Merge remote-tracking branch 'origin/main' into codex/pr28-final

scripts/build-icons.py

@@ -246,11 +246,13 @@ def normalize_url(raw: str) -> str:
     - White / near-white fills (#fff, #ffffff, white) are kept as-is — they
       act as contrast cutouts in 2-colour logos (e.g. white letter on coloured
       circle).  Every other colour becomes currentColor.
-    - inline style="fill:#..." is rewritten to a fill= attribute.
+    - inline fill/stroke colours are normalized without discarding other styles.
     - Inkscape layer translate() transforms are stripped (canvas offset artefact).
     """
-    vb_match = re.search(r'viewBox="([^"]+)"', raw, re.IGNORECASE)
-    viewbox = vb_match.group(1) if vb_match else "0 0 128 128"
+    vb_match = re.search(
+        r'\bviewBox\s*=\s*(["\'])(.*?)\1', raw, flags=re.IGNORECASE
+    )
+    viewbox = vb_match.group(2).strip() if vb_match else "0 0 128 128"
     inner = re.search(r"<svg\b[^>]*>(.*?)</svg>", raw, flags=re.DOTALL | re.IGNORECASE)
     if not inner:
         raise ValueError("no <svg> in payload")
@@ -261,27 +263,55 @@ def normalize_url(raw: str) -> str:
     # Strip Inkscape canvas-offset translate transforms from group wrappers
     body = re.sub(r'\btransform="translate\([^"]+\)"', '', body)
 
-    def _rewrite_fill(m: re.Match) -> str:
+    def _normalized_fill(value: str) -> str:
         """Keep white; turn everything else into currentColor."""
-        hex_val = m.group(1).lower().strip()
+        hex_val = value.lower().strip()
         if hex_val in ("#fff", "#ffffff", "white", "#fefefe", "#fdfdfd"):
-            return 'fill="#fff"'
-        return 'fill="currentColor"'
+            return "#fff"
+        return "currentColor"
 
-    # Rewrite inline style="fill:#rrggbb" → standalone fill= attribute
+    def _normalize_inline_style(match: re.Match) -> str:
+        quote, declarations = match.group(1), match.group(2)
+        declarations = re.sub(
+            r'(\bfill\s*:\s*)(#[0-9a-fA-F]{3,8}|white)\b',
+            lambda fill: f"{fill.group(1)}{_normalized_fill(fill.group(2))}",
+            declarations,
+            flags=re.IGNORECASE,
+        )
+        declarations = re.sub(
+            r'(\bstroke\s*:\s*)#[0-9a-fA-F]{3,8}\b',
+            lambda stroke: f"{stroke.group(1)}currentColor",
+            declarations,
+            flags=re.IGNORECASE,
+        )
+        return f"style={quote}{declarations}{quote}"
+
+    # Normalize supported inline colours while preserving unrelated declarations.
     body = re.sub(
-        r'style="[^"]*fill:\s*(#[0-9a-fA-F]{3,8}|white)[^"]*"',
-        lambda m: _rewrite_fill(re.search(r'(#[0-9a-fA-F]{3,8}|white)', m.group(0))),
+        r'style\s*=\s*(["\'])(.*?)\1',
+        _normalize_inline_style,
         body,
+        flags=re.IGNORECASE,
     )
-    # Rewrite standalone fill="#rrggbb" attributes
+    # Rewrite standalone fill="#rrggbb" attributes.
     body = re.sub(
-        r'\bfill="(#[0-9a-fA-F]{3,8}|white)"',
-        _rewrite_fill,
+        r'\bfill\s*=\s*(["\'])(#[0-9a-fA-F]{3,8}|white)\1',
+        lambda m: f'fill="{_normalized_fill(m.group(2))}"',
         body,
+        flags=re.IGNORECASE,
+    )
+    body = re.sub(
+        r'\bstroke\s*=\s*(["\'])#[0-9a-fA-F]{3,8}\1',
+        'stroke="currentColor"',
+        body,
+        flags=re.IGNORECASE,
+    )
+    body = re.sub(
+        r'\bfill\s*=\s*(["\'])none\1',
+        '',
+        body,
+        flags=re.IGNORECASE,
     )
-    body = re.sub(r'\bstroke="#[0-9a-fA-F]{3,8}"', 'stroke="currentColor"', body)
-    body = re.sub(r'\bfill="none"', '', body)
     body = re.sub(r"\s+", " ", body).strip()
     return (
         f'<svg aria-hidden="true" width="24" height="24" viewBox="{viewbox}" '
@@ -292,6 +322,10 @@ def normalize_url(raw: str) -> str:
 def normalize_devicon(raw: str) -> str:
     """Devicon SVGs have colored fills and a 128x128 or 32x32 viewBox.
     Strip fills/classes and rewrite to 24x24 currentColor."""
+    viewbox_match = re.search(
+        r'\bviewBox\s*=\s*(["\'])(.*?)\1', raw, flags=re.IGNORECASE
+    )
+    viewbox = viewbox_match.group(2).strip() if viewbox_match else "0 0 128 128"
     inner = re.search(r"<svg\b[^>]*>(.*?)</svg>", raw, flags=re.DOTALL | re.IGNORECASE)
     if not inner:
         raise ValueError("no <svg> in Devicon payload")
@@ -303,7 +337,7 @@ def normalize_devicon(raw: str) -> str:
     body = re.sub(r'\bclass="[^"]*"', 'fill="currentColor"', body)
     body = re.sub(r"\s+", " ", body).strip()
     return (
-        '<svg aria-hidden="true" width="24" height="24" viewBox="0 0 128 128" '
+        f'<svg aria-hidden="true" width="24" height="24" viewBox="{viewbox}" '
         f'fill="currentColor">{body}</svg>'
     )
 

scripts/fix-mojibake.py

@@ -17,8 +17,10 @@ PAIRS = [
     ('…', '…'),  # horizontal ellipsis
     ('•', '•'),  # bullet
     ('→', '→'),  # right arrow
-    ('↑', '←'),  # left arrow
-    ('↑', '↑'),  # up arrow (same prefix collision - rare)
+    # CP1252 leaves byte 0x90 undefined, so spell the left-arrow corruption
+    # with a Unicode escape instead of embedding an invisible control character.
+    ('\u00e2\u2020\u0090', '←'),  # left arrow
+    ('↑', '↑'),  # up arrow
     ('↓', '↓'),  # down arrow
     ('≤', '≤'),  # less-than-or-equal
     ('≥', '≥'),  # greater-than-or-equal

scripts/test-build-icons-devicon.py

@@ -0,0 +1,40 @@
+#!/usr/bin/env python3
+"""Regression tests for Devicon viewBox preservation."""
+
+from __future__ import annotations
+
+import importlib.util
+import sys
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parent.parent
+BUILD = ROOT / "scripts" / "build-icons.py"
+
+
+def load_build_module():
+    sys.dont_write_bytecode = True
+    spec = importlib.util.spec_from_file_location("diagram_design_build_icons", BUILD)
+    if spec is None or spec.loader is None:
+        raise AssertionError("could not load build-icons.py")
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+def main() -> int:
+    build_icons = load_build_module()
+    svg = '<svg viewBox="0 0 32 32"><path fill="#123456" d="M0 0h32v32H0z"/></svg>'
+    normalized = build_icons.normalize_devicon(svg)
+    if 'viewBox="0 0 32 32"' not in normalized:
+        raise AssertionError(f"source viewBox was not preserved: {normalized}")
+    if 'fill="#123456"' in normalized:
+        raise AssertionError("source fill escaped currentColor normalization")
+    if 'fill="currentColor"' not in normalized:
+        raise AssertionError("normalized icon does not inherit currentColor")
+    print("PASS: Devicon normalization preserves non-128 source viewBoxes")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

scripts/test-build-icons-quoted-attributes.py

@@ -0,0 +1,58 @@
+#!/usr/bin/env python3
+"""Regression tests for quote-independent URL SVG normalization."""
+
+from __future__ import annotations
+
+import importlib.util
+import re
+import sys
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parent.parent
+BUILD = ROOT / "scripts" / "build-icons.py"
+
+
+def load_build_module():
+    sys.dont_write_bytecode = True
+    spec = importlib.util.spec_from_file_location("diagram_design_build_icons", BUILD)
+    if spec is None or spec.loader is None:
+        raise AssertionError("could not load build-icons.py")
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+def main() -> int:
+    build_icons = load_build_module()
+    svg = (
+        "<svg viewBox='0 0 32 16'>"
+        "<path style='fill:#123456;stroke:#654321;opacity:.5' d='M0 0h32v16H0z'/>"
+        "<circle fill='#abcdef' stroke='#111111' cx='8' cy='8' r='4'/>"
+        "</svg>"
+    )
+    normalized = build_icons.normalize_url(svg)
+    if 'viewBox="0 0 32 16"' not in normalized:
+        raise AssertionError(f"single-quoted viewBox was not preserved: {normalized}")
+    if any(color in normalized for color in ("#123456", "#654321", "#abcdef", "#111111")):
+        raise AssertionError(f"source colors escaped normalization: {normalized}")
+    path = re.search(r"<path\b([^>]*)>", normalized)
+    if path is None:
+        raise AssertionError(f"normalized path is missing: {normalized}")
+    path_attributes = path.group(1)
+    for declaration in ("fill:currentColor", "stroke:currentColor", "opacity:.5"):
+        if declaration not in path_attributes:
+            raise AssertionError(
+                f"path lost or failed to normalize {declaration}: {normalized}"
+            )
+    if (
+        'fill="currentColor"' not in normalized
+        or 'stroke="currentColor"' not in normalized
+    ):
+        raise AssertionError(f"standalone quoted colours were not normalized: {normalized}")
+    print("PASS: URL SVG normalization handles single-quoted attributes")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

scripts/test-fix-mojibake.py

@@ -0,0 +1,40 @@
+#!/usr/bin/env python3
+"""Regression tests for the mojibake repair utility."""
+
+from __future__ import annotations
+
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parent.parent
+SCRIPT = ROOT / "scripts" / "fix-mojibake.py"
+
+
+def main() -> int:
+    # These are the strings produced when UTF-8 arrow bytes are decoded as
+    # Windows-1252. Byte 0x90 is undefined in CP1252, hence the explicit escape.
+    source = "right →, left \u00e2\u2020\u0090, up ↑, down ↓\n"
+    expected = "right →, left ←, up ↑, down ↓\n"
+    with tempfile.TemporaryDirectory(prefix="diagram-design-mojibake-") as directory:
+        path = Path(directory) / "fixture.md"
+        path.write_text(source, encoding="utf-8")
+        result = subprocess.run(
+            [sys.executable, str(SCRIPT), str(path)],
+            capture_output=True,
+            text=True,
+            check=False,
+        )
+        if result.returncode != 0:
+            raise AssertionError(f"repair failed: {result.stdout}{result.stderr}")
+        actual = path.read_text(encoding="utf-8")
+        if actual != expected:
+            raise AssertionError(f"expected {expected!r}, got {actual!r}")
+    print("PASS: mojibake arrow sequences repair to the intended directions")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())