Skip to content
xc

SearchIndex

SearchIndex is a small full-text index: documents go in under a String id, and searches for words come back ranked best first. It is what a find-as-you-type box or a local search needs. From 0.72.

#import "SearchIndex.xc"   // not in the Foundation umbrella: import it by name

Overview

Section titled “Overview”
SearchIndex* idx = new SearchIndex();
idx.addDocument(String.withCString("a"), String.withCString("The quick brown fox"));
idx.addDocument(String.withCString("b"), String.withCString("A quick quick dog"));
Array* hits = idx.search(String.withCString("quick fox"));
// a (both words), then b (one word, twice)
SearchResult* best = (SearchResult*)hits.get(0);   // best.id is "a"

Words. Text is split into runs of letters and digits, with ASCII letters lowered, so Fox, and fox are one word. The bytes of UTF-8 count as letters, so other scripts are indexed too, without case folding. An inverted index maps each word to the documents that hold it and how often.

Queries are split the same way. search finds documents holding any of the query’s words, searchAll those holding every one. A query word ending in * matches every word it begins (qui* finds quick and quietly).

Ranking: more of the query’s words matched first, then more occurrences of them, then the order the documents were first added.

Topics

Section titled “Topics”

Documents · addDocument · removeDocument · containsDocument · documentCount · termCount

Searching · search · searchAll · SearchResult · words

Documents

Section titled “Documents”

addDocument

Section titled “addDocument”
void addDocument(String* id, String* text)

Indexes text as the document id, replacing any document of that id (which keeps its place in the tie-breaking order).

removeDocument

Section titled “removeDocument”
void removeDocument(String* id)

containsDocument

Section titled “containsDocument”
bool containsDocument(String* id)

documentCount

Section titled “documentCount”
u32 documentCount(void)

termCount

Section titled “termCount”
u32 termCount(void)

The number of distinct words indexed.

↑ Topics

Searching

Section titled “Searching”
Section titled “search”
Array* search(String* query)

The documents holding any of the query’s words, as SearchResults, best first. An empty query finds nothing.

searchAll

Section titled “searchAll”
Array* searchAll(String* query)

The documents holding every one of the query’s words, best first.

SearchResult

Section titled “SearchResult”
class SearchResult : Object
    {
    String* id;
    u32 matchedTerms;   // how many of the query's words it holds
    u32 score;          // how many times it holds them, in all
    }

words

Section titled “words”
static Array* words(String* text)

The words text splits into, in order, repeats included: what the index stores.

↑ Topics