workspace_symbol_params.md

WorkspaceSymbolParams

Short introduction

This feature allows the LSP server to provide results for more flexible search conditions like:

• A Start_Word_Text works like a Full_Text but tested word should match patters from the first letter. This is the default choice.

• A Full_Text match searches the pattern exactly in the contents.

• A regexp parses the pattern as a regular expression.

• A fuzzy match will search for some contents that contains all the characters of the pattern, in the same order, but possibly with other characters in-between. The number of characters in-between is not limited, so this mode really only makes sense when matching short text (and not, for instance, in text editors).

• Approximate allows one or two errors to appear in the match (character insertion, deletion or substitution). This is mostly suitable when matching in long texts. The implementation of this algorithm is optimized so that characters are matched only once, but the total length of the pattern is limited to 64 characters. The exact number of errors depends on the length of the pattern: patterns of length <= 4 => no error allowed patterns of length <= 10 => one error allowed long patterns => up to two errors

Change description

We extend the WorkspaceSymbolParams by adding extra fields:

• case_sensitive to take letters' case into account
• whole_word to match the whole word instead of a part of it when True
• negate to invert matching
• kind is a search kind described above

export type Search_Kind = 'Full_Text' | 'Regexp' | 'Fuzzy' | 'Approximate' | 'Start_Word_Text';

export namespace Search_Kind {
	export const Full_Text = 1;
	export const Regexp = 2;
	export const Fuzzy = 3;
	export const Approximate = 4;
	export const Start_Word_Text = 5;
}

interface WorkspaceSymbolParams {
	query: string;
	case_sensitive?: boolean;
	whole_word?: boolean;
	negate?: boolean;
	kind?: Search_Kind;
}