Dodani popravki za iskanje kandidatov in števila besedil. Dodana logika za klicanje ocrja. Popravljen bug za timeout v ateapi-ju. Popravljen swager.

swagger_server/utils/db_utils.py

@@ -78,6 +78,8 @@ def vrni_oss_dokumente(leta, vrste, kljucnebesede, udk):
         
         print(sql)
         print(params)
+
+
         
         cur.execute(sql,params)
         
@@ -89,6 +91,81 @@ def vrni_oss_dokumente(leta, vrste, kljucnebesede, udk):
     return ret
 
 
+def vrni_oss_terminoloske_kandidate(leta, vrste, kljucnebesede, udk):
+    ret = []
+  
+    try:
+        print(database_info)
+        conn = mariadb.connect(**database_info)
+        cur = conn.cursor()
+        
+        
+        
+        sql = "select distinct document_id from metadata"
+        where=""
+        params=[]
+        if (udk):
+            where_in_udk = ','.join(['%s'] * len(udk))
+            where=" udk IN (%s) " % (where_in_udk)
+            params=udk
+        
+        if (leta):
+
+            where_in_leta = ','.join(['%s'] * len(leta))
+            if (where):
+                where=where+ " AND "
+            where=where + " leto IN (%s) " % (where_in_leta)
+            params=params+leta
+        
+        if (vrste):
+            if (where):
+                where=where+ " AND "
+            where_in_vrste = ','.join(['%s'] * len(vrste))
+            where=where + " tipologija IN (%s) " % (where_in_vrste)
+            params=params+vrste
+
+        if (kljucnebesede):
+            if (where):
+                where=where+ " AND "
+            where_in_kb = ','.join(['%s'] * len(kljucnebesede))
+            where=where + " kljucnabeseda IN (%s) " % (where_in_kb)
+            params=params+kljucnebesede
+
+        if(where):
+            sql=sql+" where " + where
+        
+        print(sql)
+        print(params)
+
+        sqltk=f"""Select ngram,upos,avg(tfidf) as tfidf from (
+                    SELECT tf.ngram, tf.upos,(0.5+0.5*(tf.tf/d.maxtf))*log(152000/df.df)*(-1*log(1-((dff.df)/(1+df.df)))) as tfidf
+                    FROM ngrams_upos_tf tf, documents d,
+                        (
+                        Select ngram, upos, count(*) as df from ngrams_upos_tf TF
+                        where document_id in
+                    ({sql})
+                    group by TF.ngram, TF.upos
+                    ) dff, ngrams_upos_df df
+                    where
+                    tf.document_id=d.document_id and
+                    df.ngram=tf.ngram AND df.upos=tf.upos and
+                    dff.ngram=tf.ngram AND dff.upos=tf.upos
+                    ) X
+                    group by ngram,upos
+                    order by tfidf desc
+                    limit 100"""
+        
+        print (sqltk)
+        
+        cur.execute(sqltk,params)
+        
+        ret = list(cur)
+    except mariadb.Error as e:
+        print(f"Error connecting to MariaDB Platform: {e}")
+    
+
+    return ret
+
 # class BaseModel(Model):
 #     class Meta:
 #         database = db