File tree Expand file tree Collapse file tree
Expand file tree Collapse file tree Original file line number Diff line number Diff line change 1+ name : spider-check
2+ on :
3+ # Once a week on Friday at 00:00
4+ schedule :
5+ - cron : ' 0 0 * * 5'
6+ # Or manually
7+ workflow_dispatch :
8+ # Or when developing this workflow
9+ push :
10+ paths :
11+ - .github/workflows/spider-check.yaml
12+ pull_request :
13+ paths :
14+ - .github/workflows/spider-check.yaml
15+
16+ jobs :
17+ spider-check :
18+ env :
19+ MY_SITE : https://api.jquery.com
20+ runs-on : ubuntu-latest
21+ if : ${{ github.repository_owner == 'jquery' }} # skip on forks
22+ steps :
23+ - uses : actions/checkout@v2
24+ with :
25+ repository : jquery/hydra-link-checker
26+ ref : v2.0.0
27+ - name : Run hydra-link-checker
28+ run : python3 hydra.py "$MY_SITE"
Original file line number Diff line number Diff line change 1+ name : typesense
2+ on :
3+ # Once a day at 11:30 UTC <https://crontab.guru/>
4+ schedule :
5+ - cron : ' 30 11 * * *'
6+ # Or after a deployment
7+ push :
8+ branches :
9+ - main
10+ # Or manually
11+ workflow_dispatch :
12+
13+ jobs :
14+ typesense :
15+ name : Update Typesense
16+ if : ${{ github.repository_owner == 'jquery' }} # skip on forks
17+ runs-on : ubuntu-latest
18+ steps :
19+ - uses : actions/checkout@v3
20+ - name : Docsearch Scraper
21+ shell : bash
22+ run : |
23+ # 0.6.0.rc1: https://github.com/typesense/typesense-docsearch-scraper/issues/36
24+ docker run \
25+ -e TYPESENSE_API_KEY=${{ secrets.TYPESENSE_ADMIN_KEY }} \
26+ -e TYPESENSE_HOST="${{ secrets.TYPESENSE_HOST }}" \
27+ -e TYPESENSE_PORT="443" \
28+ -e TYPESENSE_PROTOCOL="https" \
29+ -e CONFIG="$(cat docsearch.config.json | jq -r tostring)" \
30+ typesense/docsearch-scraper:0.6.0.rc2
Load Diff This file was deleted.
Original file line number Diff line number Diff line change 1+ {
2+ "index_name" : " jquery_com" ,
3+ "start_urls" : [
4+ { "url" : " https://jquery.com" },
5+ { "url" : " https://api.jquery.com" , "selectors_key" : " api" }
6+ ],
7+ "stop_urls" : [
8+ " .com/category/"
9+ ],
10+ "selectors" : {
11+ "default" : {
12+ "lvl0" : {
13+ "selector" : " #menu-top .menu-item.current > a" ,
14+ "global" : true ,
15+ "default_value" : " Documentation"
16+ },
17+ "lvl1" : " #content h1" ,
18+ "lvl2" : " #content h2" ,
19+ "lvl3" : " #content h3" ,
20+ "lvl4" : " #content h4" ,
21+ "lvl5" : " #content h5" ,
22+ "text" : " #content p, #content li, #content tr"
23+ },
24+ "api" : {
25+ "lvl0" : {
26+ "selector" : " #categories .cat-item.current-cat > a" ,
27+ "global" : true ,
28+ "default_value" : " API"
29+ },
30+ "lvl1" : " #content h1" ,
31+ "lvl2" : " #content h2, #content h4.name" ,
32+ "lvl3" : " #content h3, #content h4:not(.name)" ,
33+ "lvl4" : " #content h5, #content strong:first-child" ,
34+ "text" : " .entry-content p, .entry-content li"
35+ }
36+ },
37+ "selectors_exclude" : [
38+ " header ~ article" ,
39+ " .returns" ,
40+ " .version-details" ,
41+ " .section-title" ,
42+ " .icon-link.toc-link" ,
43+ " [class^=toclevel]" ,
44+ " #toctitle"
45+ ],
46+ "min_indexed_level" : 2 ,
47+ "scrape_start_urls" : false
48+ }
You can’t perform that action at this time.
0 commit comments