fdnext

fdnext returns results for different memory types through a shared structure that preserves device identity, fields, units, candidates, relations, and warnings. Applications can use that structure to build detail views or records without trying to recover the meaning of each parameter from a block of descriptive text.

Keep presentation separate from data semantics: show translated labels in the interface, but use field keys and raw values in application logic. Brief summaries help with browsing, while full field groups support closer inspection. Capacity, component counts, and power-supply information remain absent when the result does not provide them.

Install the core package in a Node.js or browser application and use one engine instance for decoding, search, and capability queries. Default resources are embedded in the package, so they do not require separate data downloads. Create and reuse the engine at startup rather than initializing it for every request or component render.

Use the Node.js HTTP server or Cloudflare Workers adapter when several clients should share a lookup service. Both use common routing and result contracts, allowing rules and data to be maintained centrally. To replace the embedded resources, the application must supply its own data and handle loading, updates, and validation.

Start with representative samples using the command-line tool included in the core package. Decode a complete part number directly, or search a partial marking and inspect the candidates in detail. Text output suits manual review. For scripts, select JSON and retain both the input and result status so queries can be revisited and exceptional outcomes handled explicitly.

A completed command does not necessarily mean a chip was identified. Outcomes such as not found, ambiguous, or unsupported can still accompany a zero exit code, so automation must inspect the result itself. HTTP clients should likewise inspect the query status in the returned data and distinguish search summaries from the more complete information returned by decoding.

Decoding is limited by the rules and data included; it cannot promise recognition of every part number or replace physical measurements. Preserve missing information, ambiguity, and warnings in your integration, and check results against samples relevant to your application. Server searches have a result cap, so their behavior is not identical to calling the core library directly.

The versions link leads to GitHub release records. Follow the developer documentation to install the packages for the engine and CLI. When upgrading the engine or result structure, check the applications that use it. Automation should read the versioned JSON format, as command-line text layout may change between versions.