WIP: Add v4 remote-module ingestion (two-phase, per-commit hooks) by hjmjohnson · Pull Request #6204 · InsightSoftwareConsortium/ITK

hjmjohnson · 2026-05-04T18:33:42Z

This PR is a WIP because each new remote module adds minor new elements that need to be addressed.

PROMPT: Use the ~/src/ITK-ingest-v4/Utilities/Maintenance/RemoteModuleIngest tooling  to ingest the remote module on the latest upstream/main branch for <<MODULE_NAME>>

Adds a v4 remote-module ingestion pipeline that splits the v3 single-driver flow into two independent phases — Phase A is fully local and reversible; Phase B publishes to upstream only after the Phase A PR has merged on ITK main. Each historical commit is individually re-formatted (clang-format / black / gersemi / trailing-ws / EOF / line-ending) and retroactively conformed to ITK ghostflow rules (subject ≤78 chars, blank line after subject, exec-bit cleared on text files), so every historical commit independently passes ITK's CI gates after the merge.

WIP — opening for design feedback. Validated end-to-end on Cuberille (PR #6205): ghostflow now reports a single, unavoidable error (the upstream root commit). All other categories (long subjects, conflict markers in .orig, exec-bit on CMakeLists.txt, two-line subjects) are eliminated.

Why two phases (motivation)

In v3 (ingest-remote-module.sh) the script orchestrated upstream archival as a bonus step at the end of the ingest. Re-running the ingest after a CI failure risked re-touching the upstream remote, and the upstream-archival step was a manual checklist (per feedback_ingest_archive_readme.md) — easy to forget steps.

v4 splits these into two independent phases:

Phase	Script	Local-clone fate	Upstream side-effects
A — Ingestion	`ingest-module-v4.sh`	Throwaway	None (no pushes to upstream remote)
B — Upstream archival	`archive-remote-module.sh`	Fresh clone	Pushes one removal commit + promotes `MIGRATION_README.md` to `README.md` + flips `archived=true`

The phases share one persistent artifact: the whitelist at Utilities/Maintenance/RemoteModuleIngest/whitelists/<Module>.list. Every other piece of state is reproducible.

The central correctness property: upstream is never touched until ITK has accepted the migration.

Per-commit hook + ghostflow coverage

sanitize-history.py walks every commit on the rewritten ingestion branch via git_filter_repo's blob_callback + commit_callback. Because blob_callback is filename-blind by design (blobs stream before commits), the filter mirrors ITK's pre-commit excludes via content sniffing. For typical module ingests, the upstream whitelist has already pruned ThirdParty/, Data/, build/, and pixi/ paths, so the exclude set narrows to a handful of content-detectable extensions.

ITK pre-commit hook	v4 coverage
`clang-format` (`.c .cc .cxx .h .hxx`)	Sniff: `#include`, `#pragma`, `namespace itk`, `template <`, ... in first 8K bytes
`black` (`.py`)	Sniff: shebang or top-level `import`/`from`/`def`/`class`
`gersemi` (`.cmake .wrap CMakeLists.txt`)	Sniff: `cmake_minimum_required`, `project(`, `add_library(`, ITK macros (`itk_module_`, `itk_add_test`, `itk_wrap_`)
`trailing-whitespace`	Universal text fixer (every text-classified blob)
`end-of-file-fixer`	Universal text fixer
`mixed-line-ending` (CRLF→LF)	Universal text fixer
`check-shebang-scripts-are-executable`	Not implemented — post-merge gate catches it

ITK ghostflow rule	v4 coverage
Subject line ≤ 78 chars	`truncate_subject_to_body()` — excess moves to commit body, preferring word-break
Subject followed by blank line	`ensure_blank_line_after_subject()` — inserts `\n` if missing
No leftover merge-conflict markers	Deny-pass strips `.orig`, `.rej`, `.BACKUP.`, `.LOCAL.`, `.REMOTE.`, `.BASE.` from all history
No exec-bit on text files	`commit_callback` walks `file_changes` and clears mode `100755 → 100644` for known text extensions
No CI scaffolding under module subtree	Deny-pass strips `Dockerfile`, `.github/`, `.travis.yml`, `azure-pipelines*`, `circle.yml`, `.gitlab-ci.yml`, `appveyor.yml`, `pyproject.toml`, `setup.py`, etc.
Root commit	Unfixable — every upstream remote module has one; surfaces as the single remaining ghostflow error

Heuristic commit-prefix mapping

When a historical commit subject does not start with a recognized ITK prefix (BUG:/COMP:/DOC:/ENH:/PERF:/STYLE:/WIP:), sanitize-history.py auto-prepends one based on the table below. Every decision is logged with <sha> <chosen-prefix> <original-subject> so the operator can audit before merge.

Subject contains...	Prefix
`fix`, `bug`, `crash`, `segfault`, `leak`, `regression`, `corrupt`, `deadlock`	`BUG:`
`cmake`, `compil`, `warning`, `build`, `ci` / `ci:`, `clang` , `gcc`, `msvc`	`COMP:`
`doc`, `readme`, `comment`, `license`, `sphinx`, `doxygen`	`DOC:`
`style`, `whitespace`, `format`, `rename`, `lint`, `clean up`, `reorder`	`STYLE:`
`perf`, `optim`, `speed`, `faster`, `efficien`	`PERF:`
anything else	`ENH:` (default)

These rules are conservative — when in doubt, default to ENH:. Misclassifications get caught by the operator reading the sidecar log.

Smoke-test 1: IOMeshSTL (--dry-run)

Phase A — Ingestion (LOCAL, throwaway)
  Module:        IOMeshSTL
  DestGroup:     IO

After filter-repo: 91 commits, 24 merges
sanitize-history complete:
  commits walked:        97
  prefix auto-prepended: 42
  blobs scanned:         149
  blobs reformatted:     108
  by-kind sniff:         cxx=106 py=0 cmake=28 text=9 skip=6

Final rewritten history:
  Commits:  143 -> 89
  Merges:   44 -> 24
Verifying every commit subject has a valid ITK prefix...
  OK: every commit has a valid prefix.

Topology preserved (44→24 merges; non-zero, no linearization).

Smoke-test 2: Cuberille (real ingest, PR #6205) — proves ghostflow conformance

PR #6205 stacks on this PR and is the first real consumer. Ghostflow on PR #6205 reports exactly one error:

Errors:
  - commit 49774c0545... not allowed; it is a root commit.

That's the only remaining failure mode — the upstream module's genesis commit. All four other ghostflow categories that originally fired on Cuberille (long subjects on 6 commits, conflict markers in .h.orig on commit d23f9207, exec-bit on CMakeLists.txt in the root commit, two-line subject on 6127d75f41) are gone after this PR's hardening landed.

Sanitize-history.py counters from the Cuberille run that demonstrate the hardening is doing real work:

  prefix auto-prepended: 36
  subjects truncated:    5
  blank lines inserted:  33
  exec-bits cleared:     1
  blobs scanned:         147
  blobs reformatted:     94
  by-kind sniff:         cxx=102 py=1 cmake=28 text=5 skip=11

Build + 25/25 Cuberille tests pass locally.

Files added

Utilities/Maintenance/RemoteModuleIngest/INGESTION_STRATEGY_v4.md — design document
Utilities/Maintenance/RemoteModuleIngest/ingest-module-v4.sh — Phase A driver
Utilities/Maintenance/RemoteModuleIngest/archive-remote-module.sh — Phase B driver
Utilities/Maintenance/RemoteModuleIngest/sanitize-history.py — git_filter_repo callback driver
Utilities/Maintenance/RemoteModuleIngest/whitelists/IOMeshSTL.list — first concrete whitelist (smoke-test target)

The v3 scripts (ingest-remote-module.sh, normalize-ingest-commits.py, etc.) are intentionally retained in tree until v4 has been used for two successful ingests.

Open questions for reviewers

Gersemi version drift. Local install is 0.24.0; ITK pre-commit pins 0.19.3. Confirmed in practice on Cuberille: the post-merge pre-commit run --all-files gate catches one CMake file per ingest where the two versions disagree, fixed via a STYLE: follow-up commit. Acceptable, or should v4 verify version match before running?
Subject truncation can fall mid-word when a long word starts before offset 40 (e.g. STYLE: Merge pull request #21 from jhlegarreta/ImproveCuberilleImageToMeshFilt lost the erStyle suffix). The full original subject is preserved in the body, but the visible subject is uglier. Reviewer preference for raising the word-break minimum, or accept ghostflow-clean-but-cosmetically-truncated?
check-shebang-scripts-are-executable is currently deferred. Acceptable post-merge gate, or worth adding now?

Splits the v3 single-driver flow into two independent phases so the upstream-archival publish step never runs on a working tree that filter-repo has rewritten: - Phase A (ingest-module-v4.sh): local, throwaway clone; rewrites history into Modules/<Group>/<Module>; never touches the upstream remote. - Phase B (archive-remote-module.sh): fresh shallow clone; pushes one removal commit + promotes MIGRATION_README.md to README.md; flips archived=true. Gated on the Phase A PR being merged. The shared sanitize-history.py walks the rewritten branch and applies the same hooks ITK's .pre-commit-config.yaml enforces: - clang-format on every C/C++ blob (content sniff) - black on every Python blob (content sniff) - gersemi on every CMake blob using ITK's .gersemi.config - trailing-whitespace, end-of-file-fixer, mixed-line-ending on every text-classified blob - VTK / SVG / hash-content sidecar files are skipped intact - Every commit subject without an ITK prefix gets one heuristically auto-prepended; decisions are logged for operator audit filter-repo's blob_callback is filename-blind by design, so excludes are approximated via content sniffing. IOMeshSTL whitelist is the first concrete example; smoke-tested via --dry-run successfully (143->89 commits, 44->24 merges preserved, 108/149 blobs reformatted). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

The whitelist admits whole directories (test/, wrapping/, ...) but some upstream remote modules place CI / packaging scaffolding inside those (e.g. test/Docker/, .github/, azure-pipelines/). Without a deny-pass after the whitelist filter, those leak into ITK history. Mirrors v3's --invert-paths --path-glob pass; discovered missing on the first Cuberille v4 ingest (test/Docker/{Dockerfile,build,run,test}.sh leaked).

dzenanz · 2026-05-04T18:39:04Z

@@ -0,0 +1,25 @@
+# IOMeshSTL — files that migrate into ITK at Modules/IO/MeshSTL/


Is there reason this list is kept per remote?

dzenanz · 2026-05-04T18:45:16Z

@@ -0,0 +1,539 @@
+#!/usr/bin/env python3
+"""sanitize-history.py — Apply ITK formatting and commit-prefix conventions


It feels like "sanitize history" script already exists somewhere.

Fixes the four classes of ghostflow errors that surfaced on the first Cuberille v4 ingest (PR InsightSoftwareConsortium#6205) — the only allowable remaining error should be the unavoidable upstream root commit. - Strip *.orig / *.rej / *.BACKUP.* / *.LOCAL.* / *.REMOTE.* / *.BASE.* merge-conflict-artifact files from history (catches leftover-conflict-marker errors like the .h.orig on Cuberille). - sanitize-history.py: cap every commit subject at 78 chars (ghostflow / kw-commit-msg rule); excess words move into the body. - sanitize-history.py: insert a blank line between subject and body when the original commit message lacked one (two-line subject error). - sanitize-history.py: clear the executable bit (mode 100755 -> 100644) on text-file extensions so root-commit "executable permissions but file does not look executable" stops firing.

Many ITK remote modules' itk-module.cmake reads file(READ "${MY_CURRENT_DIR}/README.rst" DOCUMENTATION) to populate the CMake DESCRIPTION variable from the upstream README. The v4 whitelist intentionally excludes the upstream README (the operator ships a new README.md as part of the ingest PR), so without this rewrite every commit's CMakeLists fails to configure post-ingest until the operator manually patches itk-module.cmake. sanitize-history.py now rewrites the file(READ ... README.rst ...) reference to README.md as part of the cmake-blob transform, so every historical commit on the rewritten branch is independently buildable. Surfaced on the IOMeshSTL ingest (PR following PR InsightSoftwareConsortium#6204).

…terns Add a 'Known artifacts at PR-review time' section to capture two things future ingest operators need to expect but that v3/v4 do not fix automatically: - The single ghostflow-check-main 'root commit not allowed' error that every Mode-A merge produces. Maintainers override at merge. Calling it out keeps operators from chasing a non-fix. - A short table of code-level patterns Greptile has flagged post-ingest on multiple modules (IOMeshSTL InsightSoftwareConsortium#6206, RLEImage InsightSoftwareConsortium#6208): signed-vs-unsigned size types, missing override, dead return after itkExceptionMacro, stray test args, external-friend declarations, GetBuffer-const correctness, and allocator-init invariants. The table is the durable channel for cross-ingest review knowledge — extend it as new recurring patterns surface.

@dzenanz

ExternalData stores resolved fetched content as .ExternalData_<algo>_<hash> files alongside the .cid / .md5 sidecars they correspond to. These are local fetch cache, not source — ExternalData regenerates them on demand at consumer-time. Upstream remote modules sometimes commit them inadvertently from a CTest run inside their working tree (e.g. ITKIOMeshSTL had four such files in test/Baseline/, flagged by @dzenanz on PR InsightSoftwareConsortium#6206). Add **/.ExternalData_* to the scaffolding deny-pass so they never enter ITK history.

@dzenanz

Upstream remote modules wrap their top-level CMakeLists.txt in if(NOT ITK_SOURCE_DIR) find_package(ITK REQUIRED) list(APPEND CMAKE_MODULE_PATH ${ITK_CMAKE_DIR}) include(ITKModuleExternal) else() itk_module_impl() endif() so the same file works both as a fetched remote module and as a stand-alone CMake project. In an in-tree ingested module ITK_SOURCE_DIR is always defined, so the if-branch is dead code; @dzenanz flagged it on PR InsightSoftwareConsortium#6206 (IOMeshSTL). Add patch_standalone_build_guard() that detects the idiom (allowing 2- or 4-space indent variants) and replaces it with a bare itk_module_impl() before gersemi formatting runs. Track count via sanitizer.standalone_guard_patches and print it in the run summary.

The 'Greptile post-ingest patterns' table mixed mechanical artifacts (strippable in Phase A) with semantic concerns (require human judgment). Split into two sections so the operator's checklist shrinks as the sanitizer learns more rules. Promote .ExternalData_* and the standalone-build CMake guard to the auto-sanitized table now that the previous two commits handle them.

@dzenanz

Upstream remote modules typically begin their top-level CMakeLists.txt with their own cmake_minimum_required(VERSION X.Y.Z) declaration, often 3.10.2 or earlier. ITK's top-level CMakeLists pins a higher minimum, so the per-module line is redundant and frequently *lower* than ITK's floor. @dzenanz flagged it on every ingest with the comment 'Version behind, not needed.' — most recently on PR InsightSoftwareConsortium#6215 (IOFDF). Add patch_drop_cmake_minimum_required() that detects the idiom and removes the matching line via a multiline regex. Track count via sanitizer.cmake_min_required_drops and print it in the run summary. Apply alongside patch_standalone_build_guard() in the cmake-blob branch of blob_callback.

…ze table

….txt In-tree, ITK's top-level CMakeLists pins the CMake minimum, so the per-module declaration is dead code. Pre-emptive cleanup matching the v4 ingest-pipeline rule codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required).

…mage In-tree, ITK_SOURCE_DIR is always defined and ITK's top-level CMakeLists pins the CMake minimum version. The per-module cmake_minimum_required line and if(NOT ITK_SOURCE_DIR) ... else() itk_module_impl() endif() guard are dead code. Pre-emptive cleanup matching the v4 ingest-pipeline rules codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard).

In-tree, ITK's top-level CMakeLists pins the CMake minimum, so the per-module declaration is dead code. Matches the v4 ingest-pipeline rule in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required).

…shMZ3 In-tree, ITK_SOURCE_DIR is always defined and ITK's top-level CMakeLists pins the CMake minimum version. The per-module cmake_minimum_required line and if(NOT ITK_SOURCE_DIR) ... else() itk_module_impl() endif() guard are dead code. The 'set(ITK_DIR ${CMAKE_BINARY_DIR})' line that was inside the in-tree else branch is preserved (now unconditional) since it actually had effect during the in-tree build path. Matches the v4 ingest-pipeline rules in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard).

Four already-merged remote-module ingests still carry their upstream-original boilerplate at the top of their module CMakeLists: - cmake_minimum_required(VERSION X.Y.Z) — redundant; ITK's top-level CMakeLists pins a higher minimum. - if(NOT ITK_SOURCE_DIR) ... find_package(ITK) ... else() itk_module_impl() endif() — dead code in-tree because ITK_SOURCE_DIR is always defined. Strip both from: Modules/Filtering/AnisotropicDiffusionLBR Modules/Filtering/Cuberille Modules/Filtering/LabelErodeDilate Modules/Registration/Montage Lines that had effect inside the in-tree else() branch (set(ITK_DIR ${CMAKE_BINARY_DIR}) on AnisotropicDiffusionLBR and Montage) are preserved unconditionally. The other four merged ingests (GenericLabelInterpolator, MGHIO, FastBilateral, MeshNoise) were already cleaned up during their respective ingest PRs. Matches the v4 ingest-pipeline rules now codified in PR #6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard); future ingests will not introduce this idiom.

…mage In-tree, ITK_SOURCE_DIR is always defined and ITK's top-level CMakeLists pins the CMake minimum version. The per-module cmake_minimum_required line and if(NOT ITK_SOURCE_DIR) ... else() itk_module_impl() endif() guard are dead code. Pre-emptive cleanup matching the v4 ingest-pipeline rules codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard).

In-tree, ITK's top-level CMakeLists pins the CMake minimum, so the per-module declaration is dead code. Matches the v4 ingest-pipeline rule in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required).

Four already-merged remote-module ingests still carry their upstream-original boilerplate at the top of their module CMakeLists: - cmake_minimum_required(VERSION X.Y.Z) — redundant; ITK's top-level CMakeLists pins a higher minimum. - if(NOT ITK_SOURCE_DIR) ... find_package(ITK) ... else() itk_module_impl() endif() — dead code in-tree because ITK_SOURCE_DIR is always defined. Strip both from: Modules/Filtering/AnisotropicDiffusionLBR Modules/Filtering/Cuberille Modules/Filtering/LabelErodeDilate Modules/Registration/Montage Lines that had effect inside the in-tree else() branch (set(ITK_DIR ${CMAKE_BINARY_DIR}) on AnisotropicDiffusionLBR and Montage) are preserved unconditionally. The other four merged ingests (GenericLabelInterpolator, MGHIO, FastBilateral, MeshNoise) were already cleaned up during their respective ingest PRs. Matches the v4 ingest-pipeline rules now codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard); future ingests will not introduce this idiom.

….txt In-tree, ITK's top-level CMakeLists pins the CMake minimum, so the per-module declaration is dead code. Pre-emptive cleanup matching the v4 ingest-pipeline rule codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required).

…shMZ3 In-tree, ITK_SOURCE_DIR is always defined and ITK's top-level CMakeLists pins the CMake minimum version. The per-module cmake_minimum_required line and if(NOT ITK_SOURCE_DIR) ... else() itk_module_impl() endif() guard are dead code. The 'set(ITK_DIR ${CMAKE_BINARY_DIR})' line that was inside the in-tree else branch is preserved (now unconditional) since it actually had effect during the in-tree build path. Matches the v4 ingest-pipeline rules in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard).

…mage In-tree, ITK_SOURCE_DIR is always defined and ITK's top-level CMakeLists pins the CMake minimum version. The per-module cmake_minimum_required line and if(NOT ITK_SOURCE_DIR) ... else() itk_module_impl() endif() guard are dead code. Pre-emptive cleanup matching the v4 ingest-pipeline rules codified in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required and :patch_standalone_build_guard).

In-tree, ITK's top-level CMakeLists pins the CMake minimum, so the per-module declaration is dead code. Matches the v4 ingest-pipeline rule in PR InsightSoftwareConsortium#6204 (sanitize-history.py:patch_drop_cmake_minimum_required).

…cmake Archival README.md files contain semicolons and bracket characters. CMake's foreach(arg \${ARGN}) splits on semicolons, so reading them into a DESCRIPTION argument produces CMake (dev) configure warnings for every module ingested via v4 (observed: RLEImage, SplitComponents, IOFDF, IOMeshMZ3, IOMeshSTL). Add patch_dynamic_description() to sanitize-history.py to strip the get_filename_component/file(READ README.md) preamble and replace DESCRIPTION "\${DOCUMENTATION}" with a static one-liner during history rewrite. Document the pattern in INGESTION_STRATEGY_v4.md.

Many ITK remote modules' itk-module.cmake reads file(READ "${MY_CURRENT_DIR}/README.rst" DOCUMENTATION) to populate the CMake DESCRIPTION variable from the upstream README. The v4 whitelist intentionally excludes the upstream README (the operator ships a new README.md as part of the ingest PR), so without this rewrite every commit's CMakeLists fails to configure post-ingest until the operator manually patches itk-module.cmake. sanitize-history.py now rewrites the file(READ ... README.rst ...) reference to README.md as part of the cmake-blob transform, so every historical commit on the rewritten branch is independently buildable. Surfaced on the IOMeshSTL ingest (PR following PR InsightSoftwareConsortium#6204).

dzenanz

As this is all in the maintenance directory, I believe we can accept it without really reviewing it.

…shFilter ENH: Ingest ITKSubdivisionQuadEdgeMeshFilter into Modules/Filtering (v4, stacks on #6204)

github-actions Bot added type:Infrastructure Infrastructure/ecosystem related changes, such as CMake or buildbots area:Python wrapping Python bindings for a class labels May 4, 2026

dzenanz reviewed May 4, 2026

View reviewed changes

hjmjohnson mentioned this pull request May 4, 2026

ENH: Ingest ITKCuberille into Modules/Filtering #6205

Merged

hjmjohnson added 2 commits May 4, 2026 14:05

This was referenced May 4, 2026

ENH: Ingest ITKIOMeshSTL into Modules/IO #6206

Merged

ENH: Ingest ITKRLEImage into Modules/Filtering #6208

Merged

ENH: Ingest ITKMorphologicalContourInterpolation (stacked on #6208) #6209

Merged

hjmjohnson added 4 commits May 5, 2026 05:51

hjmjohnson mentioned this pull request May 5, 2026

ENH: Master tracking — Remote-module ingestion, archival, and consolidation #6160

Open

50 tasks

ENH: Add v4 PolarTransform whitelist (header-only, mirrors Cuberille)

cea536f

hjmjohnson mentioned this pull request May 5, 2026

ENH: Ingest ITKPolarTransform into Modules/Filtering #6211

Merged

ENH: Add v4 SplitComponents whitelist (header-only, mirrors Cuberille)

bd785b0

hjmjohnson mentioned this pull request May 5, 2026

ENH: Ingest ITKSplitComponents into Modules/Filtering #6212

Merged

hjmjohnson added 4 commits May 5, 2026 08:45

ENH: Add v4 IOMeshMZ3 whitelist (mirrors IOMeshSTL — has include + src)

b91669e

ENH: Add v4 IOFDF whitelist (mirrors IOMeshSTL — has include + src)

e9e5c5a

DOC: v4 strategy — add cmake_minimum_required-stripper to auto-saniti…

91976ca

…ze table

hjmjohnson added this to the ITK 6.0 Release Candidate 1 milestone May 5, 2026

hjmjohnson mentioned this pull request May 5, 2026

COMP: Strip dead cmake_minimum_required + standalone wrapper from 4 ingests #6216

Merged

hjmjohnson mentioned this pull request May 6, 2026

BUG: Replace multi-line DESCRIPTION strings in 5 ingested itk-module.cmake #6220

Merged

This was referenced May 6, 2026

COMP: Add patch_dynamic_description() to v4 ingest sanitizer #6221

Closed

ENH: Ingest ITKSubdivisionQuadEdgeMeshFilter into Modules/Filtering (v4, stacks on #6204) #6229

Merged

dzenanz approved these changes May 7, 2026

View reviewed changes

hjmjohnson added a commit that referenced this pull request May 7, 2026

Merge pull request #6229 from hjmjohnson/ingest-SubdivisionQuadEdgeMe…

e18a952

…shFilter ENH: Ingest ITKSubdivisionQuadEdgeMeshFilter into Modules/Filtering (v4, stacks on #6204)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

WIP: Add v4 remote-module ingestion (two-phase, per-commit hooks)#6204

WIP: Add v4 remote-module ingestion (two-phase, per-commit hooks)#6204
hjmjohnson wants to merge 15 commits intoInsightSoftwareConsortium:mainfrom
hjmjohnson:ingest-strategy-v4

hjmjohnson commented May 4, 2026 •

edited

Loading

Uh oh!

dzenanz May 4, 2026

Uh oh!

dzenanz May 4, 2026

Uh oh!

dzenanz left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		@@ -0,0 +1,25 @@
		# IOMeshSTL — files that migrate into ITK at Modules/IO/MeshSTL/

		@@ -0,0 +1,539 @@
		#!/usr/bin/env python3
		"""sanitize-history.py — Apply ITK formatting and commit-prefix conventions

Uh oh!

Conversation

hjmjohnson commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dzenanz May 4, 2026

Choose a reason for hiding this comment

Uh oh!

dzenanz May 4, 2026

Choose a reason for hiding this comment

Uh oh!

dzenanz left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

hjmjohnson commented May 4, 2026 •

edited

Loading