fenix/webarchive
1
0
Fork 0
mirror of https://github.com/derfenix/webarchive.git synced 2026-03-11 21:35:34 +03:00
Code Issues Packages Projects Releases Wiki Activity

Improved single_file processor, refactoring

Browse Source 
Reduce inlined image size, get page metadata before save and put into processing queue
This commit is contained in:
derFenix
2023-11-24 14:25:17 +03:00
parent 7e53519ca0
commit 870f13f7bf
14 changed files with 325 additions and 96 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
adapters/processors/pdf.go
View File

@@ -40,9 +40,9 @@ func (p *PDF) Process(_ context.Context, url string, cache *entity.Cache) ([]ent
opts := wkhtmltopdf.NewPageOptions()
opts.PrintMediaType.Set(p.cfg.MediaPrint)
opts.JavascriptDelay.Set(200)
opts.DisableJavascript.Set(true)
opts.DisableJavascript.Set(false)
opts.LoadErrorHandling.Set("ignore")
opts.LoadMediaErrorHandling.Set("ignore")
opts.LoadMediaErrorHandling.Set("skip")
opts.FooterRight.Set("[opts]")
opts.HeaderLeft.Set(url)
opts.HeaderRight.Set(time.Now().Format(time.DateOnly))
@@ -50,9 +50,9 @@ func (p *PDF) Process(_ context.Context, url string, cache *entity.Cache) ([]ent
opts.Zoom.Set(p.cfg.Zoom)
opts.ViewportSize.Set(p.cfg.Viewport)
opts.NoBackground.Set(true)
opts.DisableLocalFileAccess.Set(true)
opts.DisableExternalLinks.Set(true)
opts.DisableInternalLinks.Set(true)
opts.DisableLocalFileAccess.Set(false)
opts.DisableExternalLinks.Set(false)
opts.DisableInternalLinks.Set(false)
var page wkhtmltopdf.PageProvider
if len(cache.Get()) > 0 {