wip: migrate to gh actions

.github/workflows/build.yml

@@ -0,0 +1,44 @@
+name: Create and publish a Docker image
+
+on:
+  push:
+    branches: ['main']
+  workflow_dispatch:
+
+env:
+  REGISTRY: ghcr.io
+
+jobs:
+  build-and-push-image:
+    strategy:
+        matrix:
+          image: [datalad-apptainer, deface, dicom_indexer, heudiconv]
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      packages: write
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Log in to the Container registry
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ env.NIDATAOPS_BOT_NAME }}
+          password: ${{ secrets.NIDATAOPS_BOT_REGISTRY_TOKEN }}
+
+      - name: Extract metadata (tags, labels) for Docker
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.GITHUB_REPOSITORY_OWNER }}.${{ matrix.image }}
+
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v5
+        with:
+          context: docker/${{ matrix.image }}
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}

docker/datalad-apptainer/Dockerfile

@@ -31,12 +31,13 @@ COPY --from=builder /usr/local/apptainer /usr/local/apptainer
 ENV PATH="/usr/local/apptainer/bin:$PATH" \
     APPTAINER_TMPDIR="/tmp-apptainer"
 RUN apk add --no-cache py3-pytest ca-certificates libseccomp squashfs-tools tzdata fuse2fs fuse-overlayfs squashfuse \
-    python3 py3-pip git openssh-client git-annex curl bzip2 bash glab\
+    python3 py3-pip git openssh-client git-annex curl bzip2 bash glab jq\
     && mkdir -p $APPTAINER_TMPDIR \
     && cp /usr/share/zoneinfo/UTC /etc/localtime \
     && apk del tzdata \
     && rm -rf /tmp/* /var/cache/apk/*
 
 RUN pip install --break-system-packages --no-cache-dir datalad datalad-container ssh_agent_setup python-gitlab
+ADD cfg_nidataops.py /usr/lib/python3.11/site-packages/datalad/resources/procedures/
 
 WORKDIR /work

docker/datalad-apptainer/cfg_nidataops.py

@@ -0,0 +1,45 @@
+#!/usr/bin/env python3
+"""Procedure to configure Git annex to add text files directly to Git"""
+
+import sys
+import os.path as op
+
+from datalad.distribution.dataset import require_dataset
+
+ds = require_dataset(
+    sys.argv[1],
+    check_installed=True,
+    purpose='configuration')
+
+nthg = {'annex.largefiles': 'nothing'}
+anthg = {'annex.largefiles': 'anything'}
+annex_largefiles = '((mimeencoding=binary)and(largerthan=0))'
+attrs = ds.repo.get_gitattributes('*')
+if not attrs.get('*', {}).get(
+        'annex.largefiles', None) == annex_largefiles:
+    ds.repo.set_gitattributes([
+        ('*', {'annex.largefiles': annex_largefiles}),
+        ('.gitignore', nthg),
+        ('.gitmodules', nthg),
+        ('.gitlab-ci.yml', nthg),
+        ('.all-contributorsrc', nthg),
+        ('.bidsignore', nthg),
+        ('*.json', nthg),
+        ('*.txt', nthg),
+        ('*.tsv', nthg),
+        ('*.nii.gz', anthg),
+        ('*.tgz', anthg),
+        ('*_scans.tsv', anthg),
+        # annex event files as they contain subjects behavioral responses
+        ('sub-*/**/*_events.tsv', anthg),
+        ('*.bk2', anthg),
+        ('*.html', anthg),
+        ('*.svg', anthg),
+        ])
+
+git_attributes_file = op.join(ds.path, '.gitattributes')
+ds.save(
+    git_attributes_file,
+    message="Setup gitattributes for ni-dataops",
+    result_renderer='disabled'
+)

docker/dicom_indexer/entrypoint.sh

@@ -3,6 +3,9 @@
 export CONTAINER_ID=$(basename $(cat /proc/1/cpuset))
 GITLAB_TOKEN_SECRET=$(cat /var/run/secrets/dicom_bot_gitlab_token 2>/dev/null)
 export GITLAB_TOKEN=${GITLAB_TOKEN_SECRET:=$GITLAB_TOKEN}
+S3_ID=$(cat /var/run/secrets/s3_id 2>/dev/null)
+S3_SECRET=$(cat /var/run/secrets/s3_secret 2>/dev/null)
+export AWS_ACCESS_KEY_ID=${S3_ID:=$AWS_ACCESS_KEY_ID} AWS_SECRET_ACCESS_KEY=${S3_SECRET:=$AWS_SECRET_ACCESS_KEY}
 export GITLAB_API_URL=https://${CI_SERVER_HOST}/api/v4
 export GIT_SSH_PORT=${GIT_SSH_PORT:=222}
 
@@ -23,14 +26,6 @@ fi
 git config --global init.defaultBranch main
 ssh-keyscan -p ${GIT_SSH_PORT} -H ${CI_SERVER_HOST} | install -m 600 /dev/stdin $HOME/.ssh/known_hosts
 
-# example
-# /usr/bin/storescp \
-#  -aet DICOM_SERVER_SEQUOIA\
-#  -pm\
-#  -od $DICOM_TMP_DIR -su ''\
-#  --eostudy-timeout ${STORESCP_STUDY_TIMEOUT:=60} \
-#  --exec-on-eostudy "python3 $DICOM_ROOT/exec_on_study_received.py #p " 2100 >> $DICOM_DATA_ROOT/storescp.log
-
 # run whatever command was passed (storescp or python index_dicoms directly)
 $@
 

docker/dicom_indexer/indexer/index_dicom.py

@@ -1,4 +1,5 @@
 import os
+import time
 import pydicom as dicom
 import argparse
 import pathlib
@@ -26,6 +27,8 @@ GITLAB_TOKEN = os.environ.get("GITLAB_TOKEN", None)
 GITLAB_BOT_USERNAME = os.environ.get("GITLAB_BOT_USERNAME", None)
 GITLAB_BOT_EMAIL = os.environ.get("GITLAB_BOT_EMAIL", None)
 BIDS_DEV_BRANCH = os.environ.get("BIDS_DEV_BRANCH", "dev")
+BIDS_BASE_BRANCH = os.environ.get("BIDS_BASE_BRANCH", "base")
+BIDS_CONVERT_BRANCHES = os.environ.get("BIDS_CONVERT_BRANCHES", 'convert/*')
 NI_DATAOPS_GITLAB_ROOT = os.environ.get("NI_DATAOPS_GITLAB_ROOT", "ni-dataops")
 
 S3_REMOTE_DEFAULT_PARAMETERS = [
@@ -202,7 +205,7 @@ def index_dicoms(
         # cannot pass message above so commit now
         dicom_session_ds.save(message=f"index dicoms from archive {archive}")  #
         # optimize git index after large import
-        #dicom_session_ds.repo.gc()  # aggressive by default
+        # dicom_session_ds.repo.gc()  # aggressive by default
         yield dicom_session_ds
 
 
@@ -296,11 +299,11 @@ def setup_gitlab_repos(
             dicom_study_ds.create(force=True)
             # add default study DS structure.
             init_dicom_study(dicom_study_ds, gitlab_group_path)
-            # initialize BIDS project
-            init_bids(gitlab_conn, dicom_study_repo, gitlab_group_path)
             # create subgroup for QC and derivatives repos
             get_or_create_gitlab_group(gitlab_conn, gitlab_group_path / "derivatives")
             get_or_create_gitlab_group(gitlab_conn, gitlab_group_path / "qc")
+            # initialize BIDS project
+            init_bids(gitlab_conn, dicom_study_repo, gitlab_group_path)
 
         dicom_study_ds.install(
             source=dicom_session_repo._attrs["ssh_url_to_repo"],
@@ -334,11 +337,23 @@ def init_bids(
         )
         bids_project_ds.push(to="origin")
         # create dev branch and push for merge requests
-        bids_project_ds.repo.checkout(BIDS_DEV_BRANCH, ["-b"])
+        for branch in [BIDS_DEV_BRANCH, BIDS_BASE_BRANCH]:
-        bids_project_ds.push(to="origin")
+            bids_project_ds.repo.checkout(branch, ["-b"])
+            bids_project_ds.push(to="origin")
         # set protected branches
-        bids_project_repo.protectedbranches.create(data={"name": "convert/*"})
+        for branch in [BIDS_CONVERT_BRANCHES, BIDS_DEV_BRANCH, BIDS_BASE_BRANCH]:
-        bids_project_repo.protectedbranches.create(data={"name": "dev"})
+            bids_project_repo.protectedbranches.create(data={"name": branch})
+
+        ### avoid race conditions for first session pushed ###
+        ### otherwise heudiconv starts before the remotes are configured
+        time.sleep(5)  # wait for config pipeline to be created
+        while True:
+            pipelines = bids_project_repo.pipelines.list(all=True)
+            no_pipe = all(p.status in ["success", "failed"] for p in pipelines)
+            if no_pipe:
+                break
+            time.sleep(1)
+    return bids_project_repo
 
 
 def init_dicom_study(
@@ -384,8 +399,10 @@ def extract_session_metas(dicom_session_ds: dlad.Dataset) -> dict:
             dic = dicom.read_file(dicom_session_ds.pathobj / f, stop_before_pixels=True)
         except Exception as e:  # TODO: what exception occurs when non-dicom ?
             continue
+        metas = {k: str(getattr(dic, k)).replace("^", "/") for k in SESSION_META_KEYS}
+        metas["StudyDescriptionPath"] = metas["StudyDescription"].split("/")
         # return at first dicom found
-        return {k: str(getattr(dic, k)).replace("^", "/") for k in SESSION_META_KEYS}
+        return metas
     raise InputError("no dicom found")
 
 
@@ -463,7 +480,7 @@ def export_to_s3(
     # git-annex initremote remotename ...
     remote_name = s3_url.hostname
     s3_path = s3_url.path
-    if '{' in s3_path:
+    if "{" in s3_path:
         s3_path = s3_path.format(**session_metas)
     _, bucket_name, *fileprefix = pathlib.Path(s3_path).parts
     fileprefix.append(session_metas["StudyInstanceUID"] + "/")

docker/dicom_indexer/indexer/repo_templates/bids/.bidsignore

@@ -1,3 +1,10 @@
+# localizers/scouts converted in BIDS for QC/QA
 **/anat/*localizer*
 **/anat/*scout*
-**/*__dup*
+# reproin/heudiconv duplicated scans mechanism
+**/*dup*
+**/*_defacemaskreg.mat
+# follow MIDS extension wip for c-spine data
+**/*_bp-*
+# follows SWI BEP wip
+**/swi/*

docker/dicom_indexer/indexer/repo_templates/bids/.gitlab-ci.yml

@@ -1,6 +1,5 @@
 include:
   - local: /.ci-env.yml
   - project: "$NI_DATAOPS_GITLAB_ROOT/ci-pipelines"
-    ref: refactor
     file:
       - 'ci-pipelines/bids/bids_repo.yml'

docker/dicom_indexer/indexer/repo_templates/dicom_study/.gitlab-ci.yml

@@ -2,6 +2,5 @@
 include:
   - local: /.ci-env.yml
   - project: "$NI_DATAOPS_GITLAB_ROOT/ci-pipelines"
-    ref: refactor
     file:
       - 'ci-pipelines/sources/dicoms_study.yml'