chore(local): add S3 migration and CRAN-source helper scripts

- migrate-s3-hetzner-to-backblaze.sh: rclone-based bucket migration helper.
- find-R-api-packages.sh: scan CRAN package sources for a C-API usage pattern.
- query-pkgs-without-old-versions.R, test-package-loading.R: ad-hoc helpers.
This commit is contained in:
Patrick Schratz 2026-06-14 13:02:48 +02:00
commit a9ad907f9c
Signed by: pat-s
GPG key ID: 3C6318841EF78925
4 changed files with 494 additions and 0 deletions

View file

@ -0,0 +1,34 @@
#!/bin/bash
# Directory to clone repos into
WORKDIR="cran_repos"
mkdir -p "$WORKDIR"
cd "$WORKDIR"
# GitHub API paginates results, so we loop through pages
PAGE=1
PER_PAGE=100
MATCHES=()
while :; do
# Fetch a page of repos
REPOS=$(curl -s "https://api.github.com/orgs/cran/repos?per_page=$PER_PAGE&page=$PAGE" | jq -r '.[].clone_url')
[ -z "$REPOS" ] && break
for REPO_URL in $REPOS; do
REPO_NAME=$(basename "$REPO_URL" .git)
# Skip if already cloned
[ -d "$REPO_NAME" ] && continue
git clone --depth 1 "$REPO_URL" "$REPO_NAME" >/dev/null 2>&1
if [ -d "$REPO_NAME/src" ]; then
# Search for Rinternals.h in src/
if grep -r -q 'R_VERSION < R_Version(' "$REPO_NAME/src"; then
echo "$REPO_NAME"
fi
fi
# Clean up to save space
rm -rf "$REPO_NAME"
done
PAGE=$((PAGE + 1))
done

View file

@ -0,0 +1,90 @@
#!/usr/bin/env bash
set -euo pipefail
# Migrate S3 buckets from Hetzner Object Storage to Backblaze B2 via rclone.
#
# Prerequisites:
# 1. Install rclone: https://rclone.org/install/
# 2. Configure two rclone remotes:
# rclone config create hetzner s3 \
# provider=Other \
# env_auth=false \
# access_key_id=YOUR_HETZNER_KEY \
# secret_access_key=YOUR_HETZNER_SECRET \
# endpoint=fsn1.your-objectstorage.com # adjust region
#
# rclone config create backblaze s3 \
# provider=Other \
# env_auth=false \
# access_key_id=YOUR_B2_KEY \
# secret_access_key=YOUR_B2_APP_KEY \
# endpoint=s3.us-west-004.backblazeb2.com # adjust region
#
# Usage:
# ./migrate-s3-hetzner-to-backblaze.sh <src:dst> [src:dst] ...
# ./migrate-s3-hetzner-to-backblaze.sh hetzner-bucket:backblaze-bucket
HETZNER_REMOTE="${HETZNER_REMOTE:-hetzner}"
BACKBLAZE_REMOTE="${BACKBLAZE_REMOTE:-backblaze}"
RCLONE_FLAGS="${RCLONE_FLAGS:---transfers=64 --checkers=64 --fast-list}"
if [[ $# -eq 0 ]]; then
echo "Usage: $0 <src-bucket:dst-bucket> [src-bucket:dst-bucket...]"
echo ""
echo " Each argument is a source:destination bucket pair separated by a colon."
echo ""
echo "Environment variables:"
echo " HETZNER_REMOTE rclone remote name for Hetzner (default: hetzner)"
echo " BACKBLAZE_REMOTE rclone remote name for Backblaze (default: backblaze)"
echo " RCLONE_FLAGS extra rclone flags (default: --transfers=16 --checkers=16 --fast-list)"
echo " DRY_RUN=1 show what would be copied without copying"
exit 1
fi
for cmd in rclone; do
if ! command -v "$cmd" &>/dev/null; then
echo "Error: $cmd is not installed." >&2
exit 1
fi
done
# Verify remotes exist
for remote in "$HETZNER_REMOTE" "$BACKBLAZE_REMOTE"; do
if ! rclone listremotes | grep -q "^${remote}:$"; then
echo "Error: rclone remote '${remote}' not found. Run 'rclone config' to set it up." >&2
exit 1
fi
done
DRY_RUN_FLAG=""
if [[ "${DRY_RUN:-0}" == "1" ]]; then
DRY_RUN_FLAG="--dry-run"
echo "=== DRY RUN MODE ==="
fi
for pair in "$@"; do
if [[ "$pair" != *:* ]]; then
echo "Error: '$pair' is not a valid src:dst pair. Use format 'hetzner-bucket:backblaze-bucket'." >&2
exit 1
fi
src_bucket="${pair%%:*}"
dst_bucket="${pair#*:}"
src="${HETZNER_REMOTE}:${src_bucket}"
dst="${BACKBLAZE_REMOTE}:${dst_bucket}"
echo ""
echo "--- Migrating: ${src} -> ${dst} ---"
# shellcheck disable=SC2086
rclone sync \
${RCLONE_FLAGS} \
${DRY_RUN_FLAG} \
--progress \
"$src" "$dst"
echo "--- Done: ${src_bucket} -> ${dst_bucket} ---"
done
echo ""
echo "Migration complete."

View file

@ -0,0 +1,39 @@
library(s3fs)
# List all files under contrib/<pkg>/
all_files <- s3fs::s3_dir_ls(
"s3://devxy-r-package-binaries-hel1/arm64/alpine322/latest/src/contrib/",
recurse = TRUE,
type = "file"
)
pattern <- ".*/src/contrib/([^/_]+)_.*"
matches <- regmatches(all_files, regexec(pattern, all_files))
pkg_names <- unique(
vapply(
matches,
function(x) if (length(x) > 1) x[2] else NA_character_,
character(1)
)
)
pkg_names <- pkg_names[!is.na(pkg_names)]
# For each package, check if Archive/<pkg>/ contains any files
no_archive_files <- character(0)
for (pkg in pkg_names) {
archive_dir1 <- sprintf(
"s3://devxy-r-package-binaries-hel1/arm64/alpine322/latest/src/contrib/Archive/%s",
pkg
)
archive_files <- unique(c(
tryCatch(
s3fs::s3_dir_ls(archive_dir1, recurse = TRUE),
error = function(e) character(0)
)
))
if (length(archive_files) == 0) {
no_archive_files <- c(no_archive_files, pkg)
}
}
print(no_archive_files)

View file

@ -0,0 +1,331 @@
install.packages(
"pak",
repos = sprintf(
"https://r-lib.github.io/p/pak/stable/%s/%s/%s",
.Platform$pkgType,
R.Version()$os,
R.Version()$arch
)
)
Sys.setenv(PKG_SYSREQS = TRUE)
all_pkgs <- rownames(available.packages())
to_skip = c("ABRSQOL", "ACA", "ACE.CoCo")
all_pkgs = setdiff(all_pkgs, to_skip)
for (i in all_pkgs) {
message(sprintf("\nInstalling %s", i))
pak::pkg_install(i)
library(i, character.only = TRUE)
}
# Example data
all_pkgs <- rownames(available.packages())
to_skip <- c(
"ABRSQOL",
"ACA",
"ACE.CoCo",
"ACEsimFit",
"ACNE",
"absorber",
"adapt4pv",
"adaptMCMC",
"addhaz",
"ADDT",
"ahaz",
"arm",
"arules",
"arulesCBA",
"aster2",
"BayesFactor",
"bc3net",
"bgsmtr",
"biglasso",
"BinNonNor",
"BinNor",
"bioassayR",
"birankr",
"BiRewire",
"bolasso",
"Boptbd",
"Brobdingnag",
"BSW",
"BTLLasso",
"bvartools",
"cAIC4",
"Category",
"celda",
"centiserve",
"cjoint",
"clinical",
"clipper",
"CodataGS",
"conos",
"CopulaInference",
"covEB",
"cplm",
"CRTgeeDR",
"cthreshER",
"ctmcmove",
"curephEM",
"CVST",
"dcGSA",
"dclone",
"dcsvm",
"DelayedArray",
"dglars",
"dhglm",
"disordR",
"distrom",
"dmm",
"DNABarcodes",
"DoubleCone",
"DRR",
"DTRlearn2",
"DWDLargeR",
"eds",
"EMCluster",
"EMMREML",
"evalITR",
"EventPointer",
"evola",
"excursions",
"expm",
"fanc",
"FAS",
"fastadi",
"fastPLS",
"fastRG",
"fdaPDE",
"flare",
"FoReco",
"frailtyHL",
"freebird",
"FSTpackage",
"gamlr",
"gamlss.lasso",
"gamm4",
"gbmt",
"gdim",
"gdistance",
"GeDS",
"geeM",
"genlasso",
"GenOrd",
"GenoScan",
"geomorph",
"geostatsp",
"GhostKnockoff",
"GIGSEA",
"GlarmaVarSel",
"glmm",
"glmmrBase",
"glmmrOptim",
"glmnet",
"glober",
"GPvam",
"graphpcor",
"gremlin",
"growthrate",
"grpCox",
"GSD",
"HelpersMG",
"hglm",
"hglm.data",
"hibayes",
"hierSDR",
"HMTL",
"hsem",
"ibmdbR",
"inca",
"INLAspacetime",
"INLAtools",
"invertiforms",
"irlba",
"islasso",
"ISLET",
"isotonic.pen",
"jordan",
"kinship2",
"KnockoffScreen",
"lcpm",
"leidenAlg",
"lfe",
"lingmatch",
"LKT",
"lme4",
"lme4breeding",
"lme4GS",
"logcondiscr",
"LPmerge",
"LRMF3",
"MAP",
"marcox",
"markovchain",
"MatrixExtra",
"matter",
"MBC",
"mcen",
"mclogit",
"MCMCglmm",
"mdhglm",
"MDPtoolbox",
"mediation",
"mefa4",
"metafor",
"mgwrsar",
"mi",
"midasml",
"mind",
"monocle",
"msda",
"MuData",
"MultiGlarmaVarSel",
"MultiOrd",
"MultiVarSel",
"mvglmmRank",
"N2R",
"nadiv",
"NBtsVarSel",
"NegBinBetaBinreg",
"NetworkRiskMeasures",
"neuroim2",
"NOISeq",
"numbat",
"optbdmaeAT",
"optimbase",
"OptimModel",
"optrcdmaeAT",
"OrdNor",
"pagoda2",
"PCovR",
"pedgene",
"pedigree",
"pedigreemm",
"pense",
"PERMANOVA",
"phateR",
"PhylogeneticEM",
"pleio",
"POINT",
"PoisBinNonNor",
"PoisBinOrd",
"PoisBinOrdNonNor",
"PoisBinOrdNor",
"PoisNonNor",
"PoisNor",
"PRISMA",
"ProbitSpatial",
"prodest",
"psqn",
"qlcMatrix",
"qpcR",
"QRM",
"quadrupen",
"QZ",
"ramps",
"randnet",
"randPedPCA",
"rBMF",
"RCBR",
"RealVAMS",
"REBayes",
"recommenderlab",
"Rediscover",
"reglogit",
"RESET",
"RGE",
"RGENERATEPREC",
"riemtan",
"RNewsflow",
"robustlmm",
"rsparse",
"rSPDE",
"rwc",
"S4Arrays",
"saeMSPE",
"sbw",
"scITD",
"scoup",
"sdwd",
"SEAGLE",
"sensory",
"serrsBayes",
"sglasso",
"sharpPen",
"SiPSiC",
"SKAT",
"snpReady",
"snpStats",
"softImpute",
"sommer",
"soptdmaeA",
"SOR",
"SparseArray",
"SparseChol",
"sparseLRMatrix",
"sparsenet",
"sparsenetgls",
"sparsestep",
"spatialprobit",
"spatialreg",
"spatstat.sparse",
"speedglm",
"sRDA",
"sSDR",
"ssfa",
"stcos",
"StratifiedSampling",
"sureLDA",
"survey",
"surveyvoi",
"svydiags",
"systemfit",
"TargetScore",
"text2map",
"textir",
"textmineR",
"textTinyR",
"tmvtnorm",
"TPEA",
"triversity",
"tsapp",
"tvReg",
"uwot",
"vagam",
"VAM",
"WaveSampling",
"WGScan",
"wordspace",
"workflowsets",
"ACSSpack",
"ADDT",
"AER"
)
# Find the position of the last package in to_skip within all_pkgs
last_skip <- tail(to_skip, 1)
# Find its position in all_pkgs (returns NA if not found)
start_pos <- match(last_skip, all_pkgs)
# If not found, start from the beginning; else, start after last_skip
if (is.na(start_pos)) {
to_process <- all_pkgs
} else {
to_process <- all_pkgs[(start_pos + 1):length(all_pkgs)]
}
if (length(to_process) == 0) {
message("All packages have been processed.")
} else {
for (i in to_process) {
message(sprintf("\nInstalling %s", i))
pak::pkg_install(i)
library(i, character.only = TRUE)
}
# Update to_skip to include all up to the last processed
to_skip <- all_pkgs[1:(last_skip_pos + length(to_process))]
}