feat: also parse within formatting

Signed-off-by: Konstantin Fickel <mail@konstantinfickel.de>

src/streamer/parse/extract_tag.py

@@ -1,45 +1,84 @@
 import re
 from typing import Iterable
 from mistletoe.block_token import BlockToken
-from mistletoe.span_token import RawText
+from mistletoe.span_token import Emphasis, RawText, Strikethrough, Strong, Link
 from mistletoe.token import Token
 
 from .markdown_tag import Tag
 
 
-def extract_tags(tokens: Iterable[Token]) -> list[str]:
-    return [token.content for token in tokens if isinstance(token, Tag)]
+def extract_markers_and_tags_from_single_token(
+    token: Token,
+    marker_boundary_encountered: bool,
+    return_at_first_marker: bool = False,
+) -> tuple[list[str], list[str], bool]:
+    result_markers, result_tags = [], []
+    result_marker_boundary_encountered = marker_boundary_encountered
+
+    if isinstance(token, Tag):
+        if marker_boundary_encountered:
+            result_tags.append(token.content)
+        else:
+            result_markers.append(token.content)
+    elif isinstance(token, (Emphasis, Strong, Strikethrough, Link)):
+        markers, tags, child_marker_boundary_encountered = (
+            extract_markers_and_tags_from_tokens(
+                token.children or [],
+                marker_boundary_encountered,
+                return_at_first_marker,
+            )
+        )
+        result_markers.extend(markers)
+        result_tags.extend(tags)
+        result_marker_boundary_encountered = (
+            marker_boundary_encountered or child_marker_boundary_encountered
+        )
+    elif isinstance(token, RawText) and re.match(r"^[\s]*$", token.content):
+        pass
+    else:
+        result_marker_boundary_encountered = True
+
+    return result_markers, result_tags, result_marker_boundary_encountered
+
+
+def extract_markers_and_tags_from_tokens(
+    tokens: Iterable[Token],
+    marker_boundary_encountered: bool,
+    return_at_first_marker: bool = False,
+) -> tuple[list[str], list[str], bool]:
+    result_markers, result_tags = [], []
+    result_marker_boundary_encountered = marker_boundary_encountered
+
+    for child in tokens:
+        markers, tags, child_marker_boundary_encountered = (
+            extract_markers_and_tags_from_single_token(
+                child, result_marker_boundary_encountered, return_at_first_marker
+            )
+        )
+        result_markers.extend(markers)
+        result_tags.extend(tags)
+        result_marker_boundary_encountered = (
+            marker_boundary_encountered or child_marker_boundary_encountered
+        )
+
+        if len(result_markers) > 0 and return_at_first_marker:
+            break
+
+    return result_markers, result_tags, result_marker_boundary_encountered
 
 
 def extract_markers_and_tags(block_token: BlockToken) -> tuple[list[str], list[str]]:
-    markers, tags = [], []
-    is_marker = True
-
-    if block_token.children is None:
-        return [], []
-
-    for token in block_token.children:
-        if isinstance(token, Tag):
-            if is_marker:
-                markers.append(token)
-            else:
-                tags.append(token)
-        elif not (isinstance(token, RawText) and re.match(r"^[\s]*$", token.content)):
-            is_marker = False
-
-    return extract_tags(markers), extract_tags(tags)
+    markers, tags, _ = extract_markers_and_tags_from_tokens(
+        block_token.children or [], False
+    )
+    return markers, tags
 
 
 def has_markers(block_token: BlockToken) -> bool:
-    if block_token.children is None:
-        return False
-
-    for child in block_token.children:
-        if isinstance(child, Tag):
-            return True
-        elif not (isinstance(child, RawText) and re.match(r"^[\s]*$", child.content)):
-            return False
-    return False
+    markers, _, _ = extract_markers_and_tags_from_tokens(
+        block_token.children or [], False, return_at_first_marker=True
+    )
+    return len(markers) > 0
 
 
-__all__ = ["extract_tags", "extract_markers_and_tags", "has_markers"]
+__all__ = ["extract_markers_and_tags", "has_markers"]

src/streamer/parse/markdown_tag.py

@@ -5,7 +5,7 @@ from mistletoe.span_token import SpanToken
 
 class Tag(SpanToken):
     parse_inner = False
-    pattern = re.compile(r"@([^\s]+)")
+    pattern = re.compile(r"@([^\s*\x60~\[\]]+)")
 
 
 class TagMarkdownRenderer(MarkdownRenderer):