Step 234: Add session pipeline tests

PROGRESS.md

@@ -753,3 +753,4 @@ vcpkg's imgui 1.91.9 removed the `sdl2-binding` feature (only `sdl3-binding` exi
 | 2026-02-10 | Codex | Step 231: Session anonymizer (paths, secrets, identifier anonymization, comment/string stripping, levels, deterministic seed). 1/1 tests pass. |
 | 2026-02-10 | Codex | Step 232: Session-to-trace converter (events->user msgs, RPC->tool calls/results, gap splitting, synthetic thinking). 1/1 tests pass. |
 | 2026-02-10 | Codex | Step 233: Training data pipeline CLI (load/anonymize/convert/filter/dedup/export + stats). 1/1 tests pass. |
+| 2026-02-10 | Codex | Step 234: Session pipeline tests (anonymizer round-trip, no PII, trace conversion, pipeline output count, quality filter, dedup). 1/1 tests pass. |

editor/CMakeLists.txt

@@ -1184,6 +1184,9 @@ target_link_libraries(step232_test PRIVATE nlohmann_json::nlohmann_json)
 add_executable(step233_test tests/step233_test.cpp)
 target_include_directories(step233_test PRIVATE src)
 target_link_libraries(step233_test PRIVATE nlohmann_json::nlohmann_json)
+add_executable(step234_test tests/step234_test.cpp)
+target_include_directories(step234_test PRIVATE src)
+target_link_libraries(step234_test PRIVATE nlohmann_json::nlohmann_json)
 add_executable(step213_test tests/step213_test.cpp)
 target_include_directories(step213_test PRIVATE src)
 target_link_libraries(step213_test PRIVATE nlohmann_json::nlohmann_json)

editor/src/SessionAnonymizer.h

@@ -233,6 +233,9 @@ private:
     }
 
     std::string anonymizeStringValue(const std::string& value, const std::string& key) {
+        if (key == "method" || key == "jsonrpc") {
+            return value;
+        }
         if (!key.empty() && keyLooksLikePath(key)) {
             return anonymizePath(value);
         }

editor/tests/step234_test.cpp

@@ -0,0 +1,81 @@
+// Step 234: Session pipeline tests.
+#include "SessionPipeline.h"
+#include <iostream>
+
+static void expect(bool cond, const std::string& name, int& passed, int& failed) {
+    if (cond) {
+        std::cout << "Test " << (passed + failed + 1) << " PASS: " << name << "\n";
+        ++passed;
+    } else {
+        std::cout << "Test " << (passed + failed + 1) << " FAIL: " << name << "\n";
+        ++failed;
+    }
+}
+
+static json sampleSession(const std::string& pathValue, const std::string& method) {
+    return {
+        {"sessionId", "s1"},
+        {"metadata", {{"language", "python"}}},
+        {"events", json::array({{
+            {"timestamp", "2026-02-10T10:00:00.000"},
+            {"type", "file_opened"},
+            {"payload", {{"path", pathValue}}}
+        }})},
+        {"calls", json::array({{
+            {"timestamp", "2026-02-10T10:00:01.000"},
+            {"method", method},
+            {"request", {{"params", json::object()}}},
+            {"response", {{"result", json::object()}}}
+        }})}
+    };
+}
+
+int main() {
+    int passed = 0;
+    int failed = 0;
+
+    try {
+        SessionAnonymizer anonymizer;
+        anonymizer.setLevel(SessionAnonymizer::Level::Medium);
+
+        json session = sampleSession("C:/Users/Bob/project/main.py", "getAST");
+        json anon = anonymizer.anonymize(session);
+
+        expect(session["calls"].size() == anon["calls"].size(), "structure preserved", passed, failed);
+        expect(anon.contains("events") && anon.contains("calls"), "events and calls preserved", passed, failed);
+        std::string anonPath = anon["events"][0]["payload"].value("path", "");
+        expect(anonPath.find("/project/src/") == 0, "paths anonymized", passed, failed);
+        expect(anon.dump().find("Bob") == std::string::npos, "identifiers removed", passed, failed);
+
+        SessionTraceConverter converter;
+        auto traces = converter.convert(anon);
+        expect(!traces.empty() && validateTrace(traces[0]), "trace conversion valid", passed, failed);
+
+        PipelineStats stats;
+        auto filtered = filterTraces(traces, stats);
+        std::string jsonl = exportTraces(filtered, "jsonl");
+        int lineCount = 0;
+        for (char c : jsonl) if (c == '\n') lineCount++;
+        expect(lineCount == (int)filtered.size(), "pipeline output count", passed, failed);
+
+        Trace degenerate;
+        degenerate.id = "bad";
+        degenerate.scenario = "session_replay";
+        degenerate.language = "python";
+        std::vector<Trace> withDegenerate = {traces[0], degenerate};
+        PipelineStats stats2;
+        auto filtered2 = filterTraces(withDegenerate, stats2);
+        expect(filtered2.size() == 1, "quality filter removes degenerate", passed, failed);
+
+        PipelineStats stats3;
+        auto deduped = dedupeTraces({traces[0], traces[0]}, stats3);
+        expect(deduped.size() == 1, "deduplication works", passed, failed);
+    } catch (const std::exception& ex) {
+        std::cerr << "Exception: " << ex.what() << "\n";
+        failed++;
+    }
+
+    std::cout << "\n=== Step 234 Results: " << passed << " passed, "
+              << failed << " failed ===\n";
+    return failed == 0 ? 0 : 1;
+}

sprint7_plan.md

@@ -440,7 +440,7 @@ Capture real editing sessions, anonymize them, and export as training data.
   - Validation: all output traces pass schema validation
   *New:* `editor/src/pipeline_main.cpp`
 
-- [ ] **Step 234: Session pipeline tests**
+- [x] **Step 234: Session pipeline tests**
   Tests for the recording and conversion pipeline:
   1. Recorded session round-trips through anonymizer preserving structure
   2. Anonymized sessions have no file paths or identifiable content