refactored entire code base to be in the package core so i can later implement interfaces to interact with the engine

1 files changed, 68 insertions, 0 deletions

diff --git a/core/index/index.go b/core/index/index.go
new file mode 100644
index 0000000..59b315c
--- /dev/null
+++ b/core/index/index.go
@@ -0,0 +1,68 @@
+package index
+
+import (
+	"searchEngine/core/analyzer"
+	"sync"
+)
+
+type Posting struct {
+	DocId          int
+	BodyCount      int
+	TitleCount     int
+	BodyPositions  []int
+	TitlePositions []int
+}
+
+type InvertedIndex struct {
+	Mu       sync.RWMutex
+	Data     map[string][]Posting
+	DocNames map[int]string
+}
+
+func New() *InvertedIndex {
+	return &InvertedIndex{
+		Data:     make(map[string][]Posting),
+		DocNames: make(map[int]string),
+	}
+}
+
+func (idx *InvertedIndex) Add(docID int, title string, bodyTokens []string) {
+	titleTokens := analyzer.ProcessText(title)
+
+	type fieldStats struct {
+		body           int
+		bodyPositions  []int
+		title          int
+		titlePositions []int
+	}
+	docStats := make(map[string]fieldStats)
+
+	for position, token := range titleTokens {
+		stats := docStats[token]
+		stats.title++
+		stats.titlePositions = append(stats.titlePositions, position)
+		docStats[token] = stats
+	}
+
+	for position, token := range bodyTokens {
+		stats := docStats[token]
+		stats.body++
+		stats.bodyPositions = append(stats.bodyPositions, position)
+		docStats[token] = stats
+	}
+
+	idx.Mu.Lock()
+	defer idx.Mu.Unlock()
+
+	idx.DocNames[docID] = title
+
+	for token, stats := range docStats {
+		idx.Data[token] = append(idx.Data[token], Posting{
+			DocId:          docID,
+			BodyCount:      stats.body,
+			TitleCount:     stats.title,
+			BodyPositions:  stats.bodyPositions,
+			TitlePositions: stats.titlePositions,
+		})
+	}
+}