alpcentaur/wd_importPEP
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

added first syntax of general settings yaml per country

Browse source
This commit is contained in:
corsaronero 2023-02-24 18:29:55 +00:00
parent 2e46df3260
commit b4223bca30
1 changed files with 15 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
crawlers/countries.yaml Normal file
View file

@ -0,0 +1,15 @@
# Settings for the PEP crawler per country to crawl
# Follow the syntax
nicaragua:
memberList:
link: http://legislacion.asamblea.gob.ni/Tablas%20Generales.nsf/Main.xsp
parent: [html, body, form, table, tbody, tr, td, table, tbody]
child-name: [html, body, form, table, tbody, tr, td, table, tbody, tr, td.null, a.text]
child-link: [html, body, form, table, tbody, tr, td, table, tbody, tr, td.null, a.href]
member:
info-1:
parent: [html, body, form, table, tbody]
child-name: [html, body, form, table, tbody, tr.0, td.1, span]
child-image: [html, body, form, table, tbody, tr.1, td.0, span, img]
child-role: [html, body, form, table, tbody, tr.1, td.2, span + label.1]