fix(rubygem-pdf-reader): strip 47 PDF test fixtures

[PawelWMS]

Koji build.

Summary

Remove 47 pathological PDF test fixtures from the upstream spec/data/ directory in Source1 (pdf-reader-2.4.2-spec.txz) to avoid scan failures on the SRPM. All fixtures are benign upstream regression PDFs (deliberately-malformed structures, encrypted-cipher variants) that are not consumed at AZL runtime — %files excludes spec/ entirely; the fixtures exist only during %check's rspec run.

How

The companion helper-file patch in modify_source.sh makes pdf_spec_file() in spec/support/reader_spec_helper.rb call rspec's skip only for the 47 removed filenames; it still raises ArgumentError for any other missing fixture. So %check continues running — the 47 affected context blocks in spec/integration_spec.rb get gracefully marked pending, and the remaining ~60 PDFs in spec/data/ continue to exercise the regression suite.

The repack is byte-deterministic (tar --sort=name --mtime --owner=0 --group=0 --numeric-owner | xz -T1 -9e) so the resulting SHA-512 is reproducible from a fresh clone.

Files

• base/comps/components.toml — remove the inline [components.rubygem-pdf-reader] row (dedicated component file is auto-included by the **/*.comp.toml glob).
• base/comps/rubygem-pdf-reader/rubygem-pdf-reader.comp.toml — dedicated component file with a [[...source-files]] block (replace-upstream = true) pointing at the locally-modified Source1 in pkgs_modified/rubygem-pdf-reader/ (SHA-512 49de5d0e…51fd708).
• base/comps/rubygem-pdf-reader/modify_source.sh — the deterministic repack + helper-patch script.
• locks/rubygem-pdf-reader.lock — refreshed input-fingerprint and upstream-commit.
• specs/r/rubygem-pdf-reader/{rubygem-pdf-reader.spec,sources} — re-rendered: release bumped to 12, sources SHA-512 updated.

Build impact

%check continues to run. 47 contexts in spec/integration_spec.rb become rspec "pending" instead of erroring out; the rest of the suite exercises its code paths unchanged.

[Copilot]

Pull request overview

This PR updates the rubygem-pdf-reader component to avoid SRPM scanning failures by replacing the upstream pdf-reader-2.4.2-spec.txz (Source1) with a locally-modified tarball that removes 47 pathological PDF fixtures from spec/data/, while keeping %check running by patching the rspec helper to skip only the removed fixtures.

Changes:

• Add a dedicated rubygem-pdf-reader.comp.toml that replaces the upstream Source1 tarball with a modified lookaside artifact (with replace-upstream = true + rationale).
• Add modify_source.sh to deterministically repack Source1 and patch spec/support/reader_spec_helper.rb to mark the affected tests as skipped/pending instead of failing.
• Refresh the lock + re-rendered spec artifacts (Release bump and updated Source1 SHA-512), and remove the inline component entry from components.toml.

Reviewed changes

Copilot reviewed 5 out of 6 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
base/comps/components.toml	Removes the inline rubygem-pdf-reader entry now that customization lives in a dedicated .comp.toml.
base/comps/rubygem-pdf-reader/rubygem-pdf-reader.comp.toml	Replaces upstream Source1 with a locally-modified pdf-reader-2.4.2-spec.txz in pkgs_modified/, including a replacement rationale.
base/comps/rubygem-pdf-reader/modify_source.sh	Adds a deterministic repack + helper patch workflow to strip fixtures and keep %check functioning.
locks/rubygem-pdf-reader.lock	Updates the input fingerprint to reflect the component definition change.
specs/r/rubygem-pdf-reader/rubygem-pdf-reader.spec	Rendered spec update (release bump) corresponding to the component change.
specs/r/rubygem-pdf-reader/sources	Updates the SHA-512 for the replaced Source1 tarball.

Comments suppressed due to low confidence (1)

base/comps/rubygem-pdf-reader/modify_source.sh:131

• This log line / comment still references "23 stripped fixtures" but FIXTURES_TO_REMOVE is documented as 47 entries and the rest of the script uses ${#FIXTURES_TO_REMOVE[@]}. Consider updating these references to 47 (or make the message dynamic) so the script output and comments stay accurate if the list changes.

echo "[5/6] Patching ${HELPER_FILE} (skip only for the 23 stripped fixtures; still raise otherwise)"
helper="${EXTRACT_DIR}/${HELPER_FILE}"
if [[ ! -f "${helper}" ]]; then
    echo "ERROR: helper missing from upstream tarball: ${HELPER_FILE}" >&2
    exit 1
fi
# The original helper has a single `else / raise ArgumentError, ...` arm for
# the missing-fixture case. We surgically insert a new `elsif` arm BEFORE the
# existing `else`, gated by an explicit allow-list of the 23 stripped
# basenames (with `.pdf` extension). Any other missing fixture still raises,
# so an unrelated upstream change that drops a different fixture surfaces as
# a hard error rather than a silent skip.

[reubeno]

Left a comment. Could you please also update the commit and PR to use conventional commit syntax, as described in CONTRIBUTING.md, please?

[reubeno]

Per our discussion, changes to this file should really be applied using a proper visible patch that's not buried in this removal script. I think we can take this change for now, but I'd like you to file a tracking bug on it for each such script that does this sort of thing.