Give every frame fragment a single content root - #97
Merged
Merged
Conversation
Overlay 6 and 8 each carried their letter in a group sitting beside #frame_overlay rather than inside it, where 5 and 7 - the same two overlays with a different letter - put theirs. The suffixes those groups carry, -3 and -5, are what an editor adds when a group is duplicated, so this was one overlay copied rather than two drawn differently. Nothing is drawn differently. The group was the last thing in the document and becomes the last thing in #frame_overlay, which is itself last, so paint order holds; and #frame_overlay carries no attributes beyond its id, so nothing new is inherited. Checked by comparing every drawing element before and after with its ancestors' attributes folded in - 14 each, identical. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
CTG's changes. Three fragments drew part of the frame outside the group that names it: 0_160_0 and 0_560_0 left the right-hand bump as a loose <path>, and 0_060_0 held it in an unnamed group alongside a second copy of the cloverleaf that #frame already drew. Both have a counterpart that shows the intended shape - 0_160_0c and 0_060_0c, the same frames in the other identity colour - and both keep everything inside #frame, with no duplicate. 0_160_0 and 0_560_0 draw exactly what they drew. 0_060_0 loses one of its three paths, which is the duplicated cloverleaf: identical d, fill and stroke to the one kept, and opaque, so drawing it twice and drawing it once come out the same. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
CTG's changes. Five fragments split the frame in two: the filled polygon inside #frame, and the stroked outline of the same polygon in an unnamed group beside it. 0_327_0 - the same frame, undashed - already held both inside #frame, so it is what the other five now look like. Nothing is drawn differently: the outline was already painting after the fill and still does, and the group it came out of carried no attributes to inherit. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Ten fragments called their content root frame_1_ or frame_2_ - the suffix an editor adds when a group is duplicated, rather than a name anything chose - and two of those, 0_435_0 and 0_435_0c, put it on a bare <polyline> with no group at all, so there was nothing for a consumer to take hold of. Those two are now wrapped in a <g id="frame"> like the rest; the other eight are renamed. The normaliser leaves these alone by design: its generated-id test is an SVG element name followed by digits, and "frame" is not an element name, so frame_1_ reads as a name worth keeping. It is a name, just not a useful one. An id is not part of SvgFingerprint and nothing in the tree references one of these, so nothing is drawn differently. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Fourteen frames carried a <g id="octagon"> - the bounding octagon, its mod1 and mod2 guide lines, and the outFrame outline. That is drawing scaffolding for positioning a main icon or a sector modifier within the octagon, and a frame is the thing drawn around the octagon rather than within it, so it was never of any use here. It was display="none" throughout, so nothing is drawn differently, and the measured frame bounds come back identical - which also confirms foxglove was already honouring the display, rather than the group having quietly counted towards a frame's extent all this time. With this, all 225 frame fragments hold exactly one content element: 217 a <g id="frame">, and the eight overlays a <g id="frame_overlay">. Closes #79 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #79.
All 225 frame fragments now hold exactly one content element — 217 a
<g id="frame">, and the eight overlays a<g id="frame_overlay">. Nothing else at the root.The survey turned up more than the issue's ten, because it also looked at naming and at the construction guide the issue had silently discounted. Five commits, in the order they were worked out.
What was wrong
#frame_overlayrather than inside it<path>, or an unnamed<g>)#frame#frame0_327_0frame_1_/frame_2_frame; two were a bare<polyline>and are now wrapped<g id="octagon">construction guideMost of these had a sibling that already showed the intended shape —
0_160_0c,0_060_0c,0_327_0, overlays 5 and 7 — so the target was rarely a judgement call.The octagon guide is the bounding octagon, its mod1/mod2 guide lines and the
outFrameoutline: scaffolding for positioning a main icon or sector modifier within the octagon. A frame is drawn around the octagon rather than within it, so it was never of any use in these files.0_060_0is the only file that loses anything. Its unnamed group held a second copy of the cloverleaf#framealready drew, alongside the bump. Identicald, fill and stroke, and opaque, so drawing it twice and drawing it once come out the same — and0_060_0chas the bump in#framewith no duplicate, which is what it now matches.Why the normaliser had not already caught the names
FragmentNormaliser's generated-id test is an SVG element name followed by digits.frameis not an element name, soframe_1_reads as a name worth keeping. It is a name — just not a useful one.Verification
SvgFingerprintcannot guard this, for the same reason it could not guard #78: it records structure, so moving an element between groups changes it by design. Instead, every drawing element compared before and after with its ancestors' attributes folded in anddisplay:nonesubtrees left out — so a move between groups carrying nothing but an id is invisible to the comparison, and so is the removal of a hidden guide.0_060_0's duplicate, 3 drawn → 2.FragmentMeasurerre-run: frame bounds identical, which also confirms foxglove was already honouring the guide'sdisplay, rather than it quietly counting towards a frame's extent.mvn verifygreen.Next
This makes frames the second family with a settled shape, after the free canvas fragments in #78. Once #80 gives the remaining fragments their canonical ids,
FragmentShapeCheckercan cover all three rather than just free canvas.The matching changes to
jmsfx-halluxare in the hallux working tree, uncommitted.🤖 Generated with Claude Code