feat(autocomplete): sample multiple array elements for richer suggestions#147
Merged
bellicose100xp merged 12 commits intomainfrom Feb 20, 2026
Merged
feat(autocomplete): sample multiple array elements for richer suggestions#147bellicose100xp merged 12 commits intomainfrom
bellicose100xp merged 12 commits intomainfrom
Conversation
Update collect_fields_recursive() to iterate the first 10 array elements instead of only the first, unioning their keys into the all_field_names fallback cache. This ensures fields present only in non-first elements of heterogeneous arrays appear in suggestions. Add ARRAY_SAMPLE_SIZE constant (10) in json_navigator.rs as the single source of truth for the sample limit across the codebase.
… analyzer Add extract_union_fields_from_array() to sample first N array elements and deduplicate keys using a shared HashSet. Add analyze_multi_values() as the core implementation, and make analyze_value() delegate to it. Update extract_suggestions_for_type() for ArrayOfObjects to use the new union extraction instead of only examining the first element. Fields from non-first elements now appear in suggestions. Type detection uses first-occurrence-wins semantics for overlapping keys.
Add navigate_multi() that fans out at ArrayIterator segments, returning up to sample_size values per array level. Total results are capped at MAX_NAVIGATED_VALUES (100) to prevent combinatorial explosion at deeply nested levels. This is the foundation for Phase 4 where non-executing context (map, select, builders) will use navigate_multi to surface keys from multiple array elements.
Replace single-value navigate() with navigate_multi() in get_nested_field_suggestions(), enabling multi-element array traversal for non-executing contexts (map, select, array/object builders). Combined with analyze_multi_values(), this surfaces keys from multiple array elements when autocompleting inside map(.field), select(.field), and similar constructs.
… results When jq output is destructured objects (e.g., .services[] produces streaming objects), merge keys from the first N streamed objects into a synthetic combined object. This ensures autocomplete sees all fields across heterogeneous streamed results, not just the first object. Uses entry().or_insert() for first-occurrence-wins semantics, matching the type detection behavior in other phases.
…mpling Add issue #145 integration tests verifying: - .services[]. includes extra_service_key from second element - select(. with services also shows non-first-element keys - Nested array paths surface keys across elements - Invalid paths fall back cleanly Update README autocomplete limitation to document N=10 sampling behavior instead of first-element-only.
When jq streams multiple arrays (e.g., .services[].tasks produces
[{...}]\n[{...}]), merge elements from the first N arrays into a
single combined array. This ensures autocomplete sees fields from all
streamed arrays, not just the first.
Previously only destructured objects were merged (Phase 5). This
handles the parallel case where the streamed values are arrays of
objects rather than bare objects.
…ling Fill test gaps identified by coverage audit: navigate_multi: Field on non-object values, negative index underflow, ArrayIterator on non-arrays, intermediate empty results analyze_multi_values: empty slice, scalar-only values, multiple arrays with single .[] suggestion, suppress brackets, scalar-only arrays collect_fields_recursive: array exceeding ARRAY_SAMPLE_SIZE limit parse_and_detect_type: mixed-type streams for both destructured objects and arrays, sample limit enforcement for both merge paths
Expose the array sample size (previously hardcoded to 10) as a user-configurable option under [autocomplete] in config.toml. Also raise MAX_NAVIGATED_VALUES from 100 to 1000 for better coverage of heterogeneous arrays on modern hardware.
…calar When streaming output starts with null or a scalar but subsequent values are objects, reclassify to DestructuredObjects (or ArrayOfObjects) so autocomplete can extract field suggestions from the object values.
Verify that array_sample_size is correctly forwarded and respected at both the preprocessing layer (destructured stream merging) and the result analyzer layer (array field suggestion extraction). Also trim redundant phrasing from the autocomplete known limitation in README.
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
{...}\n{...}) or streamed arrays ([{...}]\n[{...}]), keys/elements from subsequent values are merged into the first for autocomplete.nullor a scalar but subsequent values are objects, the result type is reclassified so autocomplete can still extract field suggestions.array_sample_sizein the config file.Test plan
cargo test)cargo clippy --all-targets --all-features— zero warningscargo fmt --all --check— zero formatting issuescargo build --release— zero build warningstests/fixtures/mixed-type-array.json, type.services[].deployments[]?.tasks[]?.payload.→dataappears in autocomplete