A zero-dependency, single-file HTML viewer for OPC UA Compliance Test Tool (UACTT) result files (*.results.xml / *.ctt_result_*.xml).
Open the .xml produced by a CTT run and get an instant, color-coded, collapsible view of every conformance group, unit, and test case — with a real pass/fail summary. No install, no server, no internet. Just double-click the HTML file.
Privacy: everything runs locally in your browser via the File API. Your result files never leave your machine and are never uploaded anywhere.
- 🗂️ Full result tree — every
ResultNoderendered as an indented, collapsible table (conformance groups → units → test cases → details). - 🎨 Color-coded by result — Error, Warning, Recommendation, Not Implemented, Skipped, Not Supported, OK, Backtrace — each with its own badge and a left-accent bar.
- 🔢 Honest summary counts — the pill totals count real test cases only, not the Warning/Error/Backtrace detail sub-nodes that would otherwise inflate the numbers.
- 🖱️ Click-to-filter — click any summary pill (or a test case's result badge) to show only test cases with that result. Click again to clear.
- 🔎 Live search by node name, plus quick toggles for hide Backtrace and only errors/warnings.
- 🧭 Global result banner — the overall run result and the detected file format, at a glance.
- 🔀 Automatic format detection — transparently handles result files from different CTT tool versions that use different internal result codes (see below).
- 📦 Single file, zero dependencies — pure HTML/CSS/vanilla JS. Works offline in any modern browser.
- 🖐️ Drag & drop — drop an
.xmlanywhere on the window to load it.
- Open https://cacamille3.github.io/OPCUA-CTT-Result-Viewer/ or local copy of index.html in your web browser
- Double-click it (opens in Chrome, Edge, or Firefox), or serve it however you like.
- Click Choose File and pick your CTT result
.xml— or drag the file onto the window.
That's it. Nothing is installed and nothing is transmitted.
UACTT result files from different tool versions use different testresult numeric codes for the same meaning. The viewer auto-detects the format and maps codes to a shared set of categories, so the labels, colors, and filters are always correct.
| Category | "new" format (UACTT 2.x, has groupkey) |
"legacy" format (older tool, no version attrs) |
|---|---|---|
| Error | 0 | 0 |
| Warning | 1 | 1 |
| Recommendation | 2 | — |
| Not Implemented | 3 | 2 |
| Skipped | 4 | 3 |
| Not Supported | 5 | 4 |
| OK | 6 | 5 |
| Backtrace | 7 | 6 |
Detection order: the root node's binaryversion / scriptversion / specificationversion attributes → presence of groupkey → otherwise the numeric code on a Backtrace node. The detected format is shown in the banner (… · format: new | legacy).
The summary deliberately counts only executed test cases, not container or detail nodes:
- new format: nodes carrying both
groupkeyandunitkey. - legacy format: nodes whose name ends in
.js. - In both, the per-unit setup/teardown scripts (
initialize.js,cleanup.js,beforeTest.js,afterTest.js) are excluded.
This is why the pill totals (e.g. 302 test cases) are much smaller than the raw node count — the Backtraces and Warning/Error detail children are shown in the tree but not counted as test cases.
| Badge | Meaning |
|---|---|
| 🟥 Error | Test case failed |
| 🟧 Warning | Passed with a warning |
| 🟨 Recommendation | Spec recommendation not followed |
| 🟪 Not Implemented | Test not implemented in the tool |
| ⬜ Skipped | Test skipped |
| 🟦 Not Supported | Feature not supported by the server |
| 🟩 OK | Passed |
| ▪️ Backtrace | Script location detail (hidden by default) |
Pure web platform — no frameworks, no build step, no dependencies:
- HTML5 + CSS3 (custom properties, flexbox, sticky headers)
- Vanilla JavaScript (ES6+)
DOMParserfor XML,FileReader+ File API for local file loading, Drag-and-Drop API
Everything lives in one .html file, so it's trivially portable — copy it anywhere and open it.
- The legacy-format code map for Error and Recommendation is inferred from the shift pattern of the observed codes; result files that actually contain those rows are the only way to fully confirm them. Everything observed in real files is correct. If a legacy file ever mislabels an Error row, it's a one-line change in the
SCHEMES.legacymap near the top of the<script>. - Result-code meanings and colors are all defined in two small objects (
CATandSCHEMES) at the top of the script — easy to tweak or extend for other CTT variants. - Tested with UACTT 1.04 result files (both the 2.x tool output and older
ctt_result_*.xmloutput).
Because the summary counts real test cases per group, it's handy for spotting coverage gaps between runs — e.g. a run missing an entire service set (Subscription, Monitored Item, Method) will show far fewer test cases even though the shared groups match. Load each file and compare the pill totals.
Issues and PRs welcome. Since it's a single file with no build, contributing is as simple as editing index.html and opening it in a browser.