If you were handed a folder of .sor files and told to extract the span length, total loss, and the event table into your own system, you already know the problem. There's no clean, open spec that gets you all the way there. Bellcore SR-4731 covers the base structure. The real files in the real world add vendor extensions, proprietary blocks, bit-packing quirks, and the occasional format that isn't Bellcore at all. EXFO uses .trc. Multi-wavelength files add their own wrapper as .msor. Every brand of OTDR has done something a little different.
The result is predictable. The first week looks tractable. The second week turns into reverse-engineering someone else's hex dumps. By the third week you understand why the vendors sell their own report software.
There's an easier path.
The API
The parsing engine behind the free Fiber Analyzer and Fiber Analysis tools is available as a REST API at johnstonetechs.com/developers. The same code that powers the browser tools runs behind the endpoint. If you've ever opened a trace in the Fiber Analyzer and seen the length, loss, and event table come back correctly, that's the output you get through the API, as JSON.
Two endpoints are live:
POST /api/v1/parse takes one trace file and returns one structured result. Supported formats are .sor and .msor (Bellcore SR-4731) and EXFO .trc. The response includes span length, total loss, worst reflectance, the full event table (every connector, splice, and reflection with distance, loss, and reflectance), and instrument metadata (make, model, wavelength, pulse width, averaging time).
POST /api/v1/analyze takes a whole job (many files, or a zip) and isn't limited to OTDR traces. A real job is a mix, and this endpoint reads that mix: OTDR traces, OLTS and iOLM link-loss results, endface inspection files, and optical power-meter files. It groups the files into fibers, runs the same pass/fail logic as the Fiber Analysis tool, and returns a per-fiber verdict: passing, fail, no-continuity, or not-tested. Each fiber comes back with a plain-language summary and the measured numbers. This is the endpoint that replaces a human opening hundreds of files by hand.
Full reference docs, request and response shapes, and request-a-key instructions are at /developers. What follows is enough to see what it feels like to call.
What you get back
A parse response gives you the measured results and the instrument context together, in one JSON payload. The structure matches what you'd see in the browser Fiber Analyzer: a fiber-level summary (length, total loss, worst reflectance, overall pass/fail against standard thresholds), the event list, and the metadata block that identifies the test equipment and settings. See the reference for the full schema.
The analyze response adds a layer on top: per-fiber verdicts, reasoning across the whole job (resolving each file's identity, flagging conflicts where the filename and the file's internal ID disagree, handling bidirectional shots, and listing any files it couldn't place), and the actionable summary that the Fiber Analysis tool produces for its users.
One minimal example
Base64-encode a file and POST it:
curl -X POST https://johnstonetechs.com/api/v1/parse \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d "{\"filename\":\"span12.sor\",\"file_base64\":\"$(base64 -w0 span12.sor)\"}"That's the shape of the request. One file in, structured JSON out. See /developers for the full request schema, the response reference, and the error codes.
Want a key? The API is in early access while we bring on the first teams. The docs and the request form are on the developers page.
Read the API docs →Beyond one file
For bulk work, /api/v1/analyze accepts a whole job and returns a verdict per fiber. This is the endpoint that fits a reporting pipeline or an automated acceptance workflow. Instead of parsing files one at a time and writing your own pass/fail logic, you post the job and get back a result set your dashboard or QA system can consume directly. For a concrete picture of that integration, see how a contractor automated fiber acceptance instead of hiring a review team.
How to get access
The API is in early access. Request a key at johnstonetechs.com/developers. The browser-based Fiber Analyzer and Fiber Analysis tools stay free for interactive use. The API is the supported path for programmatic and bulk work.
Nothing is stored. Files are parsed in memory for the request and discarded.
Related
- Fiber Trace API docs, the full reference and request-a-key form
- Automating fiber acceptance testing, a worked integration
- Fiber Analyzer, the free interactive tool that runs the same engine
- Fiber Analysis, whole-job pass/fail in the browser
- How to open a .sor file
- All field guides