Hangul Filler (U+3164): What It Is and How to Remove It
The Hangul filler is the character U+3164. It belongs to the Korean script, it draws as an empty gap in most fonts, and it is the one that people usually mean when they talk about a blank character that is not a space. Because Unicode classes it as a letter, ordinary space-trimming does not touch it. This guide explains what it is, why it looks empty, what can go wrong when it appears in your text, and how to remove it.
What is the Hangul filler?
The Hangul filler is a character in the block called Hangul Compatibility Jamo. Its official name is HANGUL FILLER and its code is U+3164. It was inherited from older Korean character standards, where a filler stood in for a missing piece when a syllable was built from its parts. It was never meant to be a space. It is a letter with no shape, and that is what makes it behave in surprising ways.
| Fact | Value |
|---|---|
| Name | HANGUL FILLER |
| Code point | U+3164, decimal 12644 |
| Unicode category | Letter, other (Lo) |
| UTF-8 bytes | E3 85 A4 (three bytes) |
| HTML | ㅤ or ㅤ |
| In code | "\u3164" in Python and JavaScript |
The family of Hangul fillers
There are four fillers, and all of them look blank.
| Code | Name | Notes |
|---|---|---|
U+3164 | Hangul filler | The one most often met in text. Usually a wide blank. |
U+FFA0 | Halfwidth Hangul filler | The narrow version, from the halfwidth and fullwidth forms block. |
U+115F | Hangul choseong filler | Stands for a missing first part of a syllable block. |
U+1160 | Hangul jungseong filler | Stands for a missing middle part of a syllable block. |
The last two belong to the conjoining jamo block, where letters combine into syllables. They are rarer in everyday text.
Why the Hangul filler looks blank
Unicode marks U+3164 as a default-ignorable code point. That is a property for characters that should not produce a visible mark by themselves. A program that has no glyph for such a character is supposed to draw nothing instead of showing a missing-character box. The result is that a font that does not contain a shape for U+3164 usually shows an empty gap instead of an error. In many fonts the gap is about as wide as a Korean syllable, though the exact width depends on the font.
Why it acts like a letter, not a space
Because U+3164 is a letter and not white space, the usual space tools miss it. Each of these checks was run on the character itself:
| Check | Result for U+3164 |
|---|---|
Python str.strip() | Does not remove it |
Python str.isspace() | False |
Python str.isalnum() | True, because it is a letter |
JavaScript trim() | Does not remove it |
JavaScript /\s/ | No match |
| Length of a string that contains only this character | 1 |
The practical effect: a form that rejects empty input after trimming accepts a value made only of U+3164, and a check that asks for "at least one letter" can accept it too. The value appears empty to a person and full to a program. A cell like that is the classic "empty but not blank" spreadsheet problem, covered in our guide to the blank character.
Where you might meet it
- In names and messages in games and chat apps, where a name that seems blank is really made of this character.
- In exported data. A CSV or a database column that holds a value you cannot see.
- In Korean text that was built from separate jamo, or converted between old and new encodings.
- In pasted web text where a site used it as a spacer.
Risks of leaving it in
- Hidden mismatches. Two names that look the same differ by one character, so lookups and de-duplication fail.
- Moderation. Many services block or strip it, because it is used to produce names that cannot be read. A string that contains it may be rejected or altered.
- Accessibility. A screen reader may skip it or read out its name. Either way, someone who depends on one gets a worse experience.
- Display. A font that draws a box for it, or an app that shows a placeholder, makes the "blank" suddenly visible.
How to find the Hangul filler
The AI Text Cleaner lists and removes U+3164, along with the other Hangul fillers, zero-width characters, special spaces and control codes.
# Python: print the position of each filler
for i, ch in enumerate(text):
if ch in "\u3164\uffa0\u115f\u1160":
print(i, f"U+{ord(ch):04X}")
On the command line, this prints each line of a file that contains U+3164, with its line number:
perl -CSD -ne 'print "$.: $_" if /\x{3164}/' file.txt
In Excel or Google Sheets, this counts it in cell A1:
=LEN(A1) - LEN(SUBSTITUTE(A1, UNICHAR(12644), ""))
How to remove it
# Python: remove all four fillers
import re
text = re.sub("[\u3164\uffa0\u115f\u1160]", "", text)
// JavaScript
text = text.replace(/[\u3164\uffa0\u115f\u1160]/g, "");
# Command line
perl -CSD -pe 's/\x{3164}//g' file.txt > cleaned.txt
In a spreadsheet, wrap the cell in SUBSTITUTE as shown above, with an empty string as the replacement. After that, run the text through the cleaner to deal with the zero-width characters and special spaces that often travel with it.
What to use instead
- For a gap in a layout, use a real space, or better, spacing set in CSS. In HTML, a space that must not break is
; see our notes on the non-breaking space. - For a placeholder in data, write a plain marker such as "N/A" or leave the field truly empty.
- For test data, write the character as an escape in your code,
"\u3164", so it is visible in the source and nobody copies it by accident.
Quick answers
Is the Hangul filler a space? No. It is a letter in the Korean script that has no visible shape.
Does it display the same everywhere? No. Most fonts draw it as a blank, but some draw a box, and its width changes from font to font.
Is it safe to delete? In English text, yes. If the text is Korean and the filler is part of a jamo sequence, check with someone who reads Korean before deleting it.