ExportBars - bars to CSV that pandas reads without arguments - script for MetaTrader 5

ExportBars is a script that writes the bars of any symbol and timeframe to a CSV file prepared for analysis outside the terminal. The file loads with a plain pandas.read_csv() call, with no encoding, separator or date-format arguments. It covers the routine first step of any research: getting the terminal's own quotes into Python or a spreadsheet exactly as the terminal stores them.

EURUSD H1 exported with the default settings: the raw file and the same file loaded with pandas.read_csv()

The problem

A file written by an MQL5 program with the default flags is UTF-16, which read_csv does not parse until the encoding is passed explicitly. TimeToString produces dates such as 2026.10.02 10:00, with dots and without seconds, which need a custom format string. Prices converted without the symbol's digits may come out with a different number of decimals from row to row.

ExportBars fixes these three points in the file itself, so the loading code stays one line and the same file can be handed to somebody else without instructions. A switchable separator is added for Excel installations that use a comma as the decimal mark, where a comma-separated file is split into the wrong columns.

What the file looks like

• One header row: time, open, high, low, close, tick_volume, spread, real_volume.
• The time column is the bar opening time in server time, written in ISO 8601 form: YYYY-MM-DD HH:MM:SS.
• Prices are written with the number of digits of the symbol (SYMBOL_DIGITS), never in scientific notation.
• Rows are in chronological order, oldest bar first, the same order as on the chart.
• The text is single-byte (FILE_ANSI), not UTF-16.

Input parameters

Loading the file

With the default comma separator no arguments are required. Because the timestamps are in ISO 8601 form, adding parse_dates turns the time column into datetime values without a format string. For a file exported with a semicolon or a tab, pass the matching sep argument.

import pandas as pd

df = pd.read_csv("EURUSD_H1.csv", parse_dates=["time"])

Notes

• The result is reported in the Experts log: symbol, timeframe, number of bars, time of the first and the last bar, and the file path.
• If the history for the requested bars is not available, the script prints the error code and does not create a file.
• An existing file with the same name is overwritten.
• In the Last N bars mode the last row is the current bar, which is still forming and will change until it closes.
• Volumes and spread are written as they are stored in the bar history; real_volume is 0 for symbols without exchange volume.