admin/spray
1
0
Fork 0
mirror of https://github.com/chainreactors/spray.git synced 2025-09-15 11:40:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

调整爬虫的正则

Browse Source
This commit is contained in:
M09Ic 2023-01-11 11:12:40 +08:00
parent 4a774718c9
commit 8233dcefd3
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
pkg/utils.go
View File

@ -23,14 +23,14 @@ var (
ActivePath []string
Fingers fingers.Fingers
JSRegexps []*regexp.Regexp = []*regexp.Regexp{
regexp.MustCompile(`.(https{0,1}:[^\s',’"”><;()|*\[]{2,250}?[^=*\s'’><:;|()[]{3}\[]\.js)`),
regexp.MustCompile(`.(https{0,1}:[^\s'’"”><;()|*\[]{2,250}?[^=*\s'’><:;|()[]{3}\[]\.js)`),
regexp.MustCompile(`["']\s{0,6}(/{0,1}[^\s',’"”><;()|*:\[]{2,250}?[^=*\s'’|"”><^:;()\[]{3}\.\.js)`),
regexp.MustCompile(`=\s{0,6}["']{0,1}\s{0,6}(/{0,1}[^\s^',’><;()|*\[]{2,250}?[^=,\s'’"”>|<:;*()\[]{3}\.js)`),
}
URLRegexps []*regexp.Regexp = []*regexp.Regexp{
regexp.MustCompile(`=\s{0,6}(https{0,1}:[^\s',’"”><;()|*\[]{2,250})`),
regexp.MustCompile(`["']([^\s',’"”><.@;()|*\[]{2,250}\.[a-zA-Z]\w{1,4})["']`),
regexp.MustCompile(`["'](https?:[^\s',’"”><;()@|*\[]{2,250}?\.[^\s',’"”><;()|*\[]{2,250}?)["']`),
regexp.MustCompile(`=\s{0,6}(https{0,1}:[^\s'’"”><;()|*\[]{2,250})`),
regexp.MustCompile(`["']([^\s',’"”><.@;:()|*\[]{2,250}\.[a-zA-Z]\w{1,4})["']`),
regexp.MustCompile(`["'](https?:[^\s'’"”><;()@|*\[]{2,250}?\.[^\s',’"”><;()|*\[]{2,250}?)["']`),
regexp.MustCompile(`["']\s{0,6}([#,.]{0,2}/[^\s',’"”><;()|*\[]{2,250}?)\s{0,6}["']`),
regexp.MustCompile(`href\s{0,6}=\s{0,6}["'‘“]{0,1}\s{0,6}([^\s',’"”><;()|*\[]{2,250})|action\s{0,6}=\s{0,6}["'‘“]{0,1}\s{0,6}([^\s'’"“><)(]{2,250})`),
}