result
def test_non_dict_inline_node_is_skipped(self):
doc = {
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [
{"type": "text", "text": "kept"},
"stray inline string",
{"type": "text", "text": " also kept"},
],
},
],
}
result = tiptap_doc_to_text(doc)
assert result == "kept also kept"
def test_non_dict_list_item_is_skipped(self):
doc = {
"type": "doc",
"content": [
{
"type": "bulletList",
"content": [
"stray",
{
"type": "listItem",
"content": [{"type": "paragraph", "content": [{"type": "text", "text": "real"}]}],
},
],
}
],
}
result = tiptap_doc_to_text(doc)
assert "- real" in result
def test_string_attrs_are_coerced_or_ignored(self):
# Even if `attrs` itself is a string, we should not crash.
doc = {
"type": "doc",
"content": [
{"type": "ph-recording", "attrs": "not-a-dict"},
{
"type": "ph-recording",
"attrs": '{"id": "sess-json"}',
},
],
}
result = tiptap_doc_to_text(doc)
assert '<session_replay id="unknown" />' in result
assert '<session_replay id="sess-json" />' in result