👌 IMPROVE: MyST role syntax parsing (#32)

Match characters instead of using regex.
This now allows for an unlimited name length and new lines in the content.

Author: chrisjsewell

mdit_py_plugins/myst_role/index.py

@@ -4,7 +4,7 @@
 from markdown_it.common.utils import escapeHtml
 from markdown_it.rules_inline import StateInline
 
-PATTERN = re.compile(r"^\{([a-zA-Z0-9\_\-\+\:]{1,36})\}(`+)(?!`)(.+?)(?<!`)\2(?!`)")
+VALID_NAME_PATTERN = re.compile(r"^\{([a-zA-Z0-9\_\-\+\:]+)\}")
 
 
 def myst_role_plugin(md: MarkdownIt):
@@ -14,22 +14,45 @@ def myst_role_plugin(md: MarkdownIt):
 
 
 def myst_role(state: StateInline, silent: bool):
+
+    # check name
+    match = VALID_NAME_PATTERN.match(state.src[state.pos :])
+    if not match:
+        return False
+    name = match.group(1)
+
+    # check for starting backslash escape
     try:
         if state.srcCharCode[state.pos - 1] == 0x5C:  # /* \ */
             # escaped (this could be improved in the case of edge case '\\{')
             return False
     except IndexError:
         pass
 
-    match = PATTERN.search(state.src[state.pos :])
+    # scan opening tick length
+    start = pos = state.pos + match.end()
+    try:
+        while state.src[pos] == "`":
+            pos += 1
+    except IndexError:
+        return False
+
+    tick_length = pos - start
+    if not tick_length:
+        return False
+
+    # search for closing ticks
+    match = re.search("`" * tick_length, state.src[pos + 1 :])
     if not match:
         return False
-    state.pos += match.end()
+    content = state.src[pos : pos + match.start() + 1].replace("\n", " ")
 
     if not silent:
         token = state.push("myst_role", "", 0)
-        token.meta = {"name": match.group(1)}
-        token.content = match.group(3)
+        token.meta = {"name": name}
+        token.content = content
+
+    state.pos = pos + match.end() + 1
 
     return True
 
@@ -38,7 +61,5 @@ def render_myst_role(self, tokens, idx, options, env):
     token = tokens[idx]
     name = token.meta.get("name", "unknown")
     return (
-        '<code class="sphinx-role">'
-        f"{{{name}}}[{escapeHtml(token.content)}]"
-        "</code>"
+        '<code class="myst role">' f"{{{name}}}[{escapeHtml(token.content)}]" "</code>"
     )

tests/fixtures/myst_role.md

@@ -3,14 +3,30 @@ Basic:
 .
 {abc}`xyz`
 .
-<p><code class="sphinx-role">{abc}[xyz]</code></p>
+<p><code class="myst role">{abc}[xyz]</code></p>
+.
+
+Multiple:
+.
+{abc}`xyz`{def}`uvw`
+.
+<p><code class="myst role">{abc}[xyz]</code><code class="myst role">{def}[uvw]</code></p>
 .
 
 Surrounding Text:
 .
 a {abc}`xyz` b
 .
-<p>a <code class="sphinx-role">{abc}[xyz]</code> b</p>
+<p>a <code class="myst role">{abc}[xyz]</code> b</p>
+.
+
+New lines:
+.
+{abc}`xy
+z` `d
+e`
+.
+<p><code class="myst role">{abc}[xy z]</code> <code>d e</code></p>
 .
 
 In Code:
@@ -20,20 +36,26 @@ In Code:
 <p><code>{abc}`xyz`</code></p>
 .
 
+Empty content:
+.
+{name}`` a
+.
+<p>{name}`` a</p>
+.
 
 Surrounding Code:
 .
 `a` {abc}`xyz` `b`
 .
-<p><code>a</code> <code class="sphinx-role">{abc}[xyz]</code> <code>b</code></p>
+<p><code>a</code> <code class="myst role">{abc}[xyz]</code> <code>b</code></p>
 .
 
 In list:
 .
 - {abc}`xyz`
 .
 <ul>
-<li><code class="sphinx-role">{abc}[xyz]</code></li>
+<li><code class="myst role">{abc}[xyz]</code></li>
 </ul>
 .
 
@@ -42,15 +64,22 @@ In Quote:
 > {abc}`xyz` b
 .
 <blockquote>
-<p><code class="sphinx-role">{abc}[xyz]</code> b</p>
+<p><code class="myst role">{abc}[xyz]</code> b</p>
 </blockquote>
 .
 
 Multiple ticks:
 .
 {abc}``xyz``
 .
-<p><code class="sphinx-role">{abc}[xyz]</code></p>
+<p><code class="myst role">{abc}[xyz]</code></p>
+.
+
+Inner tick:
+.
+{abc}``x`yz``
+.
+<p><code class="myst role">{abc}[x`yz]</code></p>
 .
 
 Unbalanced ticks: