Conversation
…search input This adds a new input-parser module that allows developers to parse incomplete/in-progress query input from search bars and extract structured information like: - Key and value being typed - Operator being used - Cursor position context - Position information for highlighting Main exports: - parseQueryInput: Parse input and get full context - extractKeyValue: Simple key/value extraction - getTermAtPosition: Get term at cursor position - isInputComplete: Check if input is complete This enables UX features like: - Key-value highlighting in search inputs - Autocomplete suggestions based on context - Real-time validation feedback
|
Cursor Agent can help with this pull request. Just |
|
You have run out of free Bugbot PR reviews for this billing cycle. This will reset on March 18. To receive reviews on all of your PRs, visit the Cursor dashboard to activate Pro and start your 14-day free trial. |
Contributor
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Adds comprehensive tests that document the known differences between: - QueryParser (Liqe-based, for execution) - parseQueryInput (regex-based, for UI) Key divergences documented: 1. Logical keywords as values: Main parser REJECTS status:AND/OR/NOT, input parser correctly handles them (workaround: quote the value) 2. Incomplete input: Main parser rejects, input parser returns partial 3. Quote handling: Main strips quotes, input preserves them 4. Type conversion: Main converts to proper types, input keeps strings 5. Complex syntax: Main fully handles ranges/arrays, input has limits This helps developers understand when to use each parser.
…ence
Adds a new function parseQueryTokens() that returns an interleaved
sequence of terms and operators, making it easier to:
- Render query tokens as UI chips/tags
- Build visual query builders
- Implement proper syntax highlighting
Example:
parseQueryTokens('status:done AND priority:high')
// Returns:
// [
// { type: 'term', key: 'status', value: 'done', ... },
// { type: 'operator', operator: 'AND', ... },
// { type: 'term', key: 'priority', value: 'high', ... }
// ]
New exports:
- parseQueryTokens: Parse input into token sequence
- IQueryTermToken: Term token type
- IQueryOperatorToken: Operator token type
- QueryToken: Union type for both
- IQueryTokenSequence: Result type with tokens + active token info
…ural analysis Adds parseWithContext() method to QueryParser that provides: - Success/failure indication (never throws) - Full AST on success, error info on failure - Token sequence (always available, even on failure) - Cursor-aware active token detection - Structural analysis: - Query depth (parenthesis nesting) - Clause count - Operator count - Balanced parentheses/quotes check - Referenced fields list - Complexity classification (simple/moderate/complex) - Completeness check This enables building rich search UIs with: - Real-time validation feedback - Query complexity indicators - Field extraction for autocomplete - Error context for user feedback Phase 1 of parseWithContext implementation. Phases 2-3 will add: schema validation, security pre-check, autocomplete hints, and error recovery suggestions.
…d input parser Adds 32 tests verifying that parseWithContext returns identical token information as the standalone input parser (parseQueryTokens/parseQueryInput): - Token count consistency across various inputs - Token content consistency (type, key, value, operator) - Position consistency (startPosition, endPosition, raw) - Cursor/active token consistency - Edge case handling (empty, incomplete, invalid input) - Structure field consistency This ensures the integration between the two parsing approaches is correct and developers can rely on consistent behavior regardless of which API they use.
…thContext (Phase 2) Extends parseWithContext with Phase 2 features: ## Schema Validation (when schema option provided) - Validates referenced fields against provided schema - Reports unknown fields with 'unknown_field' reason - Provides typo suggestions using: - Case-insensitive matching - Prefix matching - Levenshtein distance (for short field names) - Includes allowed values from schema for enum fields ## Security Pre-check (when securityOptions provided) - Checks denied fields - Checks allowed fields whitelist - Validates dot notation usage - Checks query depth limits (with warnings at 80%) - Checks clause count limits (with warnings at 80%) - Warns about complex queries New types: - IFieldSchema: Schema definition for fields - IFieldValidationResult: Field validation results - ISecurityCheckResult: Security check results - ISecurityViolation: Security violation details - ISecurityWarning: Security warnings 28 new tests for Phase 2 features.
This phase completes the parseWithContext enhancement with: Autocomplete Suggestions: - Context-aware suggestions based on cursor position - Field suggestions with fuzzy matching (prefix, contains, Levenshtein) - Value suggestions for fields with allowedValues in schema - Boolean value suggestions for boolean fields - Operator suggestions with applicability hints for field types - Logical operator suggestions (AND, OR, NOT) between terms - Replace text and range info for easy insertion Error Recovery: - Unclosed quote detection with auto-fix - Unclosed parenthesis detection with auto-fix - Trailing operator detection with auto-fix - Missing value detection - Generic syntax error fallback - Position tracking for errors New types: - IAutocompleteSuggestions - IFieldSuggestion - IValueSuggestion - IOperatorSuggestion - IErrorRecovery Includes 30 new tests for Phase 3 features.
README.md: - Add section on Input Parsing for Search UIs - Document parseQueryInput and parseQueryTokens for real-time parsing - Document parseWithContext with schema validation and autocomplete - Document error recovery with autofix suggestions - Add React search bar example with highlighting - Update roadmap to reflect completed features BUGBOT.md: - Add query parsing security guidelines - Document two-tier parser architecture - Add common issues for parser usage - Add testing guidelines for parser features Example app (qk-next): - Replace simple regex highlighter with parseQueryTokens-based implementation - Provides accurate tokenization for key:value highlighting - Supports comparison operators, logical operators, quoted values - Highlights keys (orange), operators (gray), values (blue), logical ops (purple)
- Update @gblikas/querykit dependency to use workspace:* for local linking - Add pnpm-workspace.yaml to set up workspace packages - Fix highlightQueryHtml to detect negation from key prefix instead of isNegated property - Remove deprecated swcMinify option from next.config.ts - Remove duplicate pnpm-lock.yaml in example folder The example app now builds successfully and uses the local querykit package.
- Change @gblikas/querykit dependency from workspace:* to file:../.. - Add vercel.json to configure proper build commands: - Install from root to get all workspace dependencies - Build root package first, then example app The workspace:* protocol doesn't work on Vercel as it expects npm packages. The file: protocol allows Vercel to resolve the local package properly.
- Move vercel.json from examples/qk-next to repository root - Add packageManager field to root package.json for Vercel detection - Add build:example script to build both packages in sequence - Update vercel.json to use pnpm run build:example This configuration requires Vercel project root directory to be set to '/' (the repository root) rather than the example folder.
Vercel runs commands from the detected project directory, not the repo root. Using pnpm -w flag ensures commands run from the workspace root.
Next.js needs to transpile the local querykit package to resolve exports properly, especially when using file: protocol for workspace linking.
Modern bundlers like Next.js require explicit exports declarations for proper module resolution, especially when using file: protocol. Added: - exports field with proper types/import/require/default conditions - module field pointing to dist/index.js - Subpath exports for parser, adapters, and translators
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
2 participants
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.
Add a new input parser module to enable real-time parsing of partial search queries for UI/UX features like highlighting and autocomplete.