diff options
Diffstat (limited to 'skate/url.go')
| -rw-r--r-- | skate/url.go | 6 |
1 files changed, 4 insertions, 2 deletions
diff --git a/skate/url.go b/skate/url.go index 66b9312..b104651 100644 --- a/skate/url.go +++ b/skate/url.go @@ -15,7 +15,9 @@ var ( ) // SanitizeURL applies various cleanup rules on URLs as found in references. -// XXX: Sometimes a URL contains other identifying information, like: +// Returns an empty string when no URL could be constructed. Still, many +// results will not be a URL after all. XXX: Sometimes a URL contains other +// identifying information, like: // http://agingmind.utdallas.edu/facedb/view/neutral-faces.doi:10.3758/BF03206543. // http://10.3386/w20634https://doi.org/10.3386/w20634 func SanitizeURL(s string) string { @@ -24,7 +26,7 @@ func SanitizeURL(s string) string { index = suffixarray.New([]byte(s)) indices = index.Lookup([]byte("http"), -1) ) - if !strings.HasPrefix(s, "http") { + if !strings.HasPrefix(s, "http") && !strings.HasPrefix(s, "ftp") { s = sanitizeRaw(s, index) if s == "" { return s |