Public-Mirror/meta-rules-dat
1
0
Fork 0
mirror of https://github.com/MetaCubeX/meta-rules-dat.git synced 2024-12-28 19:24:14 +08:00
Code Issues Projects Releases Wiki Activity

Remove redundant domains

Browse Source
This commit is contained in:
loyalsoldier 2020-03-02 01:05:01 +08:00
parent 1e4c41d5ae
commit bd0e1a3d7b
1 changed files with 17 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
.github/workflows/build.yml vendored
View File

@ -93,15 +93,28 @@ jobs:
cat proxy.txt >> temp-proxy.txt
cat direct.txt >> temp-direct.txt
- name: Sort and generate lists
- name: Sort and generate redundant lists
run: |
cat temp-proxy.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > $GOPATH/src/$GEOSITE_REPO/data/proxy-list
cat temp-proxy.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > proxy-list-with-redundant
cat temp-proxy.txt | sort --ignore-case -u | perl -ne 'print if not /^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/' > proxy-excluse-list
cat temp-direct.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > $GOPATH/src/$GEOSITE_REPO/data/direct-list
cat temp-direct.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > direct-list-with-redundant
cat temp-direct.txt | sort --ignore-case -u | perl -ne 'print if not /^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/' > direct-excluse-list
cat temp-reject.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > $GOPATH/src/$GEOSITE_REPO/data/reject-list
cat temp-reject.txt | sort --ignore-case -u | perl -ne '/^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/ && print "$1\n"' > reject-list-with-redundant
cat temp-reject.txt | sort --ignore-case -u | perl -ne 'print if not /^((?=^.{3,255})[a-zA-Z0-9][-_a-zA-Z0-9]{0,62}(\.[a-zA-Z0-9][-_a-zA-Z0-9]{0,62})+)/' > reject-excluse-list
- name: Remove redundant domains
run: |
chmod +x findRedundantDomain.py
./findRedundantDomain.py ./direct-list-with-redundant ./direct-list-deleted-unsort
./findRedundantDomain.py ./proxy-list-with-redundant ./proxy-list-deleted-unsort
./findRedundantDomain.py ./reject-list-with-redundant ./reject-list-deleted-unsort
sort ./direct-list-deleted-unsort > ./direct-list-deleted-sort
sort ./proxy-list-deleted-unsort > ./proxy-list-deleted-sort
sort ./reject-list-deleted-unsort > ./reject-list-deleted-sort
diff ./direct-list-deleted-sort ./direct-list-with-redundant | awk '/^>/{print $2}' > $GOPATH/src/$GEOSITE_REPO/data/direct-list
diff ./proxy-list-deleted-sort ./proxy-list-with-redundant | awk '/^>/{print $2}' > $GOPATH/src/$GEOSITE_REPO/data/proxy-list
diff ./reject-list-deleted-sort ./reject-list-with-redundant | awk '/^>/{print $2}' > $GOPATH/src/$GEOSITE_REPO/data/reject-list
- name: Add list into appropriate category file
run: |
cd $GOPATH/src/$GEOSITE_REPO/data