Skip to content

Collapse whitespace

collapse_whitespace

Collapse-whitespace strategy — reduce redundant blank lines and trailing spaces.

CollapseWhitespaceStrategy

Bases: SmeltStrategy

Collapse consecutive blank lines and strip trailing whitespace.

Source code in packages/axm-smelt/src/axm_smelt/strategies/collapse_whitespace.py
Python
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
class CollapseWhitespaceStrategy(SmeltStrategy):
    """Collapse consecutive blank lines and strip trailing whitespace."""

    @property
    def name(self) -> str:
        """Strategy identifier used in the registry."""
        return "collapse_whitespace"

    @property
    def category(self) -> str:
        """Strategy category (``whitespace``)."""
        return "whitespace"

    def apply(self, ctx: SmeltContext) -> SmeltContext:
        """Collapse redundant whitespace outside fenced code blocks."""
        if ctx.format in _STRUCTURED_FORMATS:
            return ctx

        text = ctx.text

        # Replace fenced code blocks with placeholders, collapse, then restore
        blocks: list[str] = []

        def _save_block(m: re.Match[str]) -> str:
            idx = len(blocks)
            blocks.append(m.group(0))
            return f"\x00CODEBLOCK{idx}\x00"

        stripped = _FENCED_BLOCK_RE.sub(_save_block, text)
        collapsed = _collapse(stripped)

        # Normalize newlines around placeholders so that code blocks don't
        # introduce extra blank lines: one \n before the first placeholder,
        # no separator between adjacent placeholders, one \n after the last.
        collapsed = re.sub(r"\n{2,}(\x00CODEBLOCK)", r"\n\1", collapsed)
        collapsed = re.sub(r"(CODEBLOCK\d+\x00)\n+(\x00CODEBLOCK)", r"\1\2", collapsed)
        collapsed = re.sub(r"(CODEBLOCK\d+\x00)\n{2,}", r"\1\n", collapsed)

        # Restore code blocks
        result = collapsed
        for idx, block in enumerate(blocks):
            result = result.replace(f"\x00CODEBLOCK{idx}\x00", block)

        if result == text:
            return ctx
        return SmeltContext(text=result, format=ctx.format)
category property

Strategy category (whitespace).

name property

Strategy identifier used in the registry.

apply(ctx)

Collapse redundant whitespace outside fenced code blocks.

Source code in packages/axm-smelt/src/axm_smelt/strategies/collapse_whitespace.py
Python
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
def apply(self, ctx: SmeltContext) -> SmeltContext:
    """Collapse redundant whitespace outside fenced code blocks."""
    if ctx.format in _STRUCTURED_FORMATS:
        return ctx

    text = ctx.text

    # Replace fenced code blocks with placeholders, collapse, then restore
    blocks: list[str] = []

    def _save_block(m: re.Match[str]) -> str:
        idx = len(blocks)
        blocks.append(m.group(0))
        return f"\x00CODEBLOCK{idx}\x00"

    stripped = _FENCED_BLOCK_RE.sub(_save_block, text)
    collapsed = _collapse(stripped)

    # Normalize newlines around placeholders so that code blocks don't
    # introduce extra blank lines: one \n before the first placeholder,
    # no separator between adjacent placeholders, one \n after the last.
    collapsed = re.sub(r"\n{2,}(\x00CODEBLOCK)", r"\n\1", collapsed)
    collapsed = re.sub(r"(CODEBLOCK\d+\x00)\n+(\x00CODEBLOCK)", r"\1\2", collapsed)
    collapsed = re.sub(r"(CODEBLOCK\d+\x00)\n{2,}", r"\1\n", collapsed)

    # Restore code blocks
    result = collapsed
    for idx, block in enumerate(blocks):
        result = result.replace(f"\x00CODEBLOCK{idx}\x00", block)

    if result == text:
        return ctx
    return SmeltContext(text=result, format=ctx.format)