Measured 2026-09-22 by scripts/smoke_settings_com.py --build-dwt against the
running AutoCAD 2026 (25.1s), each on a scratch Documents.Add() document:
1. RPC_E_CALL_REJECTED (-2147418111) on Documents.Count straight after
Documents.Open, after Close(False) inside document_close, and in
_ensure_document_state: AutoCAD's own message filter refuses incoming
calls while it is still switching documents. pywin32 exposes no
CoRegisterMessageFilter, so _wait_out_rejected_call waits the window out
for read-only callables (0.25 s -> 1 s pauses, 20 s budget, inside
COM_CALL_TIMEOUT) at those sites and after Documents.Add/Open; any other
com_error is not retried.
2. layer_list raised "'IAcadLayer' object has no attribute 'Color'. Did you
mean: 'color'?": acax25enu.tlb spells IAcadEntity/IAcadLayer `color` and
`Lineweight`. Late-bound dispatch resolved .Color/.LineWeight
case-insensitively; with a makepy cache present (this machine's gen_py was
regenerated on 2026-09-17, and every object a method returns is then a
generated wrapper) the lookup is exact. All 17 sites now use the typelib
spelling, which works on both wrappers; the fakes model the measured names.
3. layer_state_restore raised "'IAcadObject' object has no attribute
'Count'": Dictionaries.Item and IAcadDictionary.Item are declared
IAcadObject* (the narrowing _com_unnarrow already handles for MLEADER
styles). The save path went through the typed Dictionaries.Add, so only
the reopen paths failed. Both now un-narrow; the fakes narrow like pywin32.
4. The Freeze put on layer 0 raises 'Invalid layer' whatever the value, even
False while thawed (color/Linetype/Lineweight/Plottable/LayerOn/Lock/
ActiveLayer all fine on it), so a state saved with 0 current restored with
two warnings and 0 unapplied. Freeze is written only when it changes the
layer.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
ezdxf keeps a multi-line ATTRIB's two placement surfaces (the ATTRIB's own
insert/align_point and the embedded MTEXT's insert) in step through
transform() but not translate(): Insert.translate moves only the ATTRIB's
points (1.4.4 attrib.py has no translate). entity_move / entity_copy /
entity_array_rectangular therefore left the embedded MTEXT behind, so
block_explode burst the note 50 units from its symbol (audit clean), and
block_set_attributes / text_find_replace -- which rebuild the MTEXT and let
set_mtext re-place the ATTRIB from it -- teleported the attribute back to
the old spot while the INSERT stayed put (a regression of e7dafb8/b753a12,
which before wrote only dxf.text).
- _anchored_mtext(attrib): the virtual MTEXT re-anchored to the ATTRIB's
own placement (get_placement() through its OCS, the inverse of ezdxf's
_update_location_from_mtext). _burst_attrib and _write_attrib_value read
through it, so a drawing whose surfaces already disagree bursts and
writes at the ATTRIB's placement instead of the stale MTEXT's.
- _translate(ent, ...): translate + settle every multi-line ATTRIB's MTEXT
onto the ATTRIB (the set_mtext round trip ezdxf's own transform uses),
keeping the whole value in group 1. Used by move, copy, rectangular array.
- _transform(ent, m): transform + restore group 1 of every multi-line
ATTRIB -- BaseAttrib.transform goes through set_mtext, which mirrors only
the first line, so rotate/scale/mirror/polar array/change_space turned
'A\PB' into 'A' on the one surface an R2010 save keeps.
- block_insert of a multi-line ATTDEF: add_auto_attribs embeds through
embed_mtext with the same first-line mirror; the whole value goes into
group 1 the way _write_attrib_value keeps it.
Tests: move/copy/rectangular array carry the MTEXT (explode lands on the
symbol), set_attributes and find_replace after a move leave the attribute
where the move put it, a rotated reference moves both surfaces by the same
delta, a stale drawing bursts/writes at the ATTRIB placement, rotate/scale/
mirror/polar and a fresh insert keep the whole value across an R2010
round trip, and a single-line reference is unchanged. 13 fail before the
fix, all pass after; full suite 2511 passed.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
DXF and ActiveX layer tables are case-insensitive, but block_define collected
the primitives' layers with a case-sensitive set. A request naming 'Alpha' and
'ALPHA' with create_layers=True made ezdxf add the first, then raise
DXFTableEntryError on the second after the write — layer left behind, no block,
surfaced as an internal error instead of a named refusal. COM ran Layers.Add
twice, got the existing layer back the second time, and reported two layers
created for one that appeared; the two engines diverged and the refusal message
without create_layers named the same layer twice.
referenced_layers() in backends/block_specs.py keys the set on the folded name
and keeps the first spelling; both engines consume it. _FakeLayers in the
hardening tests now matches Item/Add case-insensitively and returns the existing
record on a duplicate Add, as the ActiveX table does, so the COM path is pinned.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
base_x/base_y were only float()-cast, so a NaN base point was written into
the BLOCK record; a primitive on a layer the drawing lacks was written as a
dangling reference doc.audit() never flags (and on COM would fail mid-loop
after Blocks.Add). validate_base_point names the key; the layer table is
probed before the first write and the missing names are refused, or created
and reported as layers_created when create_layers=true (new optional kwarg,
contract + both engines + tool).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
block_insert and block_set_attributes applied str() to whatever arrived, so
a JSON null became the text None and a non-mapping failed after the INSERT
was written. validate_attribute_values (backends/block_specs.py) runs first
on both engines: strings pass the one-line rule, ints and finite floats are
written as plain digits, everything else is a TypeError naming the tag.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Headless text_find_replace matched and wrote only dxf.text of an ATTRIB/ATTDEF, so on a multi-line attribute it reported replaced: 1 while block_get_attributes and block_explode still carried the old embedded MTEXT (and a needle on the second line was never found). The ATTRIB/ATTDEF branch now reads through _attrib_value and writes through _write_attrib_value, the same pair block_get_attributes / block_set_attributes use.
COM block_explode swallowed a failing GetAttributes() (attrs = ()), GetConstantAttributes() and Invisible, then exploded, deleted every ATTDEF placeholder and the reference and returned ok: True with no attribute text -- every value destroyed silently on a transient RPC_E_CALL_REJECTED. Those reads now happen before Explode() and propagate with nothing written; a failure after the explode deletes the members it added and the TEXTs written so far and leaves the reference in place (_undo_explode), so the drawing is never left half-burst under an exception. The fakes gain the members a live AcadBlockReference / AcadAttributeDefinition always expose (GetConstantAttributes, Constant, Invisible) instead of relying on the swallow.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two review findings on the multi-line-attribute paths of block_explode /
block_set_attributes.
COM: `_add_mtext_from_capture` wrote `InsertionPoint` before
`AttachmentPoint`. ActiveX's `AttachmentPoint` keeps the text body where it
is and relocates the anchor, so every non-TopLeft attribute was laid out
TopLeft-at-anchor with only the label moved (live: a BottomLeft two-line
note landed two lines low, BottomRight also a box width right). The anchor
is now written last, unconditionally, after the attachment and the frame.
Verified live on AutoCAD 2026: every alignment's bbox matches the control
and the end-to-end burst MTEXT sits at the ATTRIB's TextAlignmentPoint.
The fake MTEXT now relocates `InsertionPoint` on `AttachmentPoint`/`Normal`
writes the way ActiveX does, both tests assert where the anchor ended up,
and the alignment test uses BottomRight rather than the default TopLeft
that could never show the displacement.
ezdxf: `_write_attrib_value` relied on `set_mtext`, which mirrors only the
first line into `dxf.text`; below R2018 ezdxf exports no embedded MTEXT,
so the default R2010 document saved `NEW ONE` where the reader had
promised `NEW ONE\PNEW TWO\PNEW THREE`. `dxf.text` now carries the whole
value (AutoCAD's own `TextString` convention) after the MTEXT rebuild;
the R2010 round-trip test uses a `\P` value, and an R2018 round trip
pins both surfaces.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two review findings on the headless engine, both content lost with ok: True.
Multi-line attribute writer: `block_set_attributes` set only `attrib.dxf.text`,
so after an edit `block_get_attributes` reported the new value while
`block_explode` (`_burst_attrib`) burst the embedded MTEXT at its pre-edit
content -- and on the default R2010 document, where ezdxf exports no embedded
MTEXT, a save/reload carried the new value the exploded drawing never showed.
`_write_attrib_value` now rebuilds the embedded MTEXT from its own virtual
entity with the new content (`virtual_mtext_entity()` + `set_mtext(...,
graphic_properties=False)`, the round trip ezdxf's `BaseAttrib.transform`
uses), keeping placement, style and height; `_attrib_value` reports the
embedded MTEXT's full `Line1\PLine2` content for a multi-line attribute, as
AutoCAD's `TextString` does, so get / set / explode / COM agree.
Opaque members: `block_explode` iterated `virtual_entities()` without a
`skipped_entity_callback`, so ezdxf dropped every member it cannot copy
(OLE2FRAME) or transform (VIEWPORT) with a debug log, an ACAD_PROXY_ENTITY
without proxy graphics contributed nothing, and the reference was then deleted.
`_explodable_members` materializes the list before any write and refuses by
member with the typed capability `explode_opaque_members` (declared on both
engines: headless False with the reason, COM native), naming the block and
the live way out. The block_explode docstrings (backend and tool) now state
this refusal instead of claiming the xref/MINSERT gate covered everything.
Tests: eight new headless cases in tests/test_track_a_hardening.py -- the
multi-line reader, the rewrite (value on every surface, placement unchanged,
clean audit), a `\P` value and the empty value, an R2010 save/reload, the
single-line path unchanged, the opaque-member refusal (nothing written,
reference alive, members named), an ordinary block still exploding under a
uniform scale, and the capability declared on both engines.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
A multi-line ATTRIB keeps its content in an embedded MTEXT; dxf.text is only
the first line (ezdxf) or "Line1\PLine2" (AutoCAD), so the TEXT the explode
emitted silently dropped lines or rendered literal \P with a clean audit.
ezdxf: `_burst_attrib` adds `virtual_mtext_entity()` (graphic properties and
WCS placement already carried; the reference matrix applied for a constant
ATTDEF) and marks it invisible like the ATTRIB. COM: `_capture_attrib` reads
`MTextAttribute` / `MTextAttributeContent` / `MTextBoundaryWidth` and
`_add_mtext_from_capture` goes through `AddMText` (height, style, frame,
then Rotation, then the attachment mapped from the ATTRIB's Alignment).
Two INSERT kinds `virtual_entities()` cannot explode were destroyed with
ok: True: an xref (no virtual entities, reference deleted) and a MINSERT
(only the first cell resolved, the rest of the grid gone). Both are now
refused by name before any write, on both engines (`block_record.is_xref`
/ `mcount > 1` headless; `Blocks.Item(Name).IsXRef` / `AcDbMInsertBlock`
live, the latter previously failing the INSERT check with the misleading
"not a block reference").
Tests: nine new cases in tests/test_track_a_hardening.py (headless multi-line
ATTRIB and constant ATTDEF, xref and MINSERT refusals, a single-cell grid
that still explodes; COM multi-line capture, frame-before-angle, and both
named refusals). The fake ActiveX surface gains AddMText and IsXRef.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
COM: ``text.Normal`` was a plain Python tuple, which ActiveX marshals as
VT_ARRAY|VT_VARIANT and refuses with E_INVALIDARG (verified live on
AutoCAD 2026); because the write came after Explode(), the ATTDEF deletes
and AddText, the tool raised and left the drawing torn. It now goes
through ``_apoint`` like every other point write, and the fake ActiveX
surface refuses a bare tuple on Normal/InsertionPoint/TextAlignmentPoint
so a test cannot pin the wrong call shape again.
COM: every AcDbAttributeDefinition Explode() returned was deleted, but
GetAttributes() excludes constant attributes, so a constant attribute's
only visible text was destroyed with ok: True. A constant ATTDEF from the
explode (Constant member, or its tag in GetConstantAttributes()) is now
captured like an ATTRIB and replaced by a TEXT; placeholders are still
just deleted.
ezdxf: an AutoCAD-authored reference carries no ATTRIB for a constant
attribute and virtual_entities() skips ATTDEF, so the same text vanished
headlessly. A constant ATTDEF whose tag has no attached ATTRIB becomes a
TEXT transformed by the reference matrix, the path add_auto_attribs takes,
and is not converted twice when block_insert did attach one.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
The BURST path on the live engine read TextString/InsertionPoint/Height/
Rotation/Layer/Invisible and nothing else, so the +Z TEXT that AddText
builds got the ATTRIB's OCS Rotation with no Normal: a mirrored
reference's tag (Normal (0,0,-1), the same premise the ezdxf regression
pins) came back reading theta instead of 180deg-theta, un-mirrored, on
the wrong side of its anchor at theta=0, with ok: True. Style members
(StyleName, ObliqueAngle, ScaleFactor, Backward/UpsideDown, Alignment +
TextAlignmentPoint, Color) were dropped the same way.
_capture_attrib_text_members reads the ActiveX twins of the headless
_ATTRIB_TO_TEXT carry, guarded per member and never written as a guess;
a +Z Normal is not carried and a left-aligned ATTRIB drops its
TextAlignmentPoint (ActiveX refuses it). Writes go Normal/Thickness/
style -> Rotation -> InsertionPoint re-asserted in WCS when the frame
changed -> Alignment -> TextAlignmentPoint -> Layer -> Color -> Visible.
Two fake-surface tests: the mirrored/styled ATTRIB pins every write and
the Normal-before-Rotation and Normal-before-InsertionPoint order; the
bare ATTRIB pins that absent members are skipped. Both fail on the
previous commit.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two silent wrong-place defects in the BURST path. An ATTRIB is an OCS
entity, so the TEXT that replaces it now carries extrusion/thickness with
insert/align_point; a mirrored reference (ATTRIB reflected to (0,0,-1),
INSERT +Z with yscale=-1) put the tag 50 units across the mirror axis with
a clean audit. Members and TEXTs now go into the INSERT's owner layout
(ezdxf get_layout, COM ObjectIdToObject(OwnerID)) instead of the current
space: a sheet title block exploded while Model was current duplicated
geometry into model space and then died on delete_entity headless, and on
COM landed the tag text in model space with ok: True. A reference nested
in a block definition is refused before any write on both engines.
Five regression tests (mirrored INSERT, sheet-owned INSERT with Model
current, nested INSERT, and the COM owner-block / nested-refusal call
shapes) fail on the previous commit and pass here.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
virtual_entities() yields geometry only, so an exploded tagged symbol lost
its tag value silently. Every attached ATTRIB now becomes a TEXT at the
same WCS placement/height/rotation/style/layer (invisible stays
invisible), reported as attribute_texts. COM reads GetAttributes before
Explode, deletes the ATTDEF placeholders Explode returns, AddText per
value, and deletes the reference ActiveX Explode leaves behind.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Track A's task-1 gate covered block_insert only; the sibling tool still
returned a handle for a typo'd name that doc.audit() later deleted, and
both tools accepted *Model_Space (a reference cycle). One gate per engine
(_require_block / _require_block_defined) is now shared by both placing
tools: non-string name -> TypeError, unknown -> ValueError naming it,
layout block -> ValueError, all before add_blockref / InsertBlock.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
