CSV 파일의 글자 깨짐(mojibake)과 인코딩 오류 해결

What Is Mojibake?

Mojibake is the visual garbage that appears when a file saved in one character encoding is opened or interpreted in a different one. The word comes from Japanese and roughly means "unintelligible characters."

If your CSV contains sequences like these, you have an encoding problem:

| What you see | What it should be | |---|---| | ’ | ' (right single quotation mark) | | â€" | — (em dash) | | “ | " (left double quote) | | é | é (e with accent) | |  at the start of a field | BOM (invisible marker) |

Why Encoding Mismatches Happen

Modern platforms — Shopify, WooCommerce, Amazon — expect UTF-8 encoded files. But many tools produce files in other encodings:

• Microsoft Excel (Windows) saves .csv in the system locale encoding (often Windows-1252 or Latin-1) unless you specifically choose UTF-8.
• Older export tools may default to ISO-8859-1 / Latin-1.
• Copy-paste from websites can smuggle in curly quotes and em dashes encoded as multi-byte UTF-8 characters that break when the file is treated as Latin-1.

When a UTF-8 file is read as Latin-1, each multi-byte sequence gets split into individual Latin-1 characters, producing mojibake.

Diagnosing the Problem

Check the encoding of your file

On Mac/Linux:

file -I yourfile.csv
# Example output: yourfile.csv: text/plain; charset=utf-8

On Windows (PowerShell):

Get-Content yourfile.csv -Encoding Byte -TotalCount 3
# If first 3 bytes are 239 187 191 (0xEF 0xBB 0xBF), the file has a UTF-8 BOM.

Quick visual test: Open the file in a text editor (not Excel) and look at the first character and any accented letters. If the first cell starts with  or descriptions contain â€, the encoding is wrong.

Identify which encoding to use

• Most garbled files are UTF-8 read as Latin-1 (or vice versa).
• The fix is almost always: re-save as UTF-8 without BOM.

What Is the BOM?

The UTF-8 BOM (Byte Order Mark) is a three-byte sequence (EF BB BF) that some tools prepend to UTF-8 files. Windows Notepad and Excel sometimes add it automatically.

Why it matters: Import tools — including Shopify's product CSV importer — may interpret the BOM as literal characters in the first column name, turning Handle into Handle. That column is then not recognized and the entire import fails silently.

Fix in Excel

Excel is the most common cause of encoding problems. Follow these steps precisely.

Export from Excel as UTF-8

1. Open your spreadsheet in Excel.
2. Click File → Save As.
3. In the "Save as type" dropdown, choose CSV UTF-8 (Comma delimited) (.csv).
   • This option is available in Excel 2016 and later on Windows and Mac.
   • Do not choose plain "CSV (Comma delimited)" — that saves in the system locale encoding.
4. Click Save.

Re-open a mojibake file in Excel

If you already have a garbled file:

1. Do not double-click the file to open it. That uses the system locale.
2. Open a blank Excel workbook.
3. Go to Data → From Text/CSV.
4. Select your file.
5. In the import wizard, set File Origin to 65001: Unicode (UTF-8).
6. Complete the import, then save as CSV UTF-8.

Fix in Google Sheets

Google Sheets always works in UTF-8 internally. The risk is in downloading:

1. Import: Go to File → Import → Upload and select your CSV. Sheets detects encoding automatically.
2. Export: Go to File → Download → Comma Separated Values (.csv). Google Sheets exports UTF-8 without BOM. This is safe.

If you received a garbled file that Sheets is displaying incorrectly:

1. Import it via File → Import rather than opening it directly.
2. In the import dialog, check Convert text to numbers, dates, and formulas is unchecked if you have SKUs or barcodes.

Fix with VS Code (Reliable for Any File)

VS Code gives you full control over encoding and is free:

1. Open VS Code and drag your CSV file into it.
2. Look at the bottom-right status bar — it shows the detected encoding (e.g., UTF-8, Windows 1252).
3. Click the encoding label.
4. Choose "Reopen with Encoding" and select the encoding the file was actually saved in (often Western (Windows 1252) or ISO 8859-1).
5. The garbled text should now display correctly.
6. Click the encoding label again and choose "Save with Encoding → UTF-8".

This two-step process — reopen with the original encoding, save as UTF-8 — is the most reliable method for any tool.

Fix with Windows Notepad

For a quick fix on Windows (Windows 11 / 10 version 1903+):

1. Open the garbled file in Notepad.
2. Go to File → Save As.
3. In the Encoding dropdown, select UTF-8 (not "UTF-8 with BOM").
4. Save.

Note: Older versions of Notepad only offered "UTF-8 with BOM" which adds the invisible BOM marker. If your Notepad does not have a "UTF-8" (without BOM) option, use VS Code instead.

Preventing Re-encoding Corruption

Once your file is clean, protect it going forward:

• Always use "CSV UTF-8" in Excel, never plain CSV.
• Never re-open a UTF-8 CSV by double-clicking in Windows. Windows associates .csv with Excel which may open it in the locale encoding. Use File → Open from within Excel and specify the encoding, or use Data → From Text/CSV.
• Validate encoding after every export using a tool like StriveFormats or a command-line check before uploading.
• If your import platform's export tool consistently produces Latin-1, convert the file immediately after downloading using VS Code or a script before editing.

Quick Reference: Encoding Symptoms and Fixes

| Symptom | Likely cause | Fix | |---|---|---| | ’ â€" in text | UTF-8 file opened/saved as Latin-1 | Reopen as Latin-1, save as UTF-8 | |  at start of first column | UTF-8 BOM | Strip BOM; use "UTF-8 without BOM" | | é è à | UTF-8 multi-byte split | Reopen as Latin-1, save as UTF-8 | | ? replacing all special chars | Encoding lost on conversion | Original data may be lost; re-export from source | | Correct in Sheets but garbled when downloaded | Download encoding wrong | Use File → Download → CSV (Google uses UTF-8) |