generate-report.sh

#!/bin/bash

# This script generates a Markdown report for Jenkins CSP project updates based on input JSON data and a list of repositories with releases.

# Check if the input JSON file is provided
if [ -z "$1" ]; then
  echo "Usage: $0 <input-json-file> <repos-file>"
  exit 1
fi

# Input JSON file (passed as a parameter)
INPUT_JSON="$1"

# Repositories file (passed as a parameter)
REPOS_FILE="$2"

# Validate that the input JSON file exists and is readable
if [ ! -r "$INPUT_JSON" ]; then
  echo "Error: Input file '$INPUT_JSON' does not exist or is not readable" >&2
  exit 1
fi

# Validate that the repositories file exists and is readable
if [ ! -r "$REPOS_FILE" ]; then
  echo "Error: Repositories file '$REPOS_FILE' does not exist or is not readable" >&2
  exit 1
fi

# Validate JSON structure
if ! jq empty "$INPUT_JSON" 2>/dev/null; then
  echo "Error: Invalid JSON format in '$INPUT_JSON'" >&2
  exit 1
fi

# Validate required fields are present in the JSON
if ! jq -e 'all(has("repository") and has("user") and has("title") and has("number") and has("state"))' "$INPUT_JSON" >/dev/null; then
  echo "Error: JSON is missing required fields (repository, user, title, number, or state)" >&2
  exit 1
fi

# Debugging: Print the input filename
echo "Input filename: $INPUT_JSON"

# Extract the year and month number from the filename
YEAR=$(basename "$INPUT_JSON" | cut -d'_' -f5 | cut -d'-' -f1)
MONTH=$(basename "$INPUT_JSON" | cut -d'_' -f5 | cut -d'-' -f2)

# Debugging: Print the extracted year and month
echo "Extracted year: $YEAR"
echo "Extracted month: $MONTH"

# Map month number to month name
case $MONTH in
  01) MONTH_NAME="January" ;;
  02) MONTH_NAME="February" ;;
  03) MONTH_NAME="March" ;;
  04) MONTH_NAME="April" ;;
  05) MONTH_NAME="May" ;;
  06) MONTH_NAME="June" ;;
  07) MONTH_NAME="July" ;;
  08) MONTH_NAME="August" ;;
  09) MONTH_NAME="September" ;;
  10) MONTH_NAME="October" ;;
  11) MONTH_NAME="November" ;;
  12) MONTH_NAME="December" ;;
  *) MONTH_NAME="Unknown" ;;
esac

# Debugging: Print the month name
echo "Month name: $MONTH_NAME"

# Output Markdown file with year and month
OUTPUT_MD="jenkins-csp-${MONTH_NAME,,}-${YEAR}-report.md"

# Cache for first names
declare -A FIRST_NAME_CACHE

# Function to get the first name of a GitHub user
# Arguments:
#   $1 - GitHub handle of the user
# Returns:
#   First name of the user
get_first_name() {
  local github_handle=$1

  # Check if the first name is already cached
  if [[ -n "${FIRST_NAME_CACHE[$github_handle]}" ]]; then
    echo "${FIRST_NAME_CACHE[$github_handle]}"
    return
  fi

  # Fetch the full name from the GitHub API
  local full_name
  if ! full_name=$(gh api users/"$github_handle" --jq '.name // empty' 2>/dev/null); then
    echo "$github_handle"  # Fallback to GitHub handle on API error
    return
  fi

  # Extract the first name (first word of the full name)
  local first_name
  if [[ -z "$full_name" ]]; then
    first_name="$github_handle"  # Fallback to GitHub handle if no name
  else
    first_name=$(echo "$full_name" | cut -d' ' -f1)
  fi

  # Cache the first name
  FIRST_NAME_CACHE[$github_handle]=$first_name
  echo "$first_name"
}

# Function to generate the Markdown report
generate_report() {
  echo "# ${MONTH_NAME} ${YEAR} - Jenkins CSP Project Update" > "$OUTPUT_MD"
  echo "" >> "$OUTPUT_MD"

  # Summary section
  if ! SUMMARY=$(jq -r '{
    prs: length,
    repos: (group_by(.repository) | length),
    users: (group_by(.user) | length),
    open: map(select(.state == "OPEN")) | length,
    closed: map(select(.state == "CLOSED")) | length,
    merged: map(select(.state == "MERGED")) | length
  } | . += {
    open_percent: (.open / .prs * 100 | floor),
    closed_percent: (.closed / .prs * 100 | floor),
    merged_percent: (.merged / .prs * 100 | floor)
  }' "$INPUT_JSON" 2>/dev/null); then
    echo "Error: Failed to process JSON data for summary" >&2
    exit 1
  fi
  TOTAL_PRS=$(echo "$SUMMARY" | jq '.prs')
  TOTAL_REPOS=$(echo "$SUMMARY" | jq '.repos')
  TOTAL_USERS=$(echo "$SUMMARY" | jq '.users')
  OPEN_PRS=$(echo "$SUMMARY" | jq '.open')
  CLOSED_PRS=$(echo "$SUMMARY" | jq '.closed')
  MERGED_PRS=$(echo "$SUMMARY" | jq '.merged')

  echo "## Summary" >> "$OUTPUT_MD"
  echo "- Total PRs: $TOTAL_PRS" >> "$OUTPUT_MD"
  echo "- Total Repositories: $TOTAL_REPOS" >> "$OUTPUT_MD"
  echo "- Total Users: $TOTAL_USERS" >> "$OUTPUT_MD"
  echo "- Open PRs: $OPEN_PRS ($(echo "$SUMMARY" | jq '.open_percent')%)" >> "$OUTPUT_MD"
  echo "- Closed PRs: $CLOSED_PRS ($(echo "$SUMMARY" | jq '.closed_percent')%)" >> "$OUTPUT_MD"
  echo "- Merged PRs: $MERGED_PRS ($(echo "$SUMMARY" | jq '.merged_percent')%)" >> "$OUTPUT_MD"
  echo "" >> "$OUTPUT_MD"

  echo "## Pull Requests by Repository" >> "$OUTPUT_MD"
  echo "" >> "$OUTPUT_MD"

  # Step 1: Group PRs by repository
  jq -r '
    group_by(.repository)[] |
    "### \(.[0].repository)\n" + (
      # Step 2: Group PRs by user within each repository
      group_by(.user)[] |
      "#### User: \(.[0].user)\n" + (
        # Step 3: List all PRs for the user with the new format
        .[] |
        if .state == "OPEN" then
          "- \(.user) is working on [\(.title // "[No title]")](https://github.com/\(.repository)/pull/\(.number)) (\(.createdAt))\n"
        elif .state == "CLOSED" or .state == "MERGED" then
          "- \(.user) has worked on [\(.title // "[No title]")](https://github.com/\(.repository)/pull/\(.number)) (\(.createdAt))\n"
        else
          "- [\(.title // "[No title]")](https://github.com/\(.repository)/pull/\(.number)) (\(.createdAt)) - Status: \(.state // "null")\n"
        end
      )
    )
  ' "$INPUT_JSON" | while read -r line; do
    # Replace GitHub handle with the user's first name
    if [[ $line =~ ^####\ User:\ (.+)$ ]]; then
      github_handle=${BASH_REMATCH[1]}
      first_name=$(get_first_name "$github_handle")
      echo "#### User: $first_name" >> "$OUTPUT_MD"
    elif [[ $line =~ ^-\ (.+)\ (is working on|has worked on)\ \[(.+)\]\((https:\/\/github\.com\/.+)\)\ \((.+)\)$ ]]; then
      github_handle=${BASH_REMATCH[1]}
      action=${BASH_REMATCH[2]}
      title=${BASH_REMATCH[3]}
      url=${BASH_REMATCH[4]}
      date=${BASH_REMATCH[5]}
      first_name=$(get_first_name "$github_handle")
      echo "- $first_name $action [$title]($url) ($date)" >> "$OUTPUT_MD"
    else
      echo "$line" >> "$OUTPUT_MD"
    fi
  done

  echo "" >> "$OUTPUT_MD"
  # Read template sections from file
  if [ -f "report-template.md" ]; then
   cat "report-template.md" >> "$OUTPUT_MD"
 else
     echo "" >> "$OUTPUT_MD"
     echo "## Key Highlights" >> "$OUTPUT_MD"
     echo "- Continued progress in modernizing Jenkins plugins" >> "$OUTPUT_MD"
     echo "- Systematic removal of legacy JavaScript and inline event handlers" >> "$OUTPUT_MD"
     echo "- Enhanced Content Security Policy (CSP) compatibility" >> "$OUTPUT_MD"
     echo "- Proactive identification and resolution of potential security vulnerabilities" >> "$OUTPUT_MD"
     echo "" >> "$OUTPUT_MD"
     echo "## Next Steps" >> "$OUTPUT_MD"
     echo "- Continue plugin modernization efforts" >> "$OUTPUT_MD"
     echo "- Prioritize plugins with known CSP challenges" >> "$OUTPUT_MD"
     echo "- Expand CSP scanner capabilities" >> "$OUTPUT_MD"
     echo "- Collaborate with plugin maintainers to implement best practices" >> "$OUTPUT_MD"
  fi

  # Include the list of repositories with releases
  # Default to alphabetical sorting unless PRESERVE_ORDER is set
  if [ "${PRESERVE_ORDER:-0}" = "1" ]; then
   SORT_CMD="cat"
  else
    SORT_CMD="sort"
  fi

  echo "" >> "$OUTPUT_MD"
  echo "### Released plugins" >> "$OUTPUT_MD"
  echo "" >> "$OUTPUT_MD"
  if [ ! -s "$REPOS_FILE" ]; then
    echo "No plugins were released during this period." >> "$OUTPUT_MD"
    return
  fi
  # The output of the SORT_CMD command is piped into the nl command,
  # which numbers each line.
  # The -w1 option specifies the width of the line numbers,
  # and the -s'. ' option sets the separator between the line number and the line
  # content to a period followed by a space.
  # The while IFS= read -r line; do loop iterates over each line of the numbered
  # and sorted file.
  # The IFS= ensures that leading and trailing whitespace is preserved, and
  # read -r line reads each line into the variable line.
  $SORT_CMD "$REPOS_FILE" | nl -w1 -s'. ' | while IFS= read -r line; do
    # Within the loop, the repo=$(echo "$line" | awk '{print $2}') command
    # extracts the repository name from each line.
    # The awk '{print $2}' command prints the second field of the line, which
    # corresponds to the repository name.
    repo=$(echo "$line" | awk '{print $2}')
    # echo "$line Released the [$repo](https://github.com/$repo)" >> "$OUTPUT_MD"
    repo_name=${repo#*/}
    repo_name=${repo_name//-plugin/ plugin}
    echo "${line%% *} Released the [$repo_name](https://github.com/$repo)" >> "$OUTPUT_MD"
  done
}

# Run the report generation
generate_report

echo "Markdown report generated: $OUTPUT_MD"