steve/go-extractor
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Refine selectors and enhance price parsing with logging

Browse Source 
Adjusted HTML selectors for improved compatibility and updated price parsing logic to handle additional formats. Added logging to provide better debugging insights during price extraction.
This commit is contained in:
Steve Dudenhoeffer
2025-10-10 14:42:01 -04:00
parent dc43d1626a
commit 9947cae947
2 changed files with 11 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
sites/aislegopher/aislegopher.go
View File

@@ -62,13 +62,13 @@ func (c Config) GetItemFromURL(ctx context.Context, b extractor.Browser, u *url.
return res, fmt.Errorf("failed to open page: %w", err)
}
names := doc.Select("h2.h4")
names := doc.Select(".h4")
if len(names) > 0 {
res.Name, _ = names[0].Text()
}
prices := doc.Select("h4.h2")
prices := doc.Select(".h2")
if len(prices) > 0 {
priceStr, _ := prices[0].Text()

12
sites/wegmans/wegmans.go
View File

@@ -4,6 +4,7 @@ import (
"context"
"errors"
"io"
"log/slog"
"net/url"
"strconv"
"strings"
@@ -80,23 +81,28 @@ func (c Config) GetItemPrice(ctx context.Context, b extractor.Browser, u *url.UR
ID: id,
}
titles := doc.Select("h1[data-test]")
titles := doc.Select("h1[data-testid]")
if len(titles) != 0 {
res.Name, _ = titles[0].Text()
}
prices := doc.Select("span[data-test=\"amount\"] span:nth-child(1)")
prices := doc.Select("div.component--product-price:nth-child(1) > div:nth-child(1) > span:nth-child(1) > span:nth-child(2)")
slog.Info("prices", "len", len(prices))
if len(prices) != 0 {
priceStr, _ := prices[0].Text()
slog.Info("price", "0", prices[0], "text", priceStr)
priceStr = strings.ReplaceAll(priceStr, "$", "")
priceStr = strings.ReplaceAll(priceStr, ",", "")
// if there's a "/" in the price, then it's in the format of like "1.99/ea", so split it off
priceStr = strings.Split(priceStr, "/")[0]
price, _ := strconv.ParseFloat(priceStr, 64)
slog.Info("price", "0", prices[0], "text", priceStr, "price", price)
res.Price = price
}
unitPrices := doc.Select(`span[data-test="per-unit-price"]`)
unitPrices := doc.Select(`div.component--product-price:nth-child(1) span.price-per-unit`)
if len(unitPrices) != 0 {
unitPriceStr, _ := unitPrices[0].Text()