feat(ops): automate backups and private CV benchmarks

JobTrackerApi.Tests/CvCorpusHarnessTests.cs

@@ -11,6 +11,7 @@ using Microsoft.AspNetCore.Identity;
 using Microsoft.AspNetCore.Mvc;
 using Microsoft.Extensions.Configuration;
 using Microsoft.Extensions.Hosting;
+using Microsoft.Extensions.Caching.Memory;
 using Moq;
 using Xunit;
 
@@ -52,24 +53,18 @@ public sealed class CvCorpusHarnessTests
         userManager.Setup(x => x.GetUserAsync(It.IsAny<System.Security.Claims.ClaimsPrincipal>())).ReturnsAsync(user);
         userManager.Setup(x => x.UpdateAsync(It.IsAny<ApplicationUser>())).ReturnsAsync(IdentityResult.Success);
 
-        var aiService = new Mock<ISummarizerService>();
-        aiService.Setup(x => x.SummarizeSectionAsync(It.Is<string>(instruction => instruction.Contains("Extract this CV into structured JSON", StringComparison.Ordinal)), It.IsAny<string>(), 3200, 900)).ReturnsAsync(string.Empty);
-        aiService.Setup(x => x.SummarizeSectionAsync(It.Is<string>(instruction => instruction.Contains("Reconstruct this CV text extracted from a PDF", StringComparison.Ordinal)), It.IsAny<string>(), 2800, 900)).ReturnsAsync((string _, string text, int _, int __) => text);
+        var aiService = CreateSummarizerFromEnvironment();
         var cvAiNormalizer = CreateCvAiNormalizerFromEnvironment();
 
         await using var db = TestHostFactory.CreateInMemoryDb();
         var paths = CreatePaths(outputRoot);
-        var controller = new ProfileCvController(userManager.Object, aiService.Object, db, paths, null, NoOpCvAiClassifier.Instance, cvAiNormalizer)
+        var controller = new ProfileCvController(userManager.Object, aiService, db, paths, null, NoOpCvAiClassifier.Instance, cvAiNormalizer)
         {
             ControllerContext = new ControllerContext { HttpContext = new DefaultHttpContext() }
         };
 
-        var extractMethod = typeof(ProfileCvController).GetMethod("ExtractTextAsync", BindingFlags.NonPublic | BindingFlags.Static);
-        var reconstructMethod = typeof(ProfileCvController).GetMethod("MaybeReconstructStructuredCvAsync", BindingFlags.NonPublic | BindingFlags.Instance);
-        var buildMethod = typeof(ProfileCvController).GetMethod("BuildStructuredCvAsync", BindingFlags.NonPublic | BindingFlags.Instance);
+        var extractMethod = typeof(ProfileCvController).GetMethod("ExtractStructuredCvFromFileAsync", BindingFlags.NonPublic | BindingFlags.Instance);
         Assert.NotNull(extractMethod);
-        Assert.NotNull(reconstructMethod);
-        Assert.NotNull(buildMethod);
 
         var entries = new List<CvBenchmarkEntry>();
         foreach (var path in files)
@@ -83,16 +78,14 @@ public sealed class CvCorpusHarnessTests
             };
 
             var extension = Path.GetExtension(path);
-            var extractTask = (Task<string>)extractMethod!.Invoke(null, new object[] { formFile, extension })!;
-            var text = await extractTask;
+            var extractTask = (Task)extractMethod!.Invoke(controller, new object[] { formFile, extension, CancellationToken.None })!;
+            await extractTask;
+            var extraction = extractTask.GetType().GetProperty("Result")!.GetValue(extractTask)!;
+            var resultType = extraction.GetType();
+            var text = (string)resultType.GetProperty("RawText")!.GetValue(extraction)!;
             Assert.False(string.IsNullOrWhiteSpace(text));
-
-            var reconstructTask = (Task<string>)reconstructMethod!.Invoke(controller, new object[] { text, CancellationToken.None })!;
-            var normalizedText = await reconstructTask;
-            Assert.False(string.IsNullOrWhiteSpace(normalizedText));
-
-            var buildTask = (Task<StructuredCvProfile>)buildMethod!.Invoke(controller, new object[] { normalizedText, CancellationToken.None })!;
-            var structured = StructuredCvProfileJson.Normalize(await buildTask);
+            var structured = StructuredCvProfileJson.Normalize(
+                (StructuredCvProfile)resultType.GetProperty("StructuredCv")!.GetValue(extraction)!);
             Assert.NotNull(structured);
 
             var slug = Slugify(fileName);
@@ -382,9 +375,44 @@ public sealed class CvCorpusHarnessTests
         {
             client.BaseAddress = new Uri(baseUrl.Trim());
             client.Timeout = TimeSpan.FromSeconds(180);
+            var token = Environment.GetEnvironmentVariable("CV_AI_SERVICE_TOKEN");
+            if (!string.IsNullOrWhiteSpace(token)) client.DefaultRequestHeaders.Add("X-Ai-Service-Token", token);
         });
         var provider = services.BuildServiceProvider();
         var factory = provider.GetRequiredService<System.Net.Http.IHttpClientFactory>();
         return new CvAiNormalizer(factory);
     }
+
+    private static ISummarizerService CreateSummarizerFromEnvironment()
+    {
+        var baseUrl = Environment.GetEnvironmentVariable("CV_AI_BASE_URL");
+        if (string.IsNullOrWhiteSpace(baseUrl))
+        {
+            var mock = new Mock<ISummarizerService>();
+            mock.Setup(x => x.ExtractTextAsync(It.IsAny<Stream>(), It.IsAny<string>(), It.IsAny<string?>(), It.IsAny<CancellationToken>()))
+                .Returns(async (Stream stream, string fileName, string? contentType, CancellationToken cancellationToken) =>
+                {
+                    using var reader = new StreamReader(stream);
+                    var text = await reader.ReadToEndAsync(cancellationToken);
+                    return new AiTextExtractionResult(text, false, contentType ?? "text/plain", null, text.Length, fileName);
+                });
+            mock.Setup(x => x.SummarizeSectionAsync(It.Is<string>(instruction => instruction.Contains("Extract this CV into structured JSON", StringComparison.Ordinal)), It.IsAny<string>(), 3200, 900)).ReturnsAsync(string.Empty);
+            mock.Setup(x => x.SummarizeSectionAsync(It.Is<string>(instruction => instruction.Contains("Reconstruct this CV text extracted from a PDF", StringComparison.Ordinal)), It.IsAny<string>(), 2800, 900)).ReturnsAsync((string _, string text, int _, int __) => text);
+            return mock.Object;
+        }
+
+        var services = new Microsoft.Extensions.DependencyInjection.ServiceCollection();
+        services.AddMemoryCache();
+        services.AddHttpClient("ai-service", client =>
+        {
+            client.BaseAddress = new Uri(baseUrl.Trim());
+            client.Timeout = TimeSpan.FromMinutes(3);
+            var token = Environment.GetEnvironmentVariable("CV_AI_SERVICE_TOKEN");
+            if (!string.IsNullOrWhiteSpace(token)) client.DefaultRequestHeaders.Add("X-Ai-Service-Token", token);
+        });
+        var provider = services.BuildServiceProvider();
+        return new SummarizerService(
+            provider.GetRequiredService<System.Net.Http.IHttpClientFactory>(),
+            provider.GetRequiredService<IMemoryCache>());
+    }
 }

JobTrackerApi.Tests/Fixtures/CvCorpus/synthetic-developer-cv.txt

@@ -0,0 +1,24 @@
+Alex Example
+Software Developer
+alex.example@example.test | +44 20 7946 0000 | London, United Kingdom
+https://example.test/alex
+
+Professional Summary
+Software developer building reliable web applications and APIs.
+
+Technical Skills
+C#, .NET, TypeScript, SQL, Docker
+
+Professional Experience
+Software Developer | Example Systems Ltd | London, United Kingdom
+January 2022 - Present
+- Built accessible web application features using .NET and TypeScript.
+- Improved automated test coverage for critical application workflows.
+
+Education
+BSc Computer Science | Example University | London, United Kingdom
+September 2018 - June 2021
+
+Languages
+English - Native
+Norwegian Bokmal - Intermediate