Commit remaining unstaged changes from this session
- Platform settings service: IPlatformSettingsService, PlatformSettingKeys, PlatformSettingsService, SubscriptionService, AppConstants, SubscriptionExpiryBackgroundService, SubscriptionMiddleware - JobTimeEntry entity, DTOs, AutoMapper profile (ShopWorker → UserId migration) - InventoryDtos: SourceTransactionId on PowderUsageLogDto - InventoryTransactionRepository: include Job.Customer in ledger query - InventoryAiLookupService: @graph unwrap + HTML price fallback - ApplicationDbContextModelSnapshot: reflect migration changes - launchSettings.json, publish profile Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -834,6 +834,10 @@ Rules:
|
||||
// machine-readable price, SKU, and product info that would otherwise be lost.
|
||||
var structuredData = ExtractJsonLdData(html);
|
||||
|
||||
// Extract visible price text BEFORE stripping HTML as a fallback when
|
||||
// JSON-LD is absent or incomplete (e.g. JS-rendered stores).
|
||||
var htmlPriceSnippet = ExtractHtmlPriceSnippet(html);
|
||||
|
||||
// Remove script/style blocks
|
||||
html = System.Text.RegularExpressions.Regex.Replace(
|
||||
html, @"<(script|style)[^>]*>[\s\S]*?</(script|style)>", "",
|
||||
@@ -851,10 +855,11 @@ Rules:
|
||||
if (text.Length > maxChars)
|
||||
text = text[..maxChars] + "…";
|
||||
|
||||
// Prepend structured data + document links — Claude treats these as high-confidence
|
||||
// Prepend structured data + document links + price fallback — Claude treats these as high-confidence
|
||||
var header = new StringBuilder();
|
||||
if (!string.IsNullOrWhiteSpace(structuredData)) header.Append(structuredData);
|
||||
if (!string.IsNullOrWhiteSpace(docLinks)) header.Append(docLinks);
|
||||
if (!string.IsNullOrWhiteSpace(structuredData)) header.Append(structuredData);
|
||||
if (!string.IsNullOrWhiteSpace(htmlPriceSnippet)) header.Append(htmlPriceSnippet);
|
||||
if (!string.IsNullOrWhiteSpace(docLinks)) header.Append(docLinks);
|
||||
if (header.Length > 0) text = header + "\n" + text;
|
||||
|
||||
_logger.LogInformation("Fetched {Chars} chars from {Url} (structured data: {HasData}, image: {HasImage})",
|
||||
@@ -897,6 +902,46 @@ Rules:
|
||||
return null;
|
||||
}
|
||||
|
||||
/// <summary>
|
||||
/// Scans raw HTML for visible price elements — common WooCommerce/Shopify price class
|
||||
/// names and itemprop="price" microdata — and returns them as a "[Page Price]" header
|
||||
/// line. This runs on the full, untruncated HTML before tag stripping so prices that
|
||||
/// would fall past the 3,500-char page-text cutoff are still surfaced to Claude.
|
||||
/// Only used when JSON-LD structured data didn't already yield a price.
|
||||
/// </summary>
|
||||
private static string? ExtractHtmlPriceSnippet(string html)
|
||||
{
|
||||
// itemprop="price" microdata (e.g. <span itemprop="price">12.99</span>)
|
||||
var microprice = System.Text.RegularExpressions.Regex.Match(
|
||||
html,
|
||||
@"itemprop=[""']price[""'][^>]*content=[""']([0-9]+\.?[0-9]*)[""']|" +
|
||||
@"itemprop=[""']price[""'][^>]*>[\s]*\$?([0-9]+\.?[0-9]*)",
|
||||
System.Text.RegularExpressions.RegexOptions.IgnoreCase);
|
||||
if (microprice.Success)
|
||||
{
|
||||
var val = (microprice.Groups[1].Value.Trim().Length > 0
|
||||
? microprice.Groups[1].Value
|
||||
: microprice.Groups[2].Value).Trim();
|
||||
if (!string.IsNullOrEmpty(val))
|
||||
return $"[Page Price] Price found in page microdata: ${val}\n";
|
||||
}
|
||||
|
||||
// WooCommerce / common e-commerce price class names
|
||||
var classPricePattern = System.Text.RegularExpressions.Regex.Match(
|
||||
html,
|
||||
@"class=[""'][^""']*(?:price|woocommerce-Price-amount|product-price|bdi)[^""']*[""'][^>]*>" +
|
||||
@"[\s\S]{0,60}?\$\s*([0-9]+\.[0-9]{2})",
|
||||
System.Text.RegularExpressions.RegexOptions.IgnoreCase);
|
||||
if (classPricePattern.Success)
|
||||
{
|
||||
var val = classPricePattern.Groups[1].Value.Trim();
|
||||
if (!string.IsNullOrEmpty(val))
|
||||
return $"[Page Price] Price found in page HTML: ${val}\n";
|
||||
}
|
||||
|
||||
return null;
|
||||
}
|
||||
|
||||
/// <summary>
|
||||
/// Scans raw HTML for anchor tags linking to SDS or TDS documents and returns them as
|
||||
/// "[Structured Data]" lines that Claude can read and echo back in its JSON response.
|
||||
@@ -959,6 +1004,12 @@ Rules:
|
||||
/// Extracts product name, SKU, and price from JSON-LD structured data blocks.
|
||||
/// Many e-commerce sites (Shopify, WooCommerce, etc.) embed this in the page HTML
|
||||
/// even when the visible price is rendered by JavaScript.
|
||||
/// <para>
|
||||
/// Handles three common JSON-LD shapes:
|
||||
/// - Single Product object: <c>{"@type":"Product",...}</c>
|
||||
/// - Top-level array: <c>[{"@type":"Product",...},...]</c>
|
||||
/// - WooCommerce/Yoast @graph wrapper: <c>{"@graph":[{"@type":"Product",...},...]}</c>
|
||||
/// </para>
|
||||
/// </summary>
|
||||
private static string? ExtractJsonLdData(string html)
|
||||
{
|
||||
@@ -976,12 +1027,7 @@ Rules:
|
||||
using var doc = JsonDocument.Parse(jsonText);
|
||||
var root = doc.RootElement;
|
||||
|
||||
// Handle both single-object and @graph array
|
||||
IEnumerable<JsonElement> nodes = root.ValueKind == JsonValueKind.Array
|
||||
? root.EnumerateArray()
|
||||
: new[] { root };
|
||||
|
||||
foreach (var node in nodes)
|
||||
foreach (var node in FlattenJsonLdNodes(root))
|
||||
{
|
||||
if (!node.TryGetProperty("@type", out var typeEl)) continue;
|
||||
|
||||
@@ -1013,6 +1059,34 @@ Rules:
|
||||
return sb.Length > 0 ? sb.ToString() : null;
|
||||
}
|
||||
|
||||
/// <summary>
|
||||
/// Recursively flattens a JSON-LD element into individual nodes for processing.
|
||||
/// Handles a top-level array, a single object, and the WooCommerce/Yoast @graph
|
||||
/// wrapper pattern where the root object has no @type but contains an "@graph" array.
|
||||
/// </summary>
|
||||
private static IEnumerable<JsonElement> FlattenJsonLdNodes(JsonElement el)
|
||||
{
|
||||
if (el.ValueKind == JsonValueKind.Array)
|
||||
{
|
||||
foreach (var item in el.EnumerateArray())
|
||||
foreach (var n in FlattenJsonLdNodes(item))
|
||||
yield return n;
|
||||
}
|
||||
else if (el.ValueKind == JsonValueKind.Object)
|
||||
{
|
||||
// @graph wrapper: {"@graph": [...]} — recurse into the array
|
||||
if (el.TryGetProperty("@graph", out var graph) && graph.ValueKind == JsonValueKind.Array)
|
||||
{
|
||||
foreach (var n in FlattenJsonLdNodes(graph))
|
||||
yield return n;
|
||||
}
|
||||
else
|
||||
{
|
||||
yield return el;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/// <summary>
|
||||
/// Dispatches JSON-LD offer extraction for both the single-object form
|
||||
/// (<c>"offers": { ... }</c>) and the array form (<c>"offers": [ ... ]</c>)
|
||||
|
||||
Reference in New Issue
Block a user