fix: parse class suffixes, labeled dotted/thick links, and frontmatter

This commit is contained in:
Matt Van Horn
2026-08-11 19:31:39 -07:00
parent 9e50da6d67
commit a4be5bed0d
3 changed files with 94 additions and 4 deletions
+48
View File
@@ -181,9 +181,56 @@ A --> B; B --> C
("B", "C"), ("B", "C"),
]: ]:
fail("semicolon-separated flowchart statements were not split") fail("semicolon-separated flowchart statements were not split")
labeled_file = tmp / "labeled-links.mmd"
labeled_file.write_text(
"""flowchart LR
A[Start]:::warning --> B
B -. retry .-> C
C == critical ==> D
D:::danger --- E
""",
encoding="utf-8",
)
labeled = json.loads(run_extract([str(labeled_file), "--json"]))["diagrams"][0]
ids = [node["id"] for node in labeled["nodes"]]
if ids != ["A", "B", "C", "D", "E"]:
fail(f"`:::class` suffixes leaked into node ids: {ids}")
if labeled["nodes"][0]["label"] != "Start":
fail("shaped node with a class suffix lost its label")
dotted, thick = labeled["edges"][1], labeled["edges"][2]
if dotted["label"] != "retry" or dotted["style"] != "dashed":
fail("labeled dotted link did not retain its label and dashed style")
if thick["label"] != "critical" or thick["style"] != "thick":
fail("labeled thick link did not retain its label and thick style")
ok("documented shape families and edge forms normalize") ok("documented shape families and edge forms normalize")
def check_frontmatter(tmp: Path) -> None:
front_file = tmp / "frontmatter.mmd"
front_file.write_text(
"""---
title: Checkout
config:
theme: forest
---
flowchart LR
A --> B
""",
encoding="utf-8",
)
digest = run_extract([str(front_file)])
if "direction: LR" not in digest or "[0] flowchart (2n/1e)" not in digest:
fail("leading Mermaid frontmatter was not skipped before grammar detection")
if "Checkout" in digest or "forest" in digest:
fail("frontmatter config leaked into the digest")
unterminated = tmp / "unterminated-frontmatter.mmd"
unterminated.write_text("---\ntitle: Broken\nflowchart LR\nA --> B\n", encoding="utf-8")
expect_error([str(unterminated)], "not a Mermaid file")
ok("frontmatter is skipped, and an unterminated block still fails specifically")
def check_markdown_and_grammars(tmp: Path) -> None: def check_markdown_and_grammars(tmp: Path) -> None:
header = run_extract([str(README_FIXTURE)]) header = run_extract([str(README_FIXTURE)])
if "2 diagram(s)" not in header or "[1] sequenceDiagram" not in header: if "2 diagram(s)" not in header or "[1] sequenceDiagram" not in header:
@@ -480,6 +527,7 @@ def main() -> int:
check_files() check_files()
check_flowchart() check_flowchart()
check_shape_and_edge_vocabulary(tmp) check_shape_and_edge_vocabulary(tmp)
check_frontmatter(tmp)
check_markdown_and_grammars(tmp) check_markdown_and_grammars(tmp)
check_adversarial(tmp) check_adversarial(tmp)
check_errors_and_limits(tmp) check_errors_and_limits(tmp)
@@ -63,7 +63,7 @@ Load the selected `type-*.md`. Override the grammar only when the content disagr
- Start from a blank `viewBox` selected by the size preset. Mermaid positions do not exist in the source, and a renderer's positions must not be recreated. - Start from a blank `viewBox` selected by the size preset. Mermaid positions do not exist in the source, and a renderer's positions must not be recreated.
- Use semantic treatments from the chosen type. A Mermaid cylinder becomes Store/State; a rhombus stays a decision only in a flowchart; subgraphs become zones or collapsible groups. - Use semantic treatments from the chosen type. A Mermaid cylinder becomes Store/State; a rhombus stays a decision only in a flowchart; subgraphs become zones or collapsible groups.
- Ignore init themes, `style`, `classDef`, `class`, and `linkStyle`. One accent plus the ink ramp replaces the source theme. - Ignore init themes, `style`, `classDef`, `class`, inline `:::class` attachments, and `linkStyle`. One accent plus the ink ramp replaces the source theme. A leading `---` frontmatter block is title/config, so it is skipped with the same reasoning.
- Reroute all connections with the SKILL.md §6 connector rules. Mermaid edge length markers are ranking hints, not content. - Reroute all connections with the SKILL.md §6 connector rules. Mermaid edge length markers are ranking hints, not content.
- Do not add a component merely to fill space. Imports remain bounded by source meaning. - Do not add a component merely to fill space. Imports remain bounded by source meaning.
@@ -253,12 +253,37 @@ def load_blocks(path: Path) -> list[SourceBlock]:
return blocks return blocks
FRONTMATTER_MAX_LINES = 40
def _frontmatter_end(lines: list[str]) -> int:
"""Return the last line index of a leading `---` frontmatter block, or -1."""
first = next(
(index for index, line in enumerate(lines) if line.strip()),
None,
)
if first is None or lines[first].strip() != "---":
return -1
limit = min(len(lines), first + FRONTMATTER_MAX_LINES + 1)
for index in range(first + 1, limit):
if lines[index].strip() == "---":
return index
return -1
def _prepared_lines(block: SourceBlock) -> list[tuple[int, str]]: def _prepared_lines(block: SourceBlock) -> list[tuple[int, str]]:
prepared: list[tuple[int, str]] = [] prepared: list[tuple[int, str]] = []
in_directive = False in_directive = False
for offset, raw in enumerate(block.text.splitlines()): lines = block.text.splitlines()
frontmatter_end = _frontmatter_end(lines)
for offset, raw in enumerate(lines):
line_number = block.source_line + offset line_number = block.source_line + offset
stripped = raw.strip() stripped = raw.strip()
if offset <= frontmatter_end:
# Mermaid's `--- title: ... ---` frontmatter is source config; the
# redraw discards it the same way it discards `%%{init}%%`.
prepared.append((line_number, ""))
continue
if in_directive: if in_directive:
if "}%%" in stripped: if "}%%" in stripped:
in_directive = False in_directive = False
@@ -373,8 +398,20 @@ def classify_shape(expression: str) -> str:
return "rect" return "rect"
def _strip_class_suffix(text: str) -> str:
"""Drop Mermaid's `:::class` attachment; source styling is discarded."""
mask = _top_level_mask(text)
index = mask.find(":::")
if index == -1:
return text
end = index + 3
while end < len(text) and (text[end].isalnum() or text[end] in "_-"):
end += 1
return (text[:index] + text[end:]).strip()
def _parse_node_expression(expression: str) -> tuple[str, str, str] | None: def _parse_node_expression(expression: str) -> tuple[str, str, str] | None:
text = expression.strip().rstrip(";").strip() text = _strip_class_suffix(expression.strip().rstrip(";").strip())
if not text: if not text:
return None return None
match = re.match(r"^([\w.:-]+)", text, re.UNICODE) match = re.match(r"^([\w.:-]+)", text, re.UNICODE)
@@ -415,7 +452,12 @@ def _edge_operators(text: str) -> list[_Operator]:
operators: list[_Operator] = [] operators: list[_Operator] = []
occupied: list[tuple[int, int]] = [] occupied: list[tuple[int, int]] = []
text_edge = re.compile(r"--\s+(.+?)\s+(-{2,}>|--[xo])") # Labeled links carry the label between the opening and closing operator:
# `A-- text -->B`, `A-. retry .-> B`, `A== critical ==> B`, and the
# undirected forms of each.
text_edge = re.compile(
r"(?:--|-\.|==)\s+(.+?)\s+(\.-+[>xo]|\.-+|-{2,}>|--[xo]|=+>|={2,}|-{3,})"
)
for match in text_edge.finditer(mask): for match in text_edge.finditer(mask):
token = match.group(2) token = match.group(2)
style, arrowhead, bidirectional, undirected = _operator_style(token) style, arrowhead, bidirectional, undirected = _operator_style(token)