File size: 8,412 Bytes
89a1200
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c9a96c4
89a1200
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9870ed
 
 
 
 
 
 
89a1200
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
import os
import sys
import streamlit as st
from streamlit_extras.switch_page_button import switch_page
import pandas as pd
import numpy as np
import torch
import faiss
from sentence_transformers import SentenceTransformer
import csv

################################
######### Variables ############
################################
# -- Loading Variables
script_directory = os.path.dirname(os.path.abspath(sys.argv[0]))
source_df = pd.DataFrame()
destination_df = pd.DataFrame()
model = SentenceTransformer('all-mpnet-base-v2')

# -- Loading Session Data
if 'project_data' not in st.session_state:
    st.session_state.project_data = pd.read_csv(script_directory+'/data/project.csv')

################################
####### GenericFunctions #######
################################
# -- Create Embedding - all-mpnet-base-v2 - https://www.sbert.net/docs/pretrained_models.html
def embed_text(text):
    embedding = model.encode(text)
    return embedding

def embed_list(list):
    embeddings = []
    for text in list:
        embeddings.append(embed_text(text))
    return embeddings

# -- Store embeddings in a FAISS Vector database
def store_embeddings(embeddings):
    dimension = embeddings[0].shape[0]
    index = faiss.IndexFlatIP(dimension)
    index.add(np.array(embeddings))
    # faiss.write_index(index, "data/vector_db.index")
    return index

# -- Perform semantic search using embeddings
def semantic_search(query_embedding, index, k=1):
    D, I = index.search(np.array([query_embedding]), k)
    return I[0][0]

################################
####### Display of data ########
################################
# -- Streamlit Settings
st.set_page_config(layout='wide')
st.title("Mapping")

# -- Add Project Dropdown
st.text("")
st.text("")
st.text("")
col1, col2, col3 = st.columns(3)
option = col1.selectbox('Select Project',st.session_state.project_data['Project'])
col1, col2, col3 = st.columns(3)


# -- Destination File Name
st.text("")
st.text("")

col1, col2, col3 = st.columns(3)
cond = (st.session_state.project_data['Project'] == option)
result = st.session_state.project_data[cond].Destination.values[0]
with col1:
    destination_file_format = st.file_uploader(
        "Destination file name - "+str(result)+".csv",
        type="csv",
        key="destination_file_format",
        accept_multiple_files=True
    )
    
if destination_file_format is not None:
    for file in destination_file_format:
        destination_df = pd.read_csv(file)

# -- Source File Name
cond = (st.session_state.project_data['Project'] == option)
result = st.session_state.project_data[cond].Source.values[0]
with col3:
    source_file_format = st.file_uploader(
        "Source file name - "+str(result)+".csv",
        type="csv",
        key="source_file_format",
        accept_multiple_files=True
    )

if source_file_format is not None:
    for file in source_file_format:
        source_df = pd.read_csv(file)

# -- Suggest Button
st.text("")
st.text("")
col1, col2, col3 = st.columns([0.25,0.2,2.55])
if col1.button("AI Suggest"):
    st.session_state.mapping_df = pd.DataFrame(columns=["Sno","DestinationColumn","SourceColumn","Type","Expression"])
    if len(destination_df) == 0 or len(source_df) == 0: 
        st.error("Select Source and Destination Files")
    else:
        new_data = []
        
        # Source - KnowledgeBase
        input_text = source_df["Columns"].tolist()
        embeddings = embed_list(input_text)
        index = store_embeddings(embeddings)
        
        # Map to Source
        for i in range(len(destination_df)):
            search_text = destination_df.loc[i, "Columns"]
            query_embeddings = embed_text(search_text)
            result = input_text[semantic_search(query_embeddings, index)]
            row = {
                    "Sno": i+1,
                    "DestinationColumn": destination_df.loc[i, "Columns"],
                    "SourceColumn": result,
                    "Type": None, 
                    "Expression":None
                }
            new_data.append(row)
        
        # Saving Mapping and displaying
        if new_data or len(mapping_df) <0:    
            st.session_state.mapping_df = pd.concat(
                                                        [ st.session_state.mapping_df, pd.DataFrame(new_data)], 
                                                        ignore_index=True
                                                    )
        else:
            st.error("Unable to map Source and Destination Files")

# -- Save Button
if col2.button("Save"):
    if (len(destination_df) > 0 and len(source_df) > 0 and len(st.session_state.mapping_df)>0):
        cond = (st.session_state.project_data['Project'] == option)
        file_name = script_directory+'/data/'+str(st.session_state.project_data[cond].Id.values[0])+"_"+st.session_state.project_data[cond].Source.values[0]+"_"+st.session_state.project_data[cond].Destination.values[0]+'.csv'
        st.session_state.mapping_df.to_csv(file_name, index=False, sep="|",quoting=csv.QUOTE_NONE)
    else:
        st.error("Transformation not created")

# -- Load Exisitng Mapping
if col3.button("Load Mapping"):
    cond = (st.session_state.project_data['Project'] == option)
    file_name = script_directory+'/data/'+str(st.session_state.project_data[cond].Id.values[0])+"_"+st.session_state.project_data[cond].Source.values[0]+"_"+st.session_state.project_data[cond].Destination.values[0]+'.csv'
    st.session_state.mapping_df = pd.read_csv(file_name,sep="|",quoting=csv.QUOTE_NONE)

# -- Display Mapping Table
if (len(destination_df) > 0 and len(source_df) > 0 and len(st.session_state.mapping_df)>0):
    st.text("")
    st.header("Mapping Details")
    st.text("")
    st.text("")
    st.session_state.mapping_df = st.data_editor(
                                                    st.session_state.mapping_df,
                                                    height=400,
                                                    width=1200,
                                                    hide_index=True,
                                                    column_config={
                                                                        "Sno": st.column_config.TextColumn(
                                                                            "Sno"
                                                                        ),
                                                                        "DestinationColumn": st.column_config.TextColumn(
                                                                            "DestinationColumn"
                                                                        ),
                                                                        "SourceColumn": st.column_config.SelectboxColumn(
                                                                            "SourceColumn",
                                                                            width="medium",
                                                                            options= source_df["Columns"],
                                                                        ),
                                                                        "Type": st.column_config.SelectboxColumn(
                                                                            "Type",
                                                                            width="medium",
                                                                            options=[
                                                                                "Pandas",
                                                                                "Constant"
                                                                            ]
                                                                        ),
                                                                        "Expression": st.column_config.TextColumn(
                                                                            "Expression"
                                                                        )
                                                                    },
                                                                    disabled=["Sno","DestinationColumn"]
                                                )