Enhance DWG/DXF analyzer with column-wall-beam-slab detail reporting

[smitelagwar]

Summary

• classify INSERT blocks by structural naming prefixes to detect columns, walls, beams, and slabs
• add per-element tabular output including element name, layer, coordinates, width, height, and area
• include slab extraction from both slab-named blocks and slab-named closed polylines with dimensions
• compute and report nearby pair distances for columns and beams (optionally filtered by --yakinlik-esigi)
• keep axis spacing analysis and add richer full report formatting sections

CLI

• new option: --yakinlik-esigi <float> to report only close column/beam pairs within threshold distance

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c9d6e05cff

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Parse POLYLINE vertices when extracting slab polygons

_extract_slab_polygons explicitly queries both LWPOLYLINE and POLYLINE, but _polyline_to_polygon only reads coordinates via polyline.get_points(). For classic POLYLINE entities this call fails, is swallowed by the broad except, and returns None, so closed slab contours stored as POLYLINE are silently skipped and their area/count are missing from the report.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Tighten slab prefix matching to avoid false slab detection

Using SLAB_PREFIXES = ("D",) with startswith makes any block/layer name beginning with D classify as a slab, which can incorrectly pull in unrelated entities (for example common CAD names like DEFPOINTS/DOOR*) and inflate slab counts and total slab area; this impacts report accuracy whenever naming is not strictly constrained to slab-only D* identifiers.

Useful? React with 👍 / 👎.