Tune editor

folkugat_web/services/temes/search.py

@@ -1,4 +1,7 @@
+import functools
 import time
+import typing
+from collections.abc import Callable, Iterable, Iterator
 
 import Levenshtein
 from folkugat_web.config import search as config
@@ -18,9 +21,10 @@ def get_query_word_similarity(query_word: str, text_ngrams: model.NGrams) -> sea
     candidate_ngrams = ((m, ngram)
                         for m, ngrams in map(lambda i: (i, text_ngrams.get(i, [])), ns)
                         for ngram in ngrams)
-    return min(search_model.SearchMatch(distance=Levenshtein.distance(query_word, ngram)/m,
-                                        ngram=ngram)
-               for m, ngram in candidate_ngrams)
+    return min((search_model.SearchMatch(distance=Levenshtein.distance(query_word, ngram)/m,
+                                         ngram=ngram)
+                for m, ngram in candidate_ngrams),
+               default=search_model.SearchMatch(distance=float("inf"), ngram=""))
 
 
 def get_query_similarity(query: str, ngrams: model.NGrams) -> search_model.SearchMatch:
@@ -44,14 +48,24 @@ def build_result(query: str, entry: tuple[int, model.NGrams]) -> search_model.Qu
     )
 
 
-def busca_temes(query: str) -> list[model.Tema]:
+T = typing.TypeVar("T")
+
+
+def _thread(it: Iterable[T], *funcs: Callable[[Iterable], Iterable]) -> Iterable:
+    return functools.reduce(lambda i, fn: fn(i), funcs, it)
+
+
+def busca_temes(query: str, hidden: bool = False, limit: int = 20, offset: int = 0) -> list[model.Tema]:
     t0 = time.time()
     with get_connection() as con:
-        tema_id_to_ngrams = temes_q.get_tema_id_to_ngrams(con)
-        search_results = (build_result(query, entry) for entry in tema_id_to_ngrams.items())
-        filtered_results = filter(lambda qr: qr.distance <= config.SEARCH_DISTANCE_THRESHOLD, search_results)
-        # filtered_results = filter(lambda qr: True, search_results)
-        sorted_results = sorted(filtered_results, key=lambda qr: qr.distance)
-        sorted_temes = list(filter(None, map(lambda qr: temes_q.get_tema_by_id(qr.id, con), sorted_results)))
+        result = _thread(
+            temes_q.get_tema_id_to_ngrams(con).items(),
+            lambda tema_id_to_ngrams: (build_result(query, entry) for entry in tema_id_to_ngrams),
+            lambda results: filter(lambda qr: qr.distance <= config.SEARCH_DISTANCE_THRESHOLD, results),
+            lambda results: sorted(results, key=lambda qr: qr.distance),
+            lambda results: filter(None, map(lambda qr: temes_q.get_tema_by_id(qr.id, con), results)),
+            lambda results: filter(lambda t: hidden or not t.hidden, results),
+        )
+        result = list(result)[offset:offset + limit]
     logger.info(f"Search time: { int((time.time() - t0) * 1000) } ms")
-    return sorted_temes
+    return result