jriou/restockbot
Archived
1
0
Fork 0
Code Issues Pull requests Activity

fix: exclude cybertek home page (#33)

Browse source 
Signed-off-by: Julien Riou <julien@riou.xyz>
This commit is contained in:
Julien Riou 2021-07-12 08:50:36 +02:00
parent bf1a389704
commit 2c0b972849
No known key found for this signature in database
GPG key ID: FF42D23B580C89F7
1 changed files with 4 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
parser_url.go
View file

@ -214,6 +214,7 @@ func createQueryForCybertek(url string) string {
// gather first page // gather first page
LET first_page = '` + url + `' LET first_page = '` + url + `'
LET doc = DOCUMENT(first_page, {driver: "cdp"}) LET doc = DOCUMENT(first_page, {driver: "cdp"})
LET home_page = 'https://www.cybertek.fr/boutique/index.aspx'
// discover next pages // discover next pages
LET pagination = ELEMENT(doc, "div .pagination-div") LET pagination = ELEMENT(doc, "div .pagination-div")
@ -232,6 +233,7 @@ LET results = (
LET products_available = ( LET products_available = (
FOR el IN ELEMENTS(doc, "div .listing_dispo") FOR el IN ELEMENTS(doc, "div .listing_dispo")
LET url = ELEMENT(el, "a") LET url = ELEMENT(el, "a")
FILTER url.attributes.href != home_page // exclude home page
LET name = TRIM(FIRST(SPLIT(INNER_TEXT(ELEMENT(el, "div .height-txt-cat")), "-"))) LET name = TRIM(FIRST(SPLIT(INNER_TEXT(ELEMENT(el, "div .height-txt-cat")), "-")))
LET price = TO_FLOAT(SUBSTITUTE(INNER_TEXT(ELEMENT(el, "div .price_prod_resp")), "€", ".")) LET price = TO_FLOAT(SUBSTITUTE(INNER_TEXT(ELEMENT(el, "div .price_prod_resp")), "€", "."))
RETURN { RETURN {
@ -245,6 +247,7 @@ LET results = (
LET products_not_available = ( LET products_not_available = (
FOR el IN ELEMENTS(doc, "div .listing_nodispo") FOR el IN ELEMENTS(doc, "div .listing_nodispo")
LET url = ELEMENT(el, "a") LET url = ELEMENT(el, "a")
FILTER url.attributes.href != home_page // exclude home page
LET name = TRIM(FIRST(SPLIT(INNER_TEXT(ELEMENT(el, "div .height-txt-cat")), "-"))) LET name = TRIM(FIRST(SPLIT(INNER_TEXT(ELEMENT(el, "div .height-txt-cat")), "-")))
LET price = TO_FLOAT(SUBSTITUTE(INNER_TEXT(ELEMENT(el, "div .price_prod_resp")), "€", ".")) LET price = TO_FLOAT(SUBSTITUTE(INNER_TEXT(ELEMENT(el, "div .price_prod_resp")), "€", "."))
RETURN { RETURN {
@ -261,7 +264,7 @@ LET results = (
// combine all arrays to a single one // combine all arrays to a single one
RETURN FLATTEN(results) RETURN FLATTEN(results)
` `
return q return q
} }