Fix file_object_content filters 404ing due to wrong URL prefix

[PhillSimonds]

Why

The three file_object_content* Jinja2 filters shipped in #889 and #904 are 404ing at runtime. The SDK hardcodes URLs under /api/files/*, but the Infrahub backend routes those endpoints under /api/storage/files/*. Every call to file_object_content, file_object_content_by_id, or file_object_content_by_hfid raises JinjaFilterError: Client error '404 Not Found' and the artifact goes to Error.

Goal: restore the three filters to working order and add end-to-end-shaped unit tests so a URL contract drift like this one trips CI next time.

Non-goals: not changing the backend router surface, not adding new filter variants, not touching the computed-attribute gating.

Closes #954

What changed

• Behavioural change users observe: file_object_content, file_object_content_by_id, and file_object_content_by_hfid now return the actual file content instead of raising a 404 wrapper.
• Implementation: six URL strings in infrahub_sdk/object_store.py prefixed with /api/storage/files/ (async + sync). No other behaviour in ObjectStore changed.
• Tests: updated FILE_BY_STORAGE_ID_URL mock constant; added two new constants FILE_BY_ID_URL and FILE_BY_HFID_URL; added three happy-path test cases for file_object_content_by_id, file_object_content_by_hfid (string argument), and file_object_content_by_hfid (list argument). The existing missing-kind, auth-error, and binary-content tests still cover their respective paths.
• What stayed the same: SDK public API signatures, error types, existing test scenarios, and sync-vs-async parity all unchanged.

How to review

• infrahub_sdk/object_store.py: six one-line URL changes on lines 105/110/116/191/196/202. Trivial.
• tests/unit/sdk/test_infrahub_filters.py: the new happy-path tests at the bottom of TestClientDependentFilters are the interesting bits — note they hit per-filter URLs that match the new SDK paths 1:1.
• changelog/954.fixed.md: towncrier fragment.

The bug was introduced in the original filter PRs (#889, #904) and has been present on every branch since. git diff 44a7ccc..origin/infrahub-develop -- infrahub_sdk/object_store.py returns empty — the wrong URLs never changed.

How to test

All should pass. The three new happy-path tests (test_file_object_content_by_id_happy_path, test_file_object_content_by_hfid_happy_path_string, test_file_object_content_by_hfid_happy_path_list) each assert the SDK hits http://mock/api/storage/files/... — if the URL regresses, these fail immediately.

End-to-end reproduction (against a live Infrahub 1.9 stack)

Start an Infrahub 1.9 instance per the usual invoke demo.start flow. Grab the admin token (default 06438eb2-8019-4776-878c-0941b1f1d1ec in dev builds) and set:

export TOKEN=06438eb2-8019-4776-878c-0941b1f1d1ec
export INFRAHUB=http://localhost:8000

1. Load a schema with a TestingFileBlob node inheriting from CoreFileObject:

# /tmp/testing-fileblob.yml
---
version: "1.0"
nodes:
  - name: FileBlob
    namespace: Testing
    include_in_menu: true
    label: FileBlob
    default_filter: label__value
    human_friendly_id: ["label__value"]
    inherit_from: ["CoreFileObject"]
    attributes:
      - name: label
        kind: Text
        unique: true

Load it:

infrahubctl schema load /tmp/testing-fileblob.yml --wait 30

2. Upload a file object via the GraphQL multipart mutation:

echo "hello from blob A" > /tmp/blob_a.txt

OP='{"query": "mutation CreateBlob($file: Upload!) { TestingFileBlobCreate(data: {label: {value: \"blob_a\"}}, file: $file) { ok object { id hfid storage_id { value } } } }", "variables": {"file": null}}'

curl -s -X POST $INFRAHUB/graphql -H "X-INFRAHUB-KEY: $TOKEN" \
  -F "operations=$OP" \
  -F 'map={"0": ["variables.file"]}' \
  -F "0=@/tmp/blob_a.txt"

Note the returned id (node UUID), hfid[0] (blob_a), and storage_id.value — you'll plug these into the templates below.

3. Create a TestingPerson schema + data to act as the artifact target:

# /tmp/testing-person.yml
---
version: "1.0"
nodes:
  - name: Person
    namespace: Testing
    include_in_menu: true
    label: Person
    default_filter: name__value
    human_friendly_id: ["name__value"]
    inherit_from: ["CoreArtifactTarget"]
    attributes:
      - name: name
        kind: Text
        unique: true

infrahubctl schema load /tmp/testing-person.yml --wait 30

# Create a person + a group holding it
curl -s -X POST $INFRAHUB/graphql -H "X-INFRAHUB-KEY: $TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "mutation { p: TestingPersonCreate(data: {name: {value: \"John Doe\"}}) { object { id } } }"}'
# use the returned id for members:[{id: "..."}] in the group create

curl -s -X POST $INFRAHUB/graphql -H "X-INFRAHUB-KEY: $TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "mutation { CoreStandardGroupCreate(data: {name: {value: \"people\"}, members: [{id: \"<person-id>\"}]}) { object { id } } }"}'

4. Prepare a Git repo with three templates exercising each filter variant. Inside a worker container (or any directory reachable by Infrahub as a git remote):

mkdir -p /remote/file-filter-tests/templates
cd /remote/file-filter-tests

# Common query
cat > templates/person.gql <<'EOF'
query Person($name: String!) {
  TestingPerson(name__value: $name) { edges { node { name { value } } } }
}
EOF

# file_object_content by storage_id
cat > templates/by_sid.j2 <<EOF
[by_sid] {{ '<blob_a_storage_id>' | file_object_content }}
EOF

# file_object_content_by_id (node UUID)
cat > templates/by_id.j2 <<EOF
[by_id] {{ '<blob_a_node_id>' | file_object_content_by_id }}
EOF

# file_object_content_by_hfid (needs kind=...)
cat > templates/by_hfid.j2 <<EOF
[by_hfid] {{ 'blob_a' | file_object_content_by_hfid(kind='TestingFileBlob') }}
EOF

cp templates/person.gql templates/by_sid.gql
cp templates/person.gql templates/by_id.gql
cp templates/person.gql templates/by_hfid.gql

cat > .infrahub.yml <<'EOF'
---
queries:
  - { name: by_sid_q, file_path: templates/by_sid.gql }
  - { name: by_id_q, file_path: templates/by_id.gql }
  - { name: by_hfid_q, file_path: templates/by_hfid.gql }
jinja2_transforms:
  - { name: by_sid_t, query: by_sid_q, template_path: templates/by_sid.j2 }
  - { name: by_id_t, query: by_id_q, template_path: templates/by_id.j2 }
  - { name: by_hfid_t, query: by_hfid_q, template_path: templates/by_hfid.j2 }
artifact_definitions:
  - { name: By Sid, artifact_name: by-sid, parameters: { name: name__value }, content_type: text/plain, targets: people, transformation: by_sid_t }
  - { name: By Id, artifact_name: by-id, parameters: { name: name__value }, content_type: text/plain, targets: people, transformation: by_id_t }
  - { name: By Hfid, artifact_name: by-hfid, parameters: { name: name__value }, content_type: text/plain, targets: people, transformation: by_hfid_t }
EOF

git init -q --initial-branch=main && git add -A && git -c user.email=test@test -c user.name=test commit -q -m init

Substitute the real storage_id / node id into by_sid.j2 and by_id.j2 before committing.

5. Register the repo:

infrahubctl repository add file-filter-tests file:///remote/file-filter-tests

Wait ~15 seconds for the artifact pipeline to run.

6. Observe results:

curl -s -X POST $INFRAHUB/graphql -H "X-INFRAHUB-KEY: $TOKEN" -H "Content-Type: application/json" \
  -d '{"query":"{ CoreArtifact { edges { node { name { value } status { value } storage_id { value } } } } }"}' | python3 -m json.tool

• Before this PR: the three by-sid / by-id / by-hfid artifacts have status Error. Worker logs contain Filter 'file_object_content*': failed to retrieve content ... Client error '404 Not Found' for url 'http://server:8000/api/files/...'.
• After this PR: all three artifacts have status Ready. Fetching their content returns e.g. [by_sid] hello from blob A. No 404 errors anywhere in worker logs.

Quick direct HTTP verification independent of the filters:

# Before this PR, both should 404:
curl -s -w "\n%{http_code}\n" "$INFRAHUB/api/files/by-storage-id/<sid>" -H "X-INFRAHUB-KEY: $TOKEN"

# After this PR, the SDK now hits the correct path (which already works):
curl -s -w "\n%{http_code}\n" "$INFRAHUB/api/storage/files/by-storage-id/<sid>" -H "X-INFRAHUB-KEY: $TOKEN"
# → 200 + "hello from blob A"

Impact & rollout

• Backward compatibility: none broken — these filters did not work before, so no caller is depending on the current (error) behaviour. Callers that previously saw JinjaFilterError from these filters will now succeed.
• Performance: none.
• Config/env changes: none.
• Deployment notes: safe to deploy. Users who have worked around the bug by avoiding these filters can now adopt them.

Checklist

• Tests added/updated (3 new happy-path tests covering all three filter URL variants)
• Changelog entry added (changelog/954.fixed.md)
• External docs updated — no user-facing doc changes required; the filters are already documented assuming they work.
• Internal .md docs updated — none needed; no architectural change.

[codecov]

Codecov Report

❌ Patch coverage is 50.00000% with 3 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
infrahub_sdk/object_store.py	50.00%	3 Missing ⚠️

@@                 Coverage Diff                  @@
##           infrahub-develop     #956      +/-   ##
====================================================
+ Coverage             81.11%   81.19%   +0.08%     
====================================================
  Files                   134      134              
  Lines                 11314    11315       +1     
  Branches               1693     1693              
====================================================
+ Hits                   9177     9187      +10     
+ Misses                 1594     1585       -9     
  Partials                543      543              

Flag	Coverage Δ
integration-tests	41.83% <0.00%> (-0.01%)	⬇️
python-3.10	54.10% <50.00%> (+0.07%)	⬆️
python-3.11	54.10% <50.00%> (+0.07%)	⬆️
python-3.12	54.08% <50.00%> (+0.07%)	⬆️
python-3.13	54.08% <50.00%> (+0.05%)	⬆️
python-3.14	55.71% <50.00%> (+0.09%)	⬆️
python-filler-3.12	22.79% <0.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
infrahub_sdk/object_store.py	61.74% <50.00%> (+3.35%)	⬆️

... and 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[gmazoyer]

Nice catch. I don't think you need the changelog record though, since this is still unreleased code, this would appear in the changelog as a fix for something introduced in the same release.

[cloudflare-workers-and-pages]

Deploying infrahub-sdk-python with    Cloudflare Pages

Latest commit:	6fe555a
Status:	✅  Deploy successful!
Preview URL:	https://84958bda.infrahub-sdk-python.pages.dev
Branch Preview URL:	https://fix-issue-954-file-object-ur.infrahub-sdk-python.pages.dev

View logs