Attempting to add concurrency, may god have mercy on my soul

This commit is contained in:
Seednode 2022-10-20 14:22:01 -05:00
parent 9cc372d7c4
commit 1cbc84d789
4 changed files with 135 additions and 77 deletions

View File

@ -13,16 +13,55 @@ import (
"regexp" "regexp"
"strconv" "strconv"
"strings" "strings"
"sync"
"sync/atomic"
"time" "time"
"github.com/h2non/filetype" "github.com/h2non/filetype"
) )
var (
ErrNoImagesFound = fmt.Errorf("no supported image formats found")
)
type Concurrency struct {
DirectoryScans chan int
FileScans chan int
}
type Files struct {
Mutex sync.Mutex
List map[string][]string
}
type Stats struct { type Stats struct {
FilesMatched uint64 FilesMatched uint64
FilesSkipped uint64 FilesSkipped uint64
DirectoriesMatched uint64 DirectoriesMatched uint64
DirectoriesSkipped uint64 }
func (s *Stats) IncrementFilesMatched() {
atomic.AddUint64(&s.FilesMatched, 1)
}
func (s *Stats) GetFilesMatched() uint64 {
return atomic.LoadUint64(&s.FilesMatched)
}
func (s *Stats) IncrementFilesSkipped() {
atomic.AddUint64(&s.FilesSkipped, 1)
}
func (s *Stats) GetFilesSkipped() uint64 {
return atomic.LoadUint64(&s.FilesSkipped)
}
func (s *Stats) IncrementDirectoriesMatched() {
atomic.AddUint64(&s.DirectoriesMatched, 1)
}
func (s *Stats) GetDirectoriesMatched() uint64 {
return atomic.LoadUint64(&s.DirectoriesMatched)
} }
type Path struct { type Path struct {
@ -39,14 +78,18 @@ func (p *Path) Decrement() {
p.Number = p.Number - 1 p.Number = p.Number - 1
} }
var ( func appendPath(directory, path string, files *Files, stats *Stats) {
ErrNoImagesFound = fmt.Errorf("no supported image formats found") files.Mutex.Lock()
) files.List[directory] = append(files.List[directory], path)
files.Mutex.Unlock()
func appendPaths(m map[string][]string, path string, filters *Filters, stats *Stats) (map[string][]string, error) { stats.IncrementFilesMatched()
}
func appendPaths(path string, files *Files, filters *Filters, stats *Stats) error {
absolutePath, err := filepath.Abs(path) absolutePath, err := filepath.Abs(path)
if err != nil { if err != nil {
return nil, err return err
} }
directory, filename := filepath.Split(absolutePath) directory, filename := filepath.Split(absolutePath)
@ -60,41 +103,39 @@ func appendPaths(m map[string][]string, path string, filters *Filters, stats *St
filename, filename,
filters.Includes[i], filters.Includes[i],
) { ) {
m[directory] = append(m[directory], path) appendPath(directory, path, files, stats)
stats.FilesMatched += 1
return m, nil return nil
} }
} }
stats.FilesSkipped += 1 stats.IncrementFilesSkipped()
return m, nil return nil
case !filters.HasIncludes() && filters.HasExcludes(): case !filters.HasIncludes() && filters.HasExcludes():
for i := 0; i < len(filters.Excludes); i++ { for i := 0; i < len(filters.Excludes); i++ {
if strings.Contains( if strings.Contains(
filename, filename,
filters.Excludes[i], filters.Excludes[i],
) { ) {
stats.FilesSkipped += 1 stats.IncrementFilesSkipped()
return m, nil return nil
} }
} }
m[directory] = append(m[directory], path) appendPath(directory, path, files, stats)
stats.FilesMatched += 1
return m, nil return nil
case filters.HasIncludes() && filters.HasExcludes(): case filters.HasIncludes() && filters.HasExcludes():
for i := 0; i < len(filters.Excludes); i++ { for i := 0; i < len(filters.Excludes); i++ {
if strings.Contains( if strings.Contains(
filename, filename,
filters.Excludes[i], filters.Excludes[i],
) { ) {
stats.FilesSkipped += 1 stats.IncrementFilesSkipped()
return m, nil return nil
} }
} }
@ -103,21 +144,19 @@ func appendPaths(m map[string][]string, path string, filters *Filters, stats *St
filename, filename,
filters.Includes[i], filters.Includes[i],
) { ) {
m[directory] = append(m[directory], path) appendPath(directory, path, files, stats)
stats.FilesMatched += 1
return m, nil return nil
} }
} }
stats.FilesSkipped += 1 stats.IncrementFilesSkipped()
return m, nil return nil
default: default:
m[directory] = append(m[directory], path) appendPath(directory, path, files, stats)
stats.FilesMatched += 1
return m, nil return nil
} }
} }
@ -271,60 +310,69 @@ func isImage(path string) (bool, error) {
return filetype.IsImage(head), nil return filetype.IsImage(head), nil
} }
func getFiles(m map[string][]string, path string, filters *Filters, stats *Stats) (map[string][]string, error) { func getFiles(path string, files *Files, filters *Filters, stats *Stats, concurrency *Concurrency) error {
var wg sync.WaitGroup
err := filepath.WalkDir(path, func(p string, info os.DirEntry, err error) error { err := filepath.WalkDir(path, func(p string, info os.DirEntry, err error) error {
if err != nil { if err != nil {
return err return err
} }
if info.IsDir() && Filter && filters.HasExcludes() {
for i := 0; i < len(filters.Excludes); i++ {
if strings.Contains(
strings.ToLower(p),
strings.ToLower(filters.Excludes[i]),
) {
stats.DirectoriesSkipped += 1
return filepath.SkipDir
}
}
}
switch { switch {
case !Recursive && info.IsDir() && p != path: case !Recursive && info.IsDir() && p != path:
stats.DirectoriesSkipped += 1
return filepath.SkipDir return filepath.SkipDir
case !info.IsDir(): case !info.IsDir():
m, err = appendPaths(m, p, filters, stats) wg.Add(1)
if err != nil { concurrency.FileScans <- 1
return err
} go func() {
defer func() {
<-concurrency.FileScans
wg.Done()
}()
err = appendPaths(p, files, filters, stats)
if err != nil {
fmt.Println(err)
}
}()
case info.IsDir(): case info.IsDir():
stats.DirectoriesMatched += 1 stats.IncrementDirectoriesMatched()
} }
return err return err
}) })
wg.Wait()
if err != nil { if err != nil {
return nil, err return err
} }
return m, nil return nil
} }
func getFileList(paths []string, filters *Filters, stats *Stats) (map[string][]string, error) { func getFileList(paths []string, files *Files, filters *Filters, stats *Stats, concurrency *Concurrency) {
fileMap := map[string][]string{} var wg sync.WaitGroup
var err error
for i := 0; i < len(paths); i++ { for i := 0; i < len(paths); i++ {
fileMap, err = getFiles(fileMap, paths[i], filters, stats) wg.Add(1)
if err != nil { concurrency.DirectoryScans <- 1
return nil, err
} go func(i int) {
defer func() {
<-concurrency.DirectoryScans
wg.Done()
}()
err := getFiles(paths[i], files, filters, stats, concurrency)
if err != nil {
fmt.Println(err)
}
}(i)
} }
return fileMap, nil wg.Wait()
} }
func cleanFilename(filename string) string { func cleanFilename(filename string) string {
@ -345,24 +393,24 @@ func prepareDirectory(directory []string) []string {
} }
} }
func prepareDirectories(m map[string][]string, sort string) []string { func prepareDirectories(files *Files, sort string) []string {
directories := []string{} directories := []string{}
keys := make([]string, len(m)) keys := make([]string, len(files.List))
i := 0 i := 0
for k := range m { for k := range files.List {
keys[i] = k keys[i] = k
i++ i++
} }
if sort == "asc" || sort == "desc" { if sort == "asc" || sort == "desc" {
for i := 0; i < len(keys); i++ { for i := 0; i < len(keys); i++ {
directories = append(directories, prepareDirectory(m[keys[i]])...) directories = append(directories, prepareDirectory(files.List[keys[i]])...)
} }
} else { } else {
for i := 0; i < len(keys); i++ { for i := 0; i < len(keys); i++ {
directories = append(directories, m[keys[i]]...) directories = append(directories, files.List[keys[i]]...)
} }
} }
@ -371,22 +419,24 @@ func prepareDirectories(m map[string][]string, sort string) []string {
func pickFile(args []string, filters *Filters, sort string) (string, error) { func pickFile(args []string, filters *Filters, sort string) (string, error) {
stats := Stats{} stats := Stats{}
files := Files{}
files.List = make(map[string][]string)
fileMap, err := getFileList(args, filters, &stats) concurrency := Concurrency{}
if err != nil { concurrency.DirectoryScans = make(chan int, maxDirectoryScans)
return "", err concurrency.FileScans = make(chan int, maxFileScans)
}
getFileList(args, &files, filters, &stats, &concurrency)
if Count { if Count {
fmt.Printf("Choosing from %v files (skipped %v) out of %v directories (skipped %v)\n", fmt.Printf("Choosing from %v files (skipped %v) in %v directories\n",
stats.FilesMatched, stats.GetFilesMatched(),
stats.FilesSkipped, stats.GetFilesSkipped(),
stats.DirectoriesMatched, stats.GetDirectoriesMatched(),
stats.DirectoriesSkipped,
) )
} }
fileList := prepareDirectories(fileMap, sort) fileList := prepareDirectories(&files, sort)
rand.Seed(time.Now().UnixNano()) rand.Seed(time.Now().UnixNano())

View File

@ -11,6 +11,14 @@ import (
"github.com/spf13/cobra" "github.com/spf13/cobra"
) )
type MaxConcurrency int
const (
// avoid hitting default open file descriptor limits (1024)
maxDirectoryScans MaxConcurrency = 32
maxFileScans MaxConcurrency = 256
)
var Count bool var Count bool
var Filter bool var Filter bool
var Port uint16 var Port uint16

View File

@ -10,7 +10,7 @@ import (
"github.com/spf13/cobra" "github.com/spf13/cobra"
) )
var Version = "0.16.0" var Version = "0.17.0"
func init() { func init() {
rootCmd.AddCommand(versionCmd) rootCmd.AddCommand(versionCmd)

View File

@ -16,13 +16,16 @@ import (
"time" "time"
) )
const LOGDATE string = "2006-01-02T15:04:05.000-07:00"
const PREFIX string = "/src"
type Filters struct { type Filters struct {
Includes []string Includes []string
Excludes []string Excludes []string
} }
func (f *Filters) IsEmpty() bool { func (f *Filters) IsEmpty() bool {
if !f.HasIncludes() && !f.HasExcludes() { if !(f.HasIncludes() && f.HasExcludes()) {
return true return true
} }
@ -61,9 +64,6 @@ func (fn appHandler) ServeHTTP(w http.ResponseWriter, r *http.Request) {
} }
} }
const LOGDATE string = "2006-01-02T15:04:05.000-07:00"
const PREFIX string = "/src"
func splitQueryParams(query string) []string { func splitQueryParams(query string) []string {
if query == "" { if query == "" {
return []string{} return []string{}