feat(mdx-storage): Phase2 Task2.5 blockquote 변환 구현

confluence-mdx/bin/mdx_to_storage/emitter.py

@@ -84,6 +84,9 @@ def emit_block(block: Block, context: Optional[dict] = None) -> str:
     if block.type == "table":
         return _emit_markdown_table(block.content)
 
+    if block.type == "blockquote":
+        return _emit_blockquote(block.content)
+
     return ""
 
 
@@ -283,3 +286,33 @@ def _replace_cell(match: re.Match[str]) -> str:
         return f"<{tag}{attrs}>{convert_inline(inner.strip())}</{tag}>"
 
     return pattern.sub(_replace_cell, stripped)
+
+
+def _emit_blockquote(content: str) -> str:
+    raw_lines = content.splitlines()
+    stripped_lines: list[str] = []
+    for line in raw_lines:
+        line = line.lstrip()
+        if line.startswith(">"):
+            line = line[1:]
+            if line.startswith(" "):
+                line = line[1:]
+        stripped_lines.append(line)
+
+    paragraphs: list[str] = []
+    current: list[str] = []
+    for line in stripped_lines:
+        if line.strip():
+            current.append(line.strip())
+            continue
+        if current:
+            paragraphs.append(" ".join(current))
+            current = []
+    if current:
+        paragraphs.append(" ".join(current))
+
+    if not paragraphs:
+        return "<blockquote><p></p></blockquote>"
+
+    body = "".join(f"<p>{convert_inline(text)}</p>" for text in paragraphs)
+    return f"<blockquote>{body}</blockquote>"

confluence-mdx/bin/mdx_to_storage/parser.py

@@ -85,6 +85,11 @@ def parse_mdx(text: str) -> list[Block]:
             blocks.append(block)
             continue
 
+        if _is_blockquote_line(line):
+            block, i = _parse_blockquote_block(lines, i)
+            blocks.append(block)
+            continue
+
         if _is_list_line(line):
             block, i = _parse_list_block(lines, i)
             blocks.append(block)
@@ -259,6 +264,23 @@ def _parse_markdown_table_block(lines: list[str], start: int) -> tuple[Block, in
     return Block(type="table", content=content), i
 
 
+def _parse_blockquote_block(lines: list[str], start: int) -> tuple[Block, int]:
+    i = start
+    while i < len(lines):
+        current = lines[i]
+        if current == "":
+            if i + 1 < len(lines) and _is_blockquote_line(lines[i + 1]):
+                i += 1
+                continue
+            break
+        if not _is_blockquote_line(current):
+            break
+        i += 1
+
+    content = "\n".join(lines[start:i]) + "\n"
+    return Block(type="blockquote", content=content), i
+
+
 def _parse_html_block(lines: list[str], start: int) -> tuple[Block, int]:
     if lines[start].startswith("<table"):
         i = start + 1
@@ -305,6 +327,8 @@ def _starts_new_block(line: str) -> bool:
         return True
     if line.startswith("<figure"):
         return True
+    if _is_blockquote_line(line):
+        return True
     if _is_list_line(line):
         return True
     if _is_html_block_start(line):
@@ -369,3 +393,8 @@ def _is_markdown_table_separator(line: str) -> bool:
         if "-" not in cell:
             return False
     return True
+
+
+def _is_blockquote_line(line: str) -> bool:
+    stripped = line.lstrip()
+    return stripped.startswith("> ") or stripped == ">"

confluence-mdx/tests/test_mdx_to_storage/test_emitter.py

@@ -448,3 +448,38 @@ def test_emit_paragraph_then_markdown_table():
     assert "<p>Intro paragraph.</p>" in xhtml
     assert "<table><tbody>" in xhtml
     assert "<td><p>1</p></td>" in xhtml
+
+
+def test_emit_blockquote_single_paragraph():
+    mdx = "> quoted **text** and `code`\n"
+    xhtml = emit_document(parse_mdx(mdx))
+    assert xhtml == "<blockquote><p>quoted <strong>text</strong> and <code>code</code></p></blockquote>"
+
+
+def test_emit_blockquote_multiple_paragraphs():
+    mdx = "> first line\n>\n> second line\n"
+    xhtml = emit_document(parse_mdx(mdx))
+    assert xhtml == "<blockquote><p>first line</p><p>second line</p></blockquote>"
+
+
+def test_emit_blockquote_multiline_single_paragraph():
+    """Multiple `>` lines without blank separator → merged into one paragraph."""
+    mdx = "> line one\n> line two\n> line three\n"
+    xhtml = emit_document(parse_mdx(mdx))
+    assert xhtml == "<blockquote><p>line one line two line three</p></blockquote>"
+
+
+def test_emit_blockquote_in_mixed_document():
+    """Blockquote integrated with other block types."""
+    mdx = "## Section\n\n> Important note.\n\nFollowing text.\n"
+    xhtml = emit_document(parse_mdx(mdx))
+    assert "<h1>Section</h1>" in xhtml
+    assert "<blockquote><p>Important note.</p></blockquote>" in xhtml
+    assert "<p>Following text.</p>" in xhtml
+
+
+def test_emit_blockquote_empty_body():
+    """Empty blockquote `>` only → blockquote with empty paragraph."""
+    mdx = ">\n"
+    xhtml = emit_document(parse_mdx(mdx))
+    assert xhtml == "<blockquote><p></p></blockquote>"

confluence-mdx/tests/test_mdx_to_storage/test_parser.py

@@ -213,3 +213,33 @@ def test_parse_inline_code_with_pipes_stays_paragraph():
     mdx = "Expression: `a == 'x' || b == 'y'` is valid.\n"
     blocks = parse_mdx(mdx)
     assert blocks[0].type == "paragraph"
+
+
+def test_parse_blockquote_block():
+    mdx = "> quoted line\n> second line\n"
+    blocks = parse_mdx(mdx)
+    assert len(blocks) == 1
+    assert blocks[0].type == "blockquote"
+    assert "> quoted line" in blocks[0].content
+
+
+def test_parse_blockquote_with_blank_and_stop_before_next_block():
+    mdx = "> first\n>\n> second\n## Heading\n"
+    blocks = parse_mdx(mdx)
+    assert blocks[0].type == "blockquote"
+    assert blocks[1].type == "heading"
+
+
+def test_parse_blockquote_followed_by_paragraph():
+    """Blockquote ends when non-`>` line appears; next block is paragraph."""
+    mdx = "> quote\n\nParagraph after.\n"
+    blocks = parse_mdx(mdx)
+    types = [b.type for b in blocks if b.type != "empty"]
+    assert types == ["blockquote", "paragraph"]
+
+
+def test_parse_triple_gt_not_blockquote():
+    """Lines like `>>> token` (shell output) should NOT be parsed as blockquote."""
+    mdx = ">>> Service Account token\n"
+    blocks = parse_mdx(mdx)
+    assert blocks[0].type != "blockquote"