Il Corsaro Nero: improve parsing (#2792)

Little modify to better recognize full seasons
Little modify to clear results
This commit is contained in:
Jorman 2018-03-25 17:23:11 +02:00 committed by kaso17
parent 257295ed08
commit 21f8fa966d
1 changed files with 6 additions and 4 deletions

View File

@ -64,12 +64,12 @@
- name: split
args: [ "/", -1 ]
- name: urldecode
- name: replace
args: [ "_", " "]
- name: replace
args: [ " ", " "]
- name: re_replace
args: [ "_+", " "]
- name: replace
args: [ ".", " "]
- name: re_replace
args: [ "\\s{2,}", " "]
# normalize to SXXEYY format
- name: re_replace
args: ["(\\d{2})x(\\d{2})", "S$1E$2"]
@ -88,6 +88,8 @@
- name: re_replace #remove various
args: ["(Serie completa|Completa|\\[in pausa\\])", ""]
# fine prova
- name: re_replace #try to find multi episode
args: ["(S\\d{2}E\\d{2})\\s(\\d{2})", "$1-$2"]
category:
selector: td:nth-child(1) a
attribute: href