#!/usr/bin/env sh
# test-ocn-alignment.sh
#
# The languages of a document share their object numbering, and spine warns
# when they stop doing so. This checks both halves of that: that the sample
# set is silent, and that a document whose languages really do diverge is
# reported.
#
# Two fixtures, because one cannot say both things:
#
#   the sample set     every document aligned, so nothing is reported at
#                      all. This is the half that would fail if a change
#                      made the check cry wolf, and the half that changed on
#                      2026-09-18 when live-manual's ja and pl translations
#                      were mended. Until then ja was two objects short of
#                      en and pl had a stray heading at ocn 313, and this
#                      script asserted those two by name.
#
#   a broken copy      live-manual copied out of the sample set with one
#                      paragraph removed from one language, which is a count
#                      divergence and must be reported. Built here rather
#                      than relied upon, so that mending a translation
#                      cannot quietly leave the check untested. A check that
#                      never fires proves as little as one that always does.
#
# It also asserts that --strict turns the warning into a non-zero exit, and
# that without --strict the run still succeeds: a document with a diverging
# translation is still a document.
#
# Usage:
#   SpinePOD=../../markup/sisudoc-spine-samples/markup/pod-samples/pod \
#     ./test/test-ocn-alignment.sh ./bin/spine-ldc
#
# Exit codes:
#   0  the check reports what it should, and nothing else
#   1  it does not (printed)
#   2  set-up problem (no $SpinePOD, no binary, no samples)

set -e

SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
SPINE_DIR="$(cd "$SCRIPT_DIR/.." && pwd)"

if [ -z "$SpinePOD" ]; then
  echo "ERROR: \$SpinePOD is not set." >&2
  echo "  e.g. SpinePOD=../../markup/sisudoc-spine-samples/markup/pod-samples/pod" >&2
  exit 2
fi
SAMPLES_RAW="$SPINE_DIR/$SpinePOD"
if [ -d "$SAMPLES_RAW" ]; then
  SAMPLES_DIR="$(cd "$SAMPLES_RAW" && pwd)"
else
  SAMPLES_DIR="$SAMPLES_RAW"
fi
if [ ! -d "$SAMPLES_DIR" ]; then
  echo "ERROR: sample documents not found at $SAMPLES_DIR" >&2
  exit 2
fi

SPINE_BIN="${1:-}"
if [ -z "$SPINE_BIN" ]; then
  if   [ -x "$SPINE_DIR/result/bin/spine" ]; then SPINE_BIN="$SPINE_DIR/result/bin/spine"
  elif [ -x "$SPINE_DIR/bin/spine-ldc" ];    then SPINE_BIN="$SPINE_DIR/bin/spine-ldc"
  elif [ -x "$SPINE_DIR/bin/spine" ];        then SPINE_BIN="$SPINE_DIR/bin/spine"
  else
    echo "ERROR: spine binary not found. Specify path as argument." >&2
    exit 2
  fi
fi

OUT_DIR="$SCRIPT_DIR/current-ocn-alignment"
rm -rf "$OUT_DIR"
mkdir -p "$OUT_DIR"

echo "spine binary: $SPINE_BIN"
echo "samples: $SAMPLES_DIR"

FAILURES=0
fail() { echo "  FAIL: $*"; FAILURES=$((FAILURES + 1)); }

# the whole sample set, abstraction only: the check does not need outputs
echo "Abstracting the sample set ..."
set +e
$SPINE_BIN --show-abstraction --skip-output \
  --output="$OUT_DIR" "$SAMPLES_DIR"/* \
  > "$OUT_DIR/run.out" 2> "$OUT_DIR/run.err"
RUN_RC=$?
set -e

# without --strict a diverging document is still built
if [ "$RUN_RC" -ne 0 ]; then
  fail "the run exited $RUN_RC without --strict; a warning must not fail a run"
  head -5 "$OUT_DIR/run.err"
fi

# it must say nothing at all: every document in the set is aligned
n_docs=$(grep -c "^WARNING ocn alignment: " "$OUT_DIR/run.err" || true)
if [ "$n_docs" != "0" ]; then
  fail "$n_docs document(s) reported over the sample set, expected none"
  grep -A 3 "^WARNING ocn alignment: " "$OUT_DIR/run.err" | head -12 | sed 's/^/    /'
fi

# --strict over the whole set therefore passes too
set +e
$SPINE_BIN --show-abstraction --skip-output --strict \
  --output="$OUT_DIR/strict-ok" "$SAMPLES_DIR"/live-manual \
  > "$OUT_DIR/strict-ok.out" 2> "$OUT_DIR/strict-ok.err"
STRICT_OK_RC=$?
set -e
if [ "$STRICT_OK_RC" -ne 0 ]; then
  fail "--strict exited $STRICT_OK_RC on a document whose languages agree"
  head -5 "$OUT_DIR/strict-ok.err"
fi

# and now the other half: a document that really does diverge.
#
# Built here, from a copy, by removing one paragraph from one language. The
# sample set no longer contains a divergence, and a check that never fires
# proves as little as one that always does.
echo "Breaking a copy, to see the check fire ..."
BROKEN="$OUT_DIR/broken"
mkdir -p "$BROKEN"
cp -a "$SAMPLES_DIR/live-manual" "$BROKEN/live-manual"
TARGET="$BROKEN/live-manual/media/text/de/about_manual.ssi"
if [ ! -f "$TARGET" ]; then
  echo "ERROR: fixture not found: $TARGET" >&2
  exit 2
fi
# the first paragraph after the first heading, whole: a numbered object, so
# removing it renumbers everything below it in this language and no other.
#
# The whole of it, to the blank line that ends it. Dropping a single line of
# a wrapped paragraph leaves the paragraph there, one line shorter, and the
# object count unchanged: the first version of this fixture did that and the
# check correctly said nothing.
awk '
  BEGIN { state = 0 }
  state == 0 && /^[A-Za-z]/ && !/^[0-9A-D]~/ && !/^:[A-D]~/ && !/^%/ && !/^@/ \
    { state = 1; next }
  state == 1 && /^[[:space:]]*$/ { state = 2; next }
  state == 1 { next }
  { print }
' "$TARGET" > "$TARGET.new" && mv "$TARGET.new" "$TARGET"

set +e
$SPINE_BIN --show-abstraction --skip-output \
  --output="$OUT_DIR/broken-out" "$BROKEN/live-manual" \
  > "$OUT_DIR/broken.out" 2> "$OUT_DIR/broken.err"
BROKEN_RC=$?
set -e
if [ "$BROKEN_RC" -ne 0 ]; then
  fail "the run over the broken copy exited $BROKEN_RC; a warning must not fail a run"
  head -5 "$OUT_DIR/broken.err"
fi
if ! grep -q "^WARNING ocn alignment: live-manual$" "$OUT_DIR/broken.err"; then
  fail "the removed paragraph was not reported"
  head -5 "$OUT_DIR/broken.err" | sed 's/^/    /'
fi
if ! grep -q "^  \[de\] differs from \[en\]$" "$OUT_DIR/broken.err"; then
  fail "de not reported as differing from en"
fi
if ! grep -q "numbered objects: en .*, de .* (-1)" "$OUT_DIR/broken.err"; then
  fail "de's object count not reported as one short"
  grep "numbered objects" "$OUT_DIR/broken.err" | sed 's/^/    /'
fi
# only the language that was broken
n_langs=$(grep -c "^  \[.*\] differs from " "$OUT_DIR/broken.err" || true)
if [ "$n_langs" != "1" ]; then
  fail "$n_langs languages reported on the broken copy, expected 1 (de)"
  grep "^  \[.*\] differs from " "$OUT_DIR/broken.err" | sed 's/^/    /'
fi

# and --strict makes that a failed run
set +e
$SPINE_BIN --show-abstraction --skip-output --strict \
  --output="$OUT_DIR/broken-strict" "$BROKEN/live-manual" \
  > "$OUT_DIR/broken-strict.out" 2> "$OUT_DIR/broken-strict.err"
STRICT_RC=$?
set -e
if [ "$STRICT_RC" -eq 0 ]; then
  fail "--strict exited 0 on a document whose languages diverge"
fi

rm -rf "$OUT_DIR"

if [ "$FAILURES" -eq 0 ]; then
  echo "PASS: the sample set is aligned, a broken copy is reported, --strict fails"
  exit 0
else
  echo "FAIL: $FAILURES check(s)"
  exit 1
fi