"""Merging chats."""
import pandas as pd
def _merge_two_chats(df1, df2):
if df1.index.min() <= df2.index.min():
df = pd.concat([df1, df2[df2.index > df1.index.max()]])
else:
df = pd.concat([df2, df1[df1.index > df2.index.max()]])
return df
[docs]def merge_chats(dfs):
"""Merge several chats into a single one.
Can come in handy when you have old exports and new ones, and both have relevant data.
**Note:** The dataframes must have an index with the timestamps of the messages, as this is required to correctly
sort and merge the chats.
Args:
dfs (List[pandas.DataFrame]): List with the chats as DataFrames.
Returns:
pandas.DataFrame: Merged chat.
"""
# Sort from oldest
dfs = sorted(dfs, key=lambda x: x.index.min())
# Merge
df = dfs[0]
for i in range(1, len(dfs)):
df = _merge_two_chats(df, dfs[i])
return df
