job-search/internal/skiprules/file.go

package skiprules

import (
	"errors"
	"log/slog"
	"os"
	"sync"
	"time"
)

// Option configures a File.
type Option func(*File)

// WithLogger makes the file log its reloads and problems to logger instead
// of slog.Default().
func WithLogger(logger *slog.Logger) Option {
	return func(f *File) {
		f.log = logger
	}
}

// File is a rules file that is read again when it changes (see Rules).
// Create it with Open. It is safe for concurrent use.
type File struct {
	path string
	log  *slog.Logger

	mu sync.Mutex
	// rules are the last good rules, or nil while the file is missing.
	rules *Rules
	// modTime and size are the file's when it was last read, good or not, so
	// a bad file isn't read again until it changes.
	modTime time.Time
	size    int64
	// problem is the missing or unreadable file last logged, so it is
	// logged once rather than on every call; "" once the file is found.
	problem string
}

// Open reads the rules file at path. Unlike a reload, a missing or invalid
// file is an error.
func Open(path string, opts ...Option) (*File, error) {
	file := &File{
		path: path,
	}
	for _, option := range opts {
		option(file)
	}

	info, err := os.Stat(path)
	if err != nil {
		return nil, err
	}
	rules, err := Load(path)
	if err != nil {
		return nil, err
	}

	file.rules = rules
	file.modTime = info.ModTime()
	file.size = info.Size()

	return file, nil
}

// logger returns the logger the file logs to, tagged with its component.
func (f *File) logger() *slog.Logger {
	log := f.log
	if log == nil {
		log = slog.Default()
	}
	return log.With("component", "skiprules")
}

// Path returns the path of the rules file.
func (f *File) Path() string { return f.path }

// Rules returns the current rules, first reading the file again if its
// modification time or size has changed since it was last read.
//
// If the file can't be read or parsed, the error is logged and the last good
// rules are kept. If it has been deleted, a warning is logged and the result
// is nil (no rules) until it is back. A file that is missing or can't be
// checked is logged once, not on every call; an invalid one once per change.
func (f *File) Rules() *Rules {
	f.mu.Lock()
	defer f.mu.Unlock()

	info, err := os.Stat(f.path)
	if errors.Is(err, os.ErrNotExist) {
		f.rules = nil
		f.modTime = time.Time{}
		f.size = 0
		if f.newProblem("missing") {
			f.logger().Warn(
				"skip rules file missing, skipping nothing", "file", f.path,
			)
		}
		return nil
	}
	if err != nil {
		if f.newProblem(err.Error()) {
			f.logger().Error(
				"skip rules file unreadable, keeping the last good rules",
				"file", f.path,
				"error", err,
			)
		}
		return f.rules
	}

	f.problem = ""
	if info.ModTime().Equal(f.modTime) && info.Size() == f.size {
		return f.rules
	}

	f.modTime = info.ModTime()
	f.size = info.Size()
	rules, err := Load(f.path)
	if err != nil {
		f.logger().Error(
			"skip rules file invalid, keeping the last good rules",
			"file", f.path,
			"error", err,
		)
		return f.rules
	}

	f.rules = rules
	filters, active := rules.Filters()
	f.logger().Info(
		"skip rules reloaded",
		"file", f.path,
		"filters", filters,
		"active_filters", active,
	)

	return f.rules
}

// newProblem reports whether problem isn't the one last logged, and
// remembers it. Call it with f.mu held.
func (f *File) newProblem(problem string) bool {
	if problem == f.problem {
		return false
	}
	f.problem = problem
	return true
}