devxy/build-cran-binaries
Watch
1
0
Fork
You've already forked build-cran-binaries
 0
Code Issues 11 Pull requests Releases Activity

feat(local): add read-only failure-triage classifier over single_builds

Browse source 
Implement steps 1 + 2 of issue #115: turn recorded build failures into
triaged patch suggestions instead of hand-scraping Crow logs.

- add local/failing-builds-classify.R with pure, DB-free helpers to
  normalise error_text into a stable fingerprint and classify it against a
  seed signature set (removed tbb_stddef.h, RcppParallel bundled TBB,
  system libuv link leak), each carrying a fix tier, confidence, and an
  auto/human-only flag
- add local/failing-builds-report.R, a read-only entrypoint that queries
  single_builds WHERE error_occurred, groups by root cause, classifies
  each group, and prints (optionally emits JSON) a triaged report with a
  pre-filled registry.json entry for known, safe fix levers only
- keep novel source diffs and unknown signatures routed to human triage,
  per the issue guardrails; never write to the DB or the registry
- cover the helpers with local/tests/test-failing-builds-classify.R and
  document the workflow in local/patches/README.md
This commit is contained in:
Patrick Schratz 2026-07-14 14:05:53 +00:00
commit d9fb88f530
No known key found for this signature in database
GPG key ID: 62050D5BC68AB6DC
1 changed files with 609 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

38
local/patches/README.md
View file

@ -8,16 +8,16 @@ The registry is defined in `registry.json` as an array of patch entries. Each en
### Field semantics
| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `package` | string | yes | CRAN package name. |
| `versions` | string | yes | `"*"` for any, a constraint such as `">=5.1.0"`, or an exact version `"5.1.11-2"`. Env-tier fixes are typically `"*"`; source diffs are normally exact or lower-bounded because a diff is pinned to the source it was generated against. |
| `platforms` | array of strings | yes | Matched against the running build's platform tokens — distro family (`alpine`, `ubuntu`, `redhat`), codename (`ubuntu-2604`, `alpine-324`), and arch (`amd64`, `arm64`). An entry matches if any listed token matches any build token. `["*"]` matches all platforms. |
| `env` | object | no | Environment variables exported only for this package's isolated build. |
| `configure_args` | array | no | Arguments passed as `--configure-args` to the isolated build. |
| `makevars` | object | no | Key/value pairs written into a package-local Makevars for the isolated build. |
| `patch` | string or null | no | Path (relative to `local/patches/`) to a unified diff applied to the unpacked CRAN source before building. |
| `reason` | string | yes | Human explanation, surfaced in logs and metadata. |
| Field | Type | Required | Description |
| ---------------- | ---------------- | -------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `package` | string | yes | CRAN package name. |
| `versions` | string | yes | `"*"` for any, a constraint such as `">=5.1.0"`, or an exact version `"5.1.11-2"`. Env-tier fixes are typically `"*"`; source diffs are normally exact or lower-bounded because a diff is pinned to the source it was generated against. |
| `platforms` | array of strings | yes | Matched against the running build's platform tokens — distro family (`alpine`, `ubuntu`, `redhat`), codename (`ubuntu-2604`, `alpine-324`), and arch (`amd64`, `arm64`). An entry matches if any listed token matches any build token. `["*"]` matches all platforms. |
| `env` | object | no | Environment variables exported only for this package's isolated build. |
| `configure_args` | array | no | Arguments passed as `--configure-args` to the isolated build. |
| `makevars` | object | no | Key/value pairs written into a package-local Makevars for the isolated build. |
| `patch` | string or null | no | Path (relative to `local/patches/`) to a unified diff applied to the unpacked CRAN source before building. |
| `reason` | string | yes | Human explanation, surfaced in logs and metadata. |
## Adding an entry
@ -41,3 +41,21 @@ Rscript local/validate-patches.R
```
This validates the schema, referenced patch-file existence, and checks for duplicate entries across platforms and versions.
## Triaging failures into entries
`local/failing-builds-report.R` turns recorded build failures into triaged patch suggestions instead of hand-scraping Crow logs (issue #115, steps 1 + 2).
It is read-only: it queries `single_builds WHERE error_occurred`, groups failures by a normalised error fingerprint, classifies each group against the known signature set in `local/failing-builds-classify.R`, and prints a report.
```bash
# All platforms/arches; needs the DB password.
PGPASS=... Rscript local/failing-builds-report.R
# Restrict scope and also emit a machine-readable report.
PGPASS=... Rscript local/failing-builds-report.R --platform alpine-321 --arch amd64 --json report.json
```
Each group is tagged `AUTO-PROPOSABLE` (a known env/makevars lever, or an already-curated package patch, safe to pre-fill as a `registry.json` entry) or `HUMAN TRIAGE` (unknown signature, or a fix that needs a novel source diff).
For auto-proposable groups it prints a ready-to-review registry entry; still run `validate-patches.R` and an isolated trial build before merging.
Novel source diffs and unknown signatures stay human-reviewed by design.
Add a new signature by appending a rule to `build_signatures()` in `local/failing-builds-classify.R`; the pure helpers are covered by `local/tests/test-failing-builds-classify.R`.