#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
PROJECT_ROOT="$(cd "$SCRIPT_DIR/.." && pwd)"
KNOWLEDGE_ROOT="$PROJECT_ROOT/knowledge/entries"
OUTPUT="$PROJECT_ROOT/knowledge-index.html"
if [ ! -d "$KNOWLEDGE_ROOT" ]; then
KNOWLEDGE_ROOT="$PROJECT_ROOT"
fi
# ---- Phase 1: Scan directories ----
echo "[1/4] Scanning knowledge directories..."
dirs=()
while IFS= read -r d; do
dirs+=("${d%/}")
done < <(ls -d "$KNOWLEDGE_ROOT"/knowledge_*/ 2>/dev/null || true)
if [ ${#dirs[@]} -eq 0 ]; then
echo " No knowledge_*/ directories found in $KNOWLEDGE_ROOT"
echo " Generated empty index (will show 'no entries' message)."
fi
n=${#dirs[@]}; echo " Found $n knowledge director$([ "$n" -ne 1 ] && echo 'ies' || echo 'y')."
# ---- Phase 2: Extract metadata ----
echo "[2/4] Extracting metadata..."
entries_js=""
count=0
for dir in "${dirs[@]}"; do
dirname=$(basename "$dir")
# Parse directory name: knowledge_YYYYMMDD_Slug
date_part=$(echo "$dirname" | sed -E 's/^knowledge_([0-9]{8})_.*/\1/')
slug=$(echo "$dirname" | sed -E 's/^knowledge_[0-9]{8}_//')
if [ -z "$date_part" ] || [ "$date_part" = "$dirname" ]; then
echo " WARNING: '$dirname' doesn't match knowledge_YYYYMMDD_Slug pattern, skipping..."
continue
fi
# Find .md file
md_file="$dir/${dirname}.md"
if [ ! -f "$md_file" ]; then
# Try to find any .md file in the directory
md_file=$(ls "$dir"/*.md 2>/dev/null | head -1)
if [ -z "$md_file" ]; then
echo " WARNING: No .md file found in '$dirname', skipping..."
continue
fi
fi
# Extract YAML frontmatter (between first and second ---)
yaml=$(sed -n '/^---$/,/^---$/p' "$md_file" | sed '1d;$d' 2>/dev/null || echo "")
# Extract title
title=$(echo "$yaml" | grep -i "^title:" | head -1 | sed 's/^[Tt]itle: *//' | sed 's/^"//;s/"$//;s/^'"'"'//;s/'"'"'$//' || echo "")
# Extract author
author=$(echo "$yaml" | grep -i "^author:" | head -1 | sed 's/^[Aa]uthor: *//' | sed 's/^"//;s/"$//' || echo "")
# Extract tags: handle both ["a","b"] and [a,b] formats
tags_raw=$(echo "$yaml" | grep -i "^tags:" | head -1 | sed 's/^[Tt]ags: *//' || echo "")
# Normalize: remove brackets, split by comma, strip quotes/spaces
tags_csv=$(echo "$tags_raw" | sed 's/^\[//;s/\]$//' | sed 's/"//g' | sed "s/'//g" | tr ',' '\n' | sed 's/^[[:space:]]*//;s/[[:space:]]*$//' | grep -v '^$' | paste -sd ',' - 2>/dev/null || echo "")
# Format tags as JS array
if [ -n "$tags_csv" ]; then
tags_js=""
IFS=',' read -ra TAG_ARR <<< "$tags_csv"
for t in "${TAG_ARR[@]}"; do
t_trimmed=$(echo "$t" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')
if [ -n "$t_trimmed" ]; then
[ -n "$tags_js" ] && tags_js+=", "
# Escape single quotes in tag
t_escaped=$(echo "$t_trimmed" | sed "s/'/\\\\'/g")
tags_js+="'$t_escaped'"
fi
done
else
tags_js=""
fi
# Fallback title
if [ -z "$title" ]; then
title=$(echo "$slug" | sed 's/_/ /g')
fi
[ -z "$author" ] && author="unknown"
# Extract body text (everything after the second ---)
body=$(awk 'BEGIN{found=0} /^---$/{found++; next} found>=2{print}' "$md_file" 2>/dev/null || echo "")
# Escape for JS string: backslash, single quote, newline
title_esc=$(echo "$title" | sed "s/\\\\/\\\\\\\\/g; s/'/\\\\'/g")
author_esc=$(echo "$author" | sed "s/\\\\/\\\\\\\\/g; s/'/\\\\'/g")
body_esc=$(echo "$body" | sed "s/\\\\/\\\\\\\\/g; s/'/\\\\'/g" | tr '\n' ' ' | sed 's/ */ /g')
slug_esc=$(echo "$slug" | sed "s/\\\\/\\\\\\\\/g; s/'/\\\\'/g")
rel_dir=$(realpath --relative-to="$PROJECT_ROOT" "$dir" 2>/dev/null || echo "$dirname")
rel_dir=${rel_dir//\\//}
rel_dir_esc=$(echo "$rel_dir" | sed "s/\\\\/\\\\\\\\/g; s/'/\\\\'/g")
# Build JS entry
entries_js+=" {"
entries_js+=" slug: '$slug_esc',"
entries_js+=" path: '$rel_dir_esc',"
entries_js+=" date: '$date_part',"
entries_js+=" title: '$title_esc',"
entries_js+=" author: '$author_esc',"
entries_js+=" tags: [$tags_js],"
entries_js+=" body: '$body_esc'"
entries_js+=" },"
entries_js+=$'\n'
count=$((count + 1))
echo " ✓ $dirname"
done
# ---- Phase 3: Generate HTML ----
echo "[3/4] Generating knowledge-index.html..."
entries_js=$(echo "$entries_js" | sed '$ s/,$//')
# Write HTML header (before __DATA_PLACEHOLDER__)
cat > "$OUTPUT" << 'HTMLEOF'
知识库索引 — github-learn
知识库索引
github-learn · 共 0 条知识条目
没有找到匹配的条目
HTMLEOF
# ---- Phase 4: Done ----
echo "[4/4] Done!"
echo " Output: $OUTPUT"
echo " Entries: $count"
echo " Open with: start knowledge-index.html"