Benjamin_Boenisch/breakpilot-lehrer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix: Remove broken getKlausurApiUrl and clean up empty lines
Some checks failed
CI / go-lint (push) Has been skipped
Details
CI / python-lint (push) Has been skipped
Details
CI / nodejs-lint (push) Has been skipped
Details
CI / test-go-school (push) Successful in 42s
Details
CI / test-go-edu-search (push) Successful in 34s
Details
CI / test-python-klausur (push) Failing after 2m51s
Details
CI / test-python-agent-core (push) Successful in 21s
Details
CI / test-nodejs-website (push) Successful in 29s
Details

Browse Source 
sed replacement left orphaned hostname references in story page
and empty lines in getApiBase functions.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Benjamin Admin
2026-04-24 16:02:04 +02:00
parent b07f802c24
commit 9ba420fa91
150 changed files with 30231 additions and 32053 deletions
Download Patch File Download Diff File Expand all files Collapse all files

350
edu-search-service/internal/search/search_query.go Normal file
View File

@@ -0,0 +1,350 @@
package search
// buildQuery constructs the OpenSearch query
func (s *Service) buildQuery(req *SearchRequest) map[string]interface{} {
// Main query
must := []map[string]interface{}{}
filter := []map[string]interface{}{}
// Text search
if req.Query != "" {
must = append(must, map[string]interface{}{
"multi_match": map[string]interface{}{
"query": req.Query,
"fields": []string{"title^3", "content_text"},
"type": "best_fields",
},
})
}
// Filters
if len(req.Filters.Language) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"language": req.Filters.Language},
})
}
if len(req.Filters.CountryHint) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"country_hint": req.Filters.CountryHint},
})
}
if len(req.Filters.SourceCategory) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"source_category": req.Filters.SourceCategory},
})
}
if len(req.Filters.DocType) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"doc_type": req.Filters.DocType},
})
}
if len(req.Filters.SchoolLevel) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"school_level": req.Filters.SchoolLevel},
})
}
if len(req.Filters.Subjects) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"subjects": req.Filters.Subjects},
})
}
if len(req.Filters.State) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"state": req.Filters.State},
})
}
if req.Filters.MinTrustScore > 0 {
filter = append(filter, map[string]interface{}{
"range": map[string]interface{}{
"trust_score": map[string]interface{}{"gte": req.Filters.MinTrustScore},
},
})
}
if req.Filters.DateFrom != "" {
filter = append(filter, map[string]interface{}{
"range": map[string]interface{}{
"fetch_time": map[string]interface{}{"gte": req.Filters.DateFrom},
},
})
}
// Build bool query
boolQuery := map[string]interface{}{}
if len(must) > 0 {
boolQuery["must"] = must
}
if len(filter) > 0 {
boolQuery["filter"] = filter
}
// Construct full query
query := map[string]interface{}{
"query": map[string]interface{}{
"bool": boolQuery,
},
"from": req.Offset,
"size": req.Limit,
"_source": []string{
"doc_id", "title", "url", "domain", "language",
"doc_type", "school_level", "subjects",
"trust_score", "quality_score", "snippet_text",
},
}
// Add highlighting if requested
if req.Include.Highlights {
query["highlight"] = map[string]interface{}{
"fields": map[string]interface{}{
"title": map[string]interface{}{},
"content_text": map[string]interface{}{"fragment_size": 150, "number_of_fragments": 3},
},
}
}
// Add function score for trust/quality boosting
query["query"] = map[string]interface{}{
"function_score": map[string]interface{}{
"query": query["query"],
"functions": []map[string]interface{}{
{
"field_value_factor": map[string]interface{}{
"field": "trust_score",
"factor": 1.5,
"modifier": "sqrt",
"missing": 0.5,
},
},
{
"field_value_factor": map[string]interface{}{
"field": "quality_score",
"factor": 1.0,
"modifier": "sqrt",
"missing": 0.5,
},
},
},
"score_mode": "multiply",
"boost_mode": "multiply",
},
}
return query
}
// buildSemanticQuery constructs a pure vector search query using k-NN
func (s *Service) buildSemanticQuery(req *SearchRequest, embedding []float32) map[string]interface{} {
filter := s.buildFilters(req)
// k-NN query for semantic search
knnQuery := map[string]interface{}{
"content_embedding": map[string]interface{}{
"vector": embedding,
"k": req.Limit + req.Offset, // Get enough results for pagination
},
}
// Add filter if present
if len(filter) > 0 {
knnQuery["content_embedding"].(map[string]interface{})["filter"] = map[string]interface{}{
"bool": map[string]interface{}{
"filter": filter,
},
}
}
query := map[string]interface{}{
"knn": knnQuery,
"from": req.Offset,
"size": req.Limit,
"_source": []string{
"doc_id", "title", "url", "domain", "language",
"doc_type", "school_level", "subjects",
"trust_score", "quality_score", "snippet_text",
},
}
// Add highlighting if requested
if req.Include.Highlights {
query["highlight"] = map[string]interface{}{
"fields": map[string]interface{}{
"title": map[string]interface{}{},
"content_text": map[string]interface{}{"fragment_size": 150, "number_of_fragments": 3},
},
}
}
return query
}
// buildHybridQuery constructs a combined BM25 + vector search query
func (s *Service) buildHybridQuery(req *SearchRequest, embedding []float32) map[string]interface{} {
filter := s.buildFilters(req)
// Build the bool query for BM25
must := []map[string]interface{}{}
if req.Query != "" {
must = append(must, map[string]interface{}{
"multi_match": map[string]interface{}{
"query": req.Query,
"fields": []string{"title^3", "content_text"},
"type": "best_fields",
},
})
}
boolQuery := map[string]interface{}{}
if len(must) > 0 {
boolQuery["must"] = must
}
if len(filter) > 0 {
boolQuery["filter"] = filter
}
// Convert embedding to []interface{} for JSON
embeddingInterface := make([]interface{}, len(embedding))
for i, v := range embedding {
embeddingInterface[i] = v
}
// Hybrid query using script_score to combine BM25 and cosine similarity
// This is a simpler approach than OpenSearch's neural search plugin
query := map[string]interface{}{
"query": map[string]interface{}{
"script_score": map[string]interface{}{
"query": map[string]interface{}{
"bool": boolQuery,
},
"script": map[string]interface{}{
"source": "cosineSimilarity(params.query_vector, 'content_embedding') + 1.0 + _score * 0.5",
"params": map[string]interface{}{
"query_vector": embeddingInterface,
},
},
},
},
"from": req.Offset,
"size": req.Limit,
"_source": []string{
"doc_id", "title", "url", "domain", "language",
"doc_type", "school_level", "subjects",
"trust_score", "quality_score", "snippet_text",
},
}
// Add highlighting if requested
if req.Include.Highlights {
query["highlight"] = map[string]interface{}{
"fields": map[string]interface{}{
"title": map[string]interface{}{},
"content_text": map[string]interface{}{"fragment_size": 150, "number_of_fragments": 3},
},
}
}
return query
}
// buildFilters constructs the filter array for queries
func (s *Service) buildFilters(req *SearchRequest) []map[string]interface{} {
filter := []map[string]interface{}{}
if len(req.Filters.Language) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"language": req.Filters.Language},
})
}
if len(req.Filters.CountryHint) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"country_hint": req.Filters.CountryHint},
})
}
if len(req.Filters.SourceCategory) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"source_category": req.Filters.SourceCategory},
})
}
if len(req.Filters.DocType) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"doc_type": req.Filters.DocType},
})
}
if len(req.Filters.SchoolLevel) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"school_level": req.Filters.SchoolLevel},
})
}
if len(req.Filters.Subjects) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"subjects": req.Filters.Subjects},
})
}
if len(req.Filters.State) > 0 {
filter = append(filter, map[string]interface{}{
"terms": map[string]interface{}{"state": req.Filters.State},
})
}
if req.Filters.MinTrustScore > 0 {
filter = append(filter, map[string]interface{}{
"range": map[string]interface{}{
"trust_score": map[string]interface{}{"gte": req.Filters.MinTrustScore},
},
})
}
if req.Filters.DateFrom != "" {
filter = append(filter, map[string]interface{}{
"range": map[string]interface{}{
"fetch_time": map[string]interface{}{"gte": req.Filters.DateFrom},
},
})
}
return filter
}
// hitToResult converts an OpenSearch hit to SearchResult
func (s *Service) hitToResult(source map[string]interface{}, score float64, highlight map[string][]string, include SearchInclude) SearchResult {
result := SearchResult{
DocID: getString(source, "doc_id"),
Title: getString(source, "title"),
URL: getString(source, "url"),
Domain: getString(source, "domain"),
Language: getString(source, "language"),
DocType: getString(source, "doc_type"),
SchoolLevel: getString(source, "school_level"),
Subjects: getStringArray(source, "subjects"),
Scores: Scores{
BM25: score,
Trust: getFloat(source, "trust_score"),
Quality: getFloat(source, "quality_score"),
Final: score, // MVP: final = BM25 * trust * quality (via function_score)
},
}
if include.Snippets {
result.Snippet = getString(source, "snippet_text")
}
if include.Highlights && highlight != nil {
if h, ok := highlight["content_text"]; ok {
result.Highlights = h
}
}
return result
}