Writing SQL and server-side JavaScript against a live dictionary now gets context-aware assistance in the editor. Completions are derived from the actual schema the statement references, not a generic keyword list, so the suggestions match the tables and columns in play.
Schema-aware SQL completion
- Statement-scoped columns. The completion provider parses the current statement's
FROM,JOIN,UPDATEandINSERT INTOclauses and offers the columns of every referenced table in theWHERE,ON,SELECT,GROUP BYandORDER BYpositions. Alias-qualified prefixes resolve to that table's columns. - Foreign-key JOIN conditions. A foreign-key endpoint exposes a table's imported and exported keys, so the
ONclause offers ready-made, alias-qualified join predicates instead of a bare column list. - Column metadata inline. Completions show a primary-key marker, a NOT NULL badge, and the column comment or default value in the documentation popup.
JavaScript that understands the dictionary
- Embedded query completion. Inside a database query call, the editor detects the SQL or XML query-string argument and delegates to the matching completion, with live table and column names.
- CRUD and function completion. The insert, update and delete calls complete table names and column keys; function calls and
require(...)complete dictionary function and module codes. - Go-to-definition. F12 or Ctrl+Click on a dictionary JavaScript function or module opens a peek widget showing its source, resolved through dedicated source endpoints.
The providers register once and defer to the editor's built-in TypeScript service for standard-library and local symbols, so dictionary-aware help is added without displacing the completions developers already rely on.