Skip to content
EDT-Bridge
English
Esc
↑↓navigate↵open⌘Jpreview
On this page

Read: diagnostics and reference

Read tools for project state, validation problems, check descriptions, tool and property reference and the platform Syntax Helper.

These tools report the state of the workspace projects and their validation problems, and return reference material: check descriptions, tool schemas, the properties a write tool accepts and the platform Syntax Helper.

Tool What it returns
edt_projects · edt_project_errors Workspace projects and validation problems from Eclipse and EDT checks. modelReady reports whether the BM model and derived data are ready; complete=false marks a partial snapshot and must not be treated as a clean result. Filter by fqn, modulePath or severity; countOnly returns counts, with the readiness diagnostics only for a partial result or with details=true; brief returns one line per problem. modulePath includes EDT markers traced to source files; unresolved markers are counted and reported with a hint. Form filters exclude sibling forms. stale identifies markers older than their source and unsynchronized identifies disk changes not yet read by the workspace. leftover marks a marker older than its file that EDT did not renew when it checked the file again: EDT keeps the marker of a query check after a data composition schema no longer has the problem. A leftover does not make the snapshot partial and needs no fix; leftoverCount counts them. refresh=true rereads the scope and builds the project; waitSeconds sets the validation wait (1-600, default 120). A timeout returns current markers with refreshed.settled=false. Include WARNING when checking generated code: EDT reports some undefined methods as warnings.
edt_support_status EDT direct-edit preflight for an existing object (projectName, objectFqn). Returns EDITABLE, READ_ONLY or UNKNOWN; unavailable policy never grants edit permission. This check does not lock the object.
edt_compare_configurations · edt_comparison_details Compare open configuration projects or their Git revisions using EDT. Includes an optional common ancestor, counts, changed objects and details down to BSL methods.
edt_check_info Returns a validation check’s description, non-compliant and compliant examples, and links to 1C development standards in Russian or English. It complements edt_project_errors, which identifies the triggered check. Search by check ID or by title. A short code such as SU68 is resolved to its check when projectName names the project that reported the problem.
edt_tool_help Lists active tools with short descriptions when tool is omitted. With tool set, reports its description, schema and availability. brief=true returns a signature with names and types. An asterisk marks a required argument.
edt_property_reference Lists editable properties of a common module, external object, extension, field, managed form, form item or form command, with each property’s type, accepted values and current value, and names the tool that takes them. A form FQN without item or command returns the form’s own properties plus its items and commands. Pass command to inspect a form command. The list comes from the validation code of the write, so a property the object lacks is not listed.
edt_platform_help Search the 1C:Enterprise platform Syntax Helper supplied with EDT by name or read a page as text. The Russian and English API reference covers objects, methods, properties, and events.

Project and runtime diagnostics

edt_projects reports model state in each available project diagnostics entry. pendingReasons lists conditions such as a model lock or pending derived data. derivedData reports the pipeline stage and the work EDT exposes for active segments. backgroundJobs is scoped to the workspace.

edt_project_errors reports diagnostics.pendingReasons, workspace jobs, validation completion and the observed problem count. An object check can finish while workspace jobs are still active. Read again with refresh and baseline omitted to sample the current state. Use complete to assess the result. For a refresh, also check refreshed.settled.

Observations use epoch milliseconds. firstObservedAt marks the first sample in the bridge’s retained history. unchangedSince marks the first sample of the current value. The first sample has lastChangeKnown=false. After a different value is sampled, lastChangeObservedAt records when the bridge saw the difference. History is limited to the current process. Metrics that EDT cannot provide include available=false and unavailableReason.

GET /status reports resources for the running bridge. The jvm section includes heap and non-heap usage, committed memory, the maximum where defined, and the JVM’s collection count and time for each garbage collector. The container section reports the current process’s cgroup v2 memory usage, limits, events, pressure, CPU quota and CPU counters on Linux when readable. The snapshot uses read-only JVM and cgroup queries. Unavailable measurements include a reason. processStartedAt records when the current process started. previousExitCause is unknown. Under heavy calls /status waits no longer than 1.5 s for the model state, the workspace jobs, resources and the tool count. If they do not come in time, the answer leaves them out, sets busy=true and lists the open projects with modelReady=null.

Validation of one object

With exact fqn or modulePath, edt_project_errors and refresh=true explicitly schedule applicable checks for that BM object and its BSL modules, including the module of a managed form. Unresolved or conflicting addresses never expand this request to the whole project. Form-model edits through the bridge also schedule module revalidation and are rechecked on the next unnarrowed refresh=true, so BSL checks such as SU243 reflect the updated Form.form. validationRequested and validationCompleted report scheduling and completion. Uninspected count-only or truncated snapshots cannot establish marker freshness.

Validation snapshots

To compare validation results, call edt_project_errors with baseline="save", then pass the returned baselineId with baseline="compare" after the change. Both operations refresh the scope and wait for validation to settle even when refresh is omitted. Use the same project and scope filters in both calls. Comparison returns only new problems plus resolved-problem counts and resolvedProblems; limit, brief and countOnly control the returned presentation. Line shifts do not turn an unchanged problem into a new one, and duplicate problems remain separate occurrences. Incomplete or truncated input is refused. Snapshots live only in this bridge session; restarting EDT invalidates them. The cache holds at most 16 snapshots and 200000 markers in total, with 100000 markers per snapshot. Saving returns the IDs it evicted.

Compare configurations

Call edt_compare_configurations with main and other. Each side has a projectName for an open EDT configuration project and an optional revision for a local Git ref of that project. Omit revision to compare the live model. For a three-way comparison, pass ancestor with the same fields. To track a supplier update, pass the adapted project as main, the new supplier revision as other, and the previous supplier revision as ancestor.

EDT runs the comparison in read-only mode. For large projects, pass async=true and poll edt_job_status. The result includes counts for the whole comparison and a page of objects. Follow nextOffset with a new call that supplies comparisonId and offset. Use edt_comparison_details with the object’s index as objectIndex. To open a child with hasDetails=true, append its index to nodeIndices and request its child page. Module details include changed BSL methods when EDT has parsed the module structure.

The summary counts metadata objects with differences in EDT. It includes order changes marked by EDT and changes to distribution and support options. Each object has one row even when several properties, modules or methods differ. Designer /CompareCfg reports differences between platform configurations and can yield different counts.

The server keeps up to four comparison results for 30 minutes. Run the comparison again if its ID expires.

Last updated on November 22, 2026

Was this page helpful?