added migrate

application/db.py

@@ -7,7 +7,7 @@ import enum
 from flask_sqlalchemy import SQLAlchemy
 from flask import Flask
 
-engine = db.create_engine('sqlite:///./test.sqlite', echo=False)
+engine = db.create_engine('mysql+mysqldb://root@server/fs?charset=utf8mb4', echo=False, encoding="utf8")
 connection = engine.connect()
 Base = declarative_base()
 Session = sessionmaker(bind=engine)

migrate.py

@@ -0,0 +1,26 @@
+import json
+import cv2
+import base64
+from application.db import Session, Recipe, Ingredient, Link, Trunk
+
+def migrate(path):
+    recs = ""
+    with open(path, encoding="utf-8") as file:
+        recs = json.load(file)
+    
+    dbSession = Session()
+    for key, value in recs.items():
+        name=key
+        resString=value[0]
+        link=value[2]
+        img=value[3].encode()
+
+        r = Recipe(name=name, instructions=resString, url=link, img=img)
+        for x, y in value[1].items():
+            a = Link(ingredient_amount=y)
+            a.ingredient = Ingredient(name=x)
+            r.ingredient.append(a)
+        dbSession.add(r)
+        dbSession.commit()
+
+migrate('./data/recs.json')

mine.py

@@ -9,6 +9,9 @@ import traceback
 import cv2
 import base64
 from application.db import Session, Recipe, Ingredient, Link, Trunk
+import nltk as nltk
+from nltk.corpus import stopwords
+
 
 header_values = {
     'name': 'Michael Foord',
@@ -123,6 +126,23 @@ def getRecipe(links):
             sleep(random.randint(0, 5))
     return recs
 
+def stemIngred():
+    dbSession = Session()
+    stopset = set(stopwords.words('german'))
+    stopset |= set("(),")
+
+    count = dbSession.query(Ingredient).count()
+    for x in dbSession.query(Ingredient).all():
+        snowball = nltk.SnowballStemmer(language='german')
+        for token in nltk.word_tokenize(x.name): 
+            if token in stopset or len(token) < 3:
+                continue
+            stemmed = snowball.stem(token)
+
+            x.trunks.append(Trunk(name=stemmed))
+            dbSession.commit()
+        print(x.ingredient_id/count)
+
 #links = getLinks()
 #with open('./data/links.json', 'w') as file:
 #    jsonString = json.dumps(links)
@@ -132,8 +152,9 @@ with open('./data/links.json') as file:
     links = json.load(file)
     
 
-recs = getRecipe(links)
+#recs = getRecipe(links)
+stemIngred()
 
-with open('./data/recs.json', 'w', encoding="utf-8") as file:
+#with open('./data/recs.json', 'w', encoding="utf-8") as file:
-    json.dump(recs, file, ensure_ascii=False)
+#    json.dump(recs, file, ensure_ascii=False)
     

test.py

@@ -2,9 +2,11 @@
 from application.db import Session, Recipe, Ingredient, Link, Trunk
 import nltk as nltk
 from nltk.corpus import stopwords
+import time
 
 dbSession = Session()
-inputArr = ["kartoffeln", "zwiebel", "steak", "würfel"] 
+inputArr = ["butter", "milch", "eier", "käse"] 
+maxMissing = 4
 
 def slow():
     recipes = dbSession.query(Recipe).all()
@@ -14,7 +16,7 @@ def slow():
     for recipe in recipes:
         rec = recipe
         recipe = recipe.ingredients()
-        if len(recipe) > len(inputArr) + 2:
+        if len(recipe) > len(inputArr) + maxMissing:
             continue
         counter = 0
         for i in inputArr:
@@ -27,11 +29,12 @@ def slow():
         if counter not in arr:
             arr[counter] = []
             
-            arr[counter].append(rec.ingredients())
+        arr[counter].append(rec.ingredients())
         #print(rec.name)
         
-
+#    for y, x in arr.items():
-    print(arr)
+#        for xx in x:
+#            print(xx)
 
 def faster():
     indx = {}
@@ -41,7 +44,7 @@ def faster():
 
             for y in x.recipe:
                 
-                if dbSession.query(Link).filter(Link.recipe_id==y.recipe_id).count() > len(inputArr) + 5:
+                if dbSession.query(Link).filter(Link.recipe_id==y.recipe_id).count() > len(inputArr) + maxMissing:
                     continue   
                 if str(y.recipe_id) not in indx:
                     indx[str(y.recipe_id)] = 0
@@ -49,28 +52,81 @@ def faster():
                 indx[str(y.recipe_id)] += 1
         
 
+    outDict = {}
     for key, value in indx.items():
-        if value >= len(inputArr):
+        ingred = dbSession.query(Recipe).filter(Recipe.recipe_id==key).first().ingredients()
-            print(dbSession.query(Recipe).filter(Recipe.recipe_id==key).first().ingredients())
+        outDict[calcOverlay(inputArr, ingred)] = (dbSession.query(Recipe).filter(Recipe.recipe_id==key).first().name, key, ingred)
-            #print(key)
     
+    print(outDict)
+
+
+def fastes():
+    indx = {}
+    inputArr2 = []
+
+    snowball = nltk.SnowballStemmer(language='german')
+    stopset = set(stopwords.words('german'))
+    for word in inputArr:
+
+        if word in stopset:
+             continue
+        inputArr2.append(snowball.stem(word))
+
+    for inpu in inputArr2:
+        ids = [] 
+        for xx in dbSession.query(Trunk).filter(Trunk.name == inpu).all():
+            for x in dbSession.query(Ingredient).filter(xx.ingredient_id == Ingredient.ingredient_id).all():
+                for y in x.recipe:
+                    
+                    if dbSession.query(Link).filter(Link.recipe_id==y.recipe_id).count() > len(inputArr) + maxMissing:
+                        continue   
+                    if str(y.recipe_id) not in indx:
+                        indx[str(y.recipe_id)] = 0
+
+                    indx[str(y.recipe_id)] += 1
+        
+    outDict = {}
+    for key, value in indx.items():
+        ingred = dbSession.query(Recipe).filter(Recipe.recipe_id==key).first().ingredients()
+        outDict[calcOverlay(inputArr, ingred)] = (dbSession.query(Recipe).filter(Recipe.recipe_id==key).first().name, key, ingred)
+    
+    print(outDict)
 #
 
-def stemIngred():
+def calcOverlay(l1, l2):
+    snowball = nltk.SnowballStemmer(language='german')
     stopset = set(stopwords.words('german'))
     stopset |= set("(),")
 
-    count = dbSession.query(Ingredient).count()
+    l1 =  [snowball.stem(l) for l in l1 ]
-    for x in dbSession.query(Ingredient).all():
+    counter = 0
-        snowball = nltk.SnowballStemmer(language='german')
+
-        for token in nltk.word_tokenize(x.name): 
+    for x in l2:
-            if token in stopset or len(token) < 3:
+        for token in nltk.word_tokenize(x): 
+            if token in stopset:
                 continue
             stemmed = snowball.stem(token)
+            for l in l1:
+                if l == stemmed:
+                    counter +=1
                      
-            x.trunks.append(Trunk(name=stemmed))
+    return counter
-            dbSession.commit()
+#
-        print(x.ingredient_id/count)
+
-#faster()    
+
-#slow()
+start = time.time()
-print(dbSession.query(Trunk.name).all())
+slow()
+end = time.time()
+print("\n", end - start, "\n")  
+
+
+start = time.time()
+faster()  
+end = time.time()
+print("\n", end - start, "\n")  
+
+
+start = time.time()
+fastes()
+end = time.time()
+print("\n", end - start, "\n")