Convert linguist attribute handling to optional.Option

Based on @KN4CK3R's work in gitea#29267. This drops the custom
`LinguistBoolAttrib` type, and uses `optional.Option` instead. I added
the `isTrue()` and `isFalse()` (function-local) helpers to make the code
easier to follow, because these names convey their goal better than
`v.ValueorDefault(false)` or `!v.ValueOrDefault(true)`.

Signed-off-by: Gergely Nagy <forgejo@gergo.csillger.hu>
This commit is contained in:
Gergely Nagy 2024-02-26 12:52:59 +01:00
parent c80406332f
commit ee39c58120
No known key found for this signature in database
4 changed files with 63 additions and 54 deletions

View file

@ -11,6 +11,7 @@ import (
"os"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/optional"
)
// CheckAttributeOpts represents the possible options to CheckAttribute
@ -322,3 +323,23 @@ func (repo *Repository) CheckAttributeReader(commitID string) (*CheckAttributeRe
return checker, deferable
}
// true if "set"/"true", false if "unset"/"false", none otherwise
func attributeToBool(attr map[string]string, name string) optional.Option[bool] {
if value, has := attr[name]; has && value != "unspecified" {
switch value {
case "set", "true":
return optional.Some(true)
case "unset", "false":
return optional.Some(false)
}
}
return optional.None[bool]()
}
func attributeToString(attr map[string]string, name string) optional.Option[string] {
if value, has := attr[name]; has && value != "unspecified" {
return optional.Some(value)
}
return optional.None[string]()
}

View file

@ -13,18 +13,6 @@ const (
bigFileSize int64 = 1024 * 1024 // 1 MiB
)
type LinguistBoolAttrib struct {
Value string
}
func (attrib *LinguistBoolAttrib) IsTrue() bool {
return attrib.Value == "set" || attrib.Value == "true"
}
func (attrib *LinguistBoolAttrib) IsFalse() bool {
return attrib.Value == "unset" || attrib.Value == "false"
}
// mergeLanguageStats mergers language names with different cases. The name with most upper case letters is used.
func mergeLanguageStats(stats map[string]int64) map[string]int64 {
names := map[string]struct {

View file

@ -12,6 +12,7 @@ import (
"strings"
"code.gitea.io/gitea/modules/analyze"
"code.gitea.io/gitea/modules/optional"
"github.com/go-enry/go-enry/v2"
"github.com/go-git/go-git/v5"
@ -53,31 +54,30 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
firstExcludedLanguage := ""
firstExcludedLanguageSize := int64(0)
isTrue := func(v optional.Option[bool]) bool {
return v.ValueOrDefault(false)
}
isFalse := func(v optional.Option[bool]) bool {
return !v.ValueOrDefault(true)
}
err = tree.Files().ForEach(func(f *object.File) error {
if f.Size == 0 {
return nil
}
isVendored := LinguistBoolAttrib{}
isGenerated := LinguistBoolAttrib{}
isDocumentation := LinguistBoolAttrib{}
isDetectable := LinguistBoolAttrib{}
isVendored := optional.None[bool]()
isGenerated := optional.None[bool]()
isDocumentation := optional.None[bool]()
isDetectable := optional.None[bool]()
if checker != nil {
attrs, err := checker.CheckPath(f.Name)
if err == nil {
if vendored, has := attrs["linguist-vendored"]; has {
isVendored = LinguistBoolAttrib{Value: vendored}
}
if generated, has := attrs["linguist-generated"]; has {
isGenerated = LinguistBoolAttrib{Value: generated}
}
if documentation, has := attrs["linguist-documentation"]; has {
isDocumentation = LinguistBoolAttrib{Value: documentation}
}
if detectable, has := attrs["linguist-detectable"]; has {
isDetectable = LinguistBoolAttrib{Value: detectable}
}
isVendored = attributeToBool(attrs, "linguist-vendored")
isGenerated = attributeToBool(attrs, "linguist-generated")
isDocumentation = attributeToBool(attrs, "linguist-documentation")
isDetectable = attributeToBool(attrs, "linguist-detectable")
if language, has := attrs["linguist-language"]; has && language != "unspecified" && language != "" {
// group languages, such as Pug -> HTML; SCSS -> CSS
group := enry.GetLanguageGroup(language)
@ -108,11 +108,11 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
}
}
if isDetectable.IsFalse() || isVendored.IsTrue() || isDocumentation.IsTrue() ||
(!isVendored.IsFalse() && analyze.IsVendor(f.Name)) ||
if isFalse(isDetectable) || isTrue(isVendored) || isTrue(isDocumentation) ||
(!isFalse(isVendored) && analyze.IsVendor(f.Name)) ||
enry.IsDotFile(f.Name) ||
enry.IsConfiguration(f.Name) ||
(!isDocumentation.IsFalse() && enry.IsDocumentation(f.Name)) {
(!isFalse(isDocumentation) && enry.IsDocumentation(f.Name)) {
return nil
}
@ -121,7 +121,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
if f.Size <= bigFileSize {
content, _ = readFile(f, fileSizeLimit)
}
if !isGenerated.IsTrue() && enry.IsGenerated(f.Name, content) {
if !isTrue(isGenerated) && enry.IsGenerated(f.Name, content) {
return nil
}
@ -142,7 +142,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
langtype := enry.GetLanguageType(language)
included = langtype == enry.Programming || langtype == enry.Markup
if !included {
if isDetectable.IsTrue() {
if isTrue(isDetectable) {
included = true
} else {
return nil

View file

@ -15,6 +15,7 @@ import (
"code.gitea.io/gitea/modules/analyze"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/optional"
"github.com/go-enry/go-enry/v2"
)
@ -77,6 +78,13 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
firstExcludedLanguage := ""
firstExcludedLanguageSize := int64(0)
isTrue := func(v optional.Option[bool]) bool {
return v.ValueOrDefault(false)
}
isFalse := func(v optional.Option[bool]) bool {
return !v.ValueOrDefault(true)
}
for _, f := range entries {
select {
case <-repo.Ctx.Done():
@ -91,26 +99,18 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
continue
}
isVendored := LinguistBoolAttrib{}
isGenerated := LinguistBoolAttrib{}
isDocumentation := LinguistBoolAttrib{}
isDetectable := LinguistBoolAttrib{}
isVendored := optional.None[bool]()
isGenerated := optional.None[bool]()
isDocumentation := optional.None[bool]()
isDetectable := optional.None[bool]()
if checker != nil {
attrs, err := checker.CheckPath(f.Name())
if err == nil {
if vendored, has := attrs["linguist-vendored"]; has {
isVendored = LinguistBoolAttrib{Value: vendored}
}
if generated, has := attrs["linguist-generated"]; has {
isGenerated = LinguistBoolAttrib{Value: generated}
}
if documentation, has := attrs["linguist-documentation"]; has {
isDocumentation = LinguistBoolAttrib{Value: documentation}
}
if detectable, has := attrs["linguist-detectable"]; has {
isDetectable = LinguistBoolAttrib{Value: detectable}
}
isVendored = attributeToBool(attrs, "linguist-vendored")
isGenerated = attributeToBool(attrs, "linguist-generated")
isDocumentation = attributeToBool(attrs, "linguist-documentation")
isDetectable = attributeToBool(attrs, "linguist-detectable")
if language, has := attrs["linguist-language"]; has && language != "unspecified" && language != "" {
// group languages, such as Pug -> HTML; SCSS -> CSS
group := enry.GetLanguageGroup(language)
@ -142,11 +142,11 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
}
}
if isDetectable.IsFalse() || isVendored.IsTrue() || isDocumentation.IsTrue() ||
(!isVendored.IsFalse() && analyze.IsVendor(f.Name())) ||
if isFalse(isDetectable) || isTrue(isVendored) || isTrue(isDocumentation) ||
(!isFalse(isVendored) && analyze.IsVendor(f.Name())) ||
enry.IsDotFile(f.Name()) ||
enry.IsConfiguration(f.Name()) ||
(!isDocumentation.IsFalse() && enry.IsDocumentation(f.Name())) {
(!isFalse(isDocumentation) && enry.IsDocumentation(f.Name())) {
continue
}
@ -179,7 +179,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
return nil, err
}
}
if !isGenerated.IsTrue() && enry.IsGenerated(f.Name(), content) {
if !isTrue(isGenerated) && enry.IsGenerated(f.Name(), content) {
continue
}
@ -201,7 +201,7 @@ func (repo *Repository) GetLanguageStats(commitID string) (map[string]int64, err
langType := enry.GetLanguageType(language)
included = langType == enry.Programming || langType == enry.Markup
if !included {
if isDetectable.IsTrue() {
if isTrue(isDetectable) {
included = true
} else {
continue