Applicants: auto-tags + deep search w/ highlight; never delete (archive instead)
CI/CD / CI · dotnet build (push) Successful in 2m1s
CI/CD / Deploy · hamkadr (push) Successful in 2m36s

- Tags: parser extracts cert/skill keywords (mmt, ICU/CCU, دیالیز, اتاق عمل,
  اورژانس, مسئول فنی, پروانه‌دار…) + role + city into TalentListing.Tags
  (+ migration); shown as chips on cards.
- Deep search on /Talent: «جستجوی عمیق» box does Postgres ILIKE across
  tags, description, person, area, role, city (every term must match);
  matches are highlighted with <mark> via SearchHighlight.
- Never delete: ShiftStatus.Archived + the admin «بایگانی گروهی» action now
  ARCHIVES aggregated posts (hidden from site, kept in DB) and leaves the
  raw crawl rows intact — a permanent archive for future analytics.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
soroush.asadi
2026-06-08 11:25:32 +03:30
parent e4dc5180ad
commit 6b657c7795
15 changed files with 1752 additions and 22 deletions
@@ -29,6 +29,7 @@ public class ParsedListing
public bool IsLicensed { get; set; } // پروانه‌دار
public string? AreaNote { get; set; } // «فقط منطقه ۱»
public List<ParsedContact> Contacts { get; set; } = new(); // phones, email, socials…
public List<string> Tags { get; set; } = new(); // cert/skill keywords for search
public List<string> Notes { get; set; } = new(); // what was/wasn't detected (shown to admin)
}
@@ -170,6 +171,12 @@ public class HeuristicListingParser : IListingParser
if (p.FacilityName is not null) p.Notes.Add($"مرکز: {p.FacilityName}");
}
// --- Tags (certs/skills for deep search): mmt, icu, پروانه‌دار, اتاق عمل … ---
p.Tags = ExtractTags(text);
if (p.RoleNames.Count > 0) p.Tags.AddRange(p.RoleNames);
if (p.IsLicensed && !p.Tags.Contains("پروانه‌دار")) p.Tags.Add("پروانه‌دار");
p.Tags = p.Tags.Distinct().ToList();
// --- Contacts (phones, email, socials — one ad may have several) ---
p.Contacts = ExtractContacts(raw ?? text);
p.Phone = p.Contacts.FirstOrDefault(c => c.Type is ContactType.Mobile or ContactType.Phone)?.Value;
@@ -355,6 +362,40 @@ public class HeuristicListingParser : IListingParser
return list.Take(8).ToList();
}
// Canonical tag → trigger words found in the post.
private static readonly (string Tag, string[] Needles)[] TagDict =
{
("mmt", new[] { "mmt", "ام ام تی", "ام‌ام‌تی" }),
("ICU", new[] { "icu", "آی سی یو", "آی‌سی‌یو" }),
("CCU", new[] { "ccu", "سی سی یو", "سی‌سی‌یو" }),
("NICU", new[] { "nicu", "ان آی سی یو", "نوزادان" }),
("BLS", new[] { "bls" }),
("ACLS", new[] { "acls" }),
("دیالیز", new[] { "دیالیز" }),
("اتاق عمل", new[] { "اتاق عمل", "اسکراب" }),
("بیهوشی", new[] { "بیهوشی" }),
("تریاژ", new[] { "تریاژ" }),
("تزریقات", new[] { "تزریقات", "تزریق" }),
("پانسمان", new[] { "پانسمان", "زخم" }),
("سونوگرافی", new[] { "سونوگرافی" }),
("رادیولوژی", new[] { "رادیولوژی" }),
("اورژانس", new[] { "اورژانس", "فوریت" }),
("مسئول فنی", new[] { "مسئول فنی" }),
("طرح", new[] { "طرح" }),
("سالمند", new[] { "سالمند" }),
("کودک", new[] { "کودک", "اطفال" }),
("همراه بیمار", new[] { "همراه بیمار" }),
("پروانه‌دار", new[] { "پروانه" }),
};
private static List<string> ExtractTags(string text)
{
var tags = new List<string>();
foreach (var (tag, needles) in TagDict)
if (ContainsAny(text, needles)) tags.Add(tag);
return tags;
}
private static string UrlHandle(string url)
{
var u = url.Split('?')[0].TrimEnd('/');
@@ -205,6 +205,7 @@ public class IngestionService
Description = raw.RawText,
Status = ShiftStatus.Open, Source = ShiftSource.Aggregated, SourceUrl = raw.SourceUrl,
Contacts = BuildContacts(d, parsed), // fresh instances per listing
Tags = BuildTags(parsed, role, city),
});
raw.Status = RawListingStatus.Normalized;
return;
@@ -260,6 +261,13 @@ public class IngestionService
raw.Status = RawListingStatus.Normalized;
}
/// <summary>Space-separated searchable tags: parsed cert/skill tags + this listing's role + city.</summary>
private static string BuildTags(ParsedListing parsed, Role role, City city)
{
var tags = new List<string>(parsed.Tags) { role.Name, city.Name };
return string.Join(" ", tags.Where(t => !string.IsNullOrWhiteSpace(t)).Distinct());
}
/// <summary>Fresh ContactMethod rows for one talent listing (parser contacts + AI phone).</summary>
private static List<ContactMethod> BuildContacts(AiStructured? d, ParsedListing parsed)
{
@@ -0,0 +1,27 @@
using System.Net;
using System.Text.RegularExpressions;
using Microsoft.AspNetCore.Html;
namespace JobsMedical.Web.Services;
/// <summary>Wraps query terms in &lt;mark&gt; for result highlighting (HTML-safe).</summary>
public static class SearchHighlight
{
public static HtmlString Mark(string? text, string? query)
{
if (string.IsNullOrEmpty(text)) return HtmlString.Empty;
var encoded = WebUtility.HtmlEncode(text);
if (string.IsNullOrWhiteSpace(query)) return new HtmlString(encoded);
var terms = query.Split(' ', StringSplitOptions.RemoveEmptyEntries | StringSplitOptions.TrimEntries)
.Where(t => t.Length >= 2)
.Select(t => Regex.Escape(WebUtility.HtmlEncode(t)))
.Distinct()
.ToList();
if (terms.Count == 0) return new HtmlString(encoded);
var pattern = string.Join("|", terms);
var marked = Regex.Replace(encoded, pattern, m => $"<mark>{m.Value}</mark>", RegexOptions.IgnoreCase);
return new HtmlString(marked);
}
}