get diary info

src/diary_cover_similarity.py

@@ -7,6 +7,42 @@ import faiss
 import numpy as np
 
 from utils.images import face_to_vec, url_to_ndarray
+from utils.es import es_query
+
+
+def save_diary_image_info(file):
+    q = {
+        "query": {
+            "bool": {
+                "filter": [{
+                    "term": {
+                        "is_online": True
+                    }
+                }, {
+                    "term": {
+                        "has_before_cover": True
+                    }
+                }, {
+                    "term": {
+                        "has_after_cover": True
+                    }
+                }, {
+                    "exists": {
+                        "field": "before_cover_url"
+                    }
+                }]
+            }
+        },
+        "_source": {
+            "include": ["id", "before_cover_url", "after_cover_url"]
+        }
+    }
+
+    with open(file, "w"):
+        step = 100
+        for i in range(0, 500000, step):
+            res_dict = es_query("diary", q, 0, step)
+            print(res_dict)
 
 
 def main():
@@ -20,37 +56,41 @@ def main():
     shape_model_path = os.path.join(model_diry, "shape_predictor_68_face_landmarks.dat")
     faiss_index_path = os.path.join(base_dir, "_index", "current.index")
 
-    face_rec = dlib.face_recognition_model_v1(facerec_model_path)
-    face_detector = dlib.get_frontal_face_detector()
-    shape_predictor = dlib.shape_predictor(shape_model_path)
-
-    if img.any():
-        faces = face_to_vec(img, face_rec, face_detector, shape_predictor)
-
-        if len(faces) > 0:
-            for face in faces:
-                print(face)
-                print("\n")
-                print("-------------")
-                face_feature = np.array(json.loads(face["feature"]))
-
-            ids = [17418645]
-            ids_np = np.array(ids).astype("int")
-            faces_np = np.array([face_feature]).astype("float32")
-            print(ids_np)
-            print(faces_np)
-
-            index = faiss.IndexHNSWFlat(128, 32)
-            print("index: ")
-            print(index)
-            index2 = faiss.IndexIDMap(index)
-            print("index2: ")
-            print(index2)
-            res = index2.add_with_ids(faces_np, ids_np)
-            print("res: ")
-            print(res)
-        else:
-            print("no faces")
+    diary_after_cover_vec_file = "./diary_after_cover_vec.txt"
+
+    save_diary_image_info(diary_after_cover_vec_file)
+
+    # face_rec = dlib.face_recognition_model_v1(facerec_model_path)
+    # face_detector = dlib.get_frontal_face_detector()
+    # shape_predictor = dlib.shape_predictor(shape_model_path)
+
+    # if img.any():
+    #     faces = face_to_vec(img, face_rec, face_detector, shape_predictor)
+
+    #     if len(faces) > 0:
+    #         for face in faces:
+    #             print(face)
+    #             print("\n")
+    #             print("-------------")
+    #             face_feature = np.array(json.loads(face["feature"]))
+
+    #         ids = [17418645]
+    #         ids_np = np.array(ids).astype("int")
+    #         faces_np = np.array([face_feature]).astype("float32")
+    #         print(ids_np)
+    #         print(faces_np)
+
+    #         index = faiss.IndexHNSWFlat(128, 32)
+    #         print("index: ")
+    #         print(index)
+    #         index2 = faiss.IndexIDMap(index)
+    #         print("index2: ")
+    #         print(index2)
+    #         res = index2.add_with_ids(faces_np, ids_np)
+    #         print("res: ")
+    #         print(res)
+    #     else:
+    #         print("no faces")
 
 
 if __name__ == "__main__":