4a9f992056
Signed-off-by: Michael Mayer <michael@liquidbytes.net>
124 lines
3.7 KiB
Go
124 lines
3.7 KiB
Go
package thumb
|
|
|
|
import "github.com/disintegration/imaging"
|
|
|
|
var (
|
|
Size = 2048
|
|
Limit = 4096
|
|
Filter = ResampleLanczos
|
|
JpegQuality = 95
|
|
JpegQualitySmall = 80
|
|
)
|
|
|
|
func MaxSize() int {
|
|
if Size > Limit {
|
|
return Size
|
|
}
|
|
|
|
return Limit
|
|
}
|
|
|
|
func InvalidSize(size int) bool {
|
|
return size < 0 || size > MaxSize()
|
|
}
|
|
|
|
const (
|
|
ResampleBlackman ResampleFilter = "blackman"
|
|
ResampleLanczos ResampleFilter = "lanczos"
|
|
ResampleCubic ResampleFilter = "cubic"
|
|
ResampleLinear ResampleFilter = "linear"
|
|
)
|
|
|
|
type ResampleFilter string
|
|
|
|
func (a ResampleFilter) Imaging() imaging.ResampleFilter {
|
|
switch a {
|
|
case ResampleBlackman:
|
|
return imaging.Blackman
|
|
case ResampleLanczos:
|
|
return imaging.Lanczos
|
|
case ResampleCubic:
|
|
return imaging.CatmullRom
|
|
case ResampleLinear:
|
|
return imaging.Linear
|
|
default:
|
|
return imaging.Lanczos
|
|
}
|
|
}
|
|
|
|
const (
|
|
ResampleFillCenter ResampleOption = iota
|
|
ResampleFillTopLeft
|
|
ResampleFillBottomRight
|
|
ResampleFit
|
|
ResampleResize
|
|
ResampleNearestNeighbor
|
|
ResampleDefault
|
|
ResamplePng
|
|
)
|
|
|
|
type ResampleOption int
|
|
|
|
var ResampleMethods = map[ResampleOption]string{
|
|
ResampleFillCenter: "center",
|
|
ResampleFillTopLeft: "left",
|
|
ResampleFillBottomRight: "right",
|
|
ResampleFit: "fit",
|
|
ResampleResize: "resize",
|
|
}
|
|
|
|
type Type struct {
|
|
Use string `json:"use"`
|
|
Source string `json:"-"`
|
|
Width int `json:"w"`
|
|
Height int `json:"h"`
|
|
Public bool `json:"-"`
|
|
Options []ResampleOption `json:"-"`
|
|
}
|
|
|
|
type TypeMap map[string]Type
|
|
|
|
var Types = TypeMap{
|
|
"tile_50": {"Lists", "tile_500", 50, 50, false, []ResampleOption{ResampleFillCenter, ResampleDefault}},
|
|
"tile_100": {"Maps", "tile_500", 100, 100, false, []ResampleOption{ResampleFillCenter, ResampleDefault}},
|
|
"tile_224": {"TensorFlow, Mosaic", "tile_500", 224, 224, false, []ResampleOption{ResampleFillCenter, ResampleDefault}},
|
|
"tile_500": {"Tiles", "", 500, 500, false, []ResampleOption{ResampleFillCenter, ResampleDefault}},
|
|
"colors": {"Color Detection", "fit_720", 3, 3, false, []ResampleOption{ResampleResize, ResampleNearestNeighbor, ResamplePng}},
|
|
"left_224": {"TensorFlow", "fit_720", 224, 224, false, []ResampleOption{ResampleFillTopLeft, ResampleDefault}},
|
|
"right_224": {"TensorFlow", "fit_720", 224, 224, false, []ResampleOption{ResampleFillBottomRight, ResampleDefault}},
|
|
"fit_720": {"Mobile, TV", "", 720, 720, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_1280": {"Mobile, HD Ready TV", "fit_2048", 1280, 1024, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_1920": {"Mobile, Full HD TV", "fit_2048", 1920, 1200, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_2048": {"Tablets, Cinema 2K", "", 2048, 2048, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_2560": {"Quad HD, Retina Display", "", 2560, 1600, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_3840": {"Ultra HD", "", 3840, 2400, false, []ResampleOption{ResampleFit, ResampleDefault}}, // Deprecated in favor of fit_4096
|
|
"fit_4096": {"Ultra HD, Retina 4K", "", 4096, 4096, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
"fit_7680": {"8K Ultra HD 2, Retina 6K", "", 7680, 4320, true, []ResampleOption{ResampleFit, ResampleDefault}},
|
|
}
|
|
|
|
var DefaultTypes = []string{
|
|
"fit_7680",
|
|
"fit_4096",
|
|
"fit_2560",
|
|
"fit_2048",
|
|
"fit_1920",
|
|
"fit_1280",
|
|
"fit_720",
|
|
"right_224",
|
|
"left_224",
|
|
"colors",
|
|
"tile_500",
|
|
"tile_224",
|
|
"tile_100",
|
|
"tile_50",
|
|
}
|
|
|
|
// Returns true if thumbnail is too large and can not be rendered at all.
|
|
func (t Type) ExceedsLimit() bool {
|
|
return t.Width > MaxSize() || t.Height > MaxSize()
|
|
}
|
|
|
|
// Returns true if thumbnail type should not be pre-rendered.
|
|
func (t Type) OnDemand() bool {
|
|
return t.Width > Size || t.Height > Size
|
|
}
|