Markdown LOC family - mehen

The Markdown LOC family separates physical lines by what they contain. A 1,000-line file with 700 lines of code fences is not the same artifact as a 1,000-line prose file, and a single SLOC count would hide that.

What mehen emits

Key	Meaning
`MD.DLOC`	Physical Markdown lines (everything).
`MD.PLOC`	Prose lines (narrative paragraphs, headings, list text).
`MD.CLOC`	Code-fence and indented-code lines.
`MD.TLOC`	Table lines.
`MD.MLOC`	Math block lines.
`MD.BLOC`	Blank lines.
`MD.ALOC`	Artifact lines = `CLOC + TLOC + MLOC` + diagram / raw HTML / MDX lines.

Derived ratios

Each component / max(1, DLOC) produces a ratio:

ArtifactLineRatio
CodeLineRatio
TableLineRatio
MathLineRatio
BlankLineRatio

These ratios drive document classification (prose-dominant vs. artifact-dominant) and anomaly detection.

Worked example

# Auth API

Brief intro paragraph.

```ts
export async function login(): Promise<Session> { /* … */ }
```

| Field | Required |
|---|---|
| `email` | yes |

loc:
  dloc: 11
  ploc: 4
  cloc: 3
  tloc: 3
  mloc: 0
  bloc: 2
  aloc: 6

References

Park, R. E. (1992). Software Size Measurement: A Framework for Counting Source Statements. CMU/SEI-92-TR-20 — the standard reference for LOC categorization that this Markdown family extends to prose / code / table / math / blank lines. SEI report.
CommonMark Spec: Container blocks and leaf blocks — the parser-level distinctions the per-construct counts rely on.

See also

Effective Content Units — review-mass normalization that uses these counts.
Section tree — per-section LOC aggregation.