JIS Function (OpenOffice Calc)
Convert half-width Latin letters, digits, and katakana into full-width text using Apache OpenOffice Calc JIS.
Verification status: Official documentation referenced; example outcomes independently reasoned or arithmetically checked where stated. Not executed in OpenOffice Calc. How formula examples are checked
Quick answer
Use JIS to convert supported half-width Latin letters, digits, punctuation, and Japanese katakana into the full-width character forms described by Apache’s Calc specification. It is the opposite direction from the ASC function and can help format Japanese customer names or product fields where consistent full-width text is required. [S1, S2]
=JIS("ABC123")
Expected text result: ABC123. This is a character conversion, not a change to the cell’s font or zoom. [S1]
What JIS changes
Apache specifies a table-driven conversion, not indiscriminate replacement of every narrow-looking symbol. It maps ordinary ASCII U+0021–U+007E, with documented special exceptions, and selected half-width katakana code points to full-width forms. Characters outside the specified mappings are copied unchanged. Half-width base kana followed by voiced/semi-voiced marks may be combined into a single full-width character. [S1]
It does not automatically turn ASCII space U+0020 into ideographic space U+3000, nor does it apply arbitrary Japanese transliteration or Latin case conversion. JIS is useful for presentation- or standards-driven Japanese text fields, but always preserve a copy of the source value. [S1]
Exact documented syntax and arguments
Apache’s implemented-feature specification lists: [S1]
JIS(Text T)
Enter it in a Calc cell as:
=JIS(A2)
=JIS("ABC123")
Text T— required: a text value or cell containing the characters to convert. One parameter only, with no optional arguments or default values in the official specification. [S1]- Return type: text; the source cell is not changed by the formula. [S1]
- Conversion limits: selected Latin, half-width katakana, and punctuation conversions; other characters are retained. Apache explicitly documents special handling of quotation marks, apostrophes, backslashes and grave accents. [S1]
- Locale: these are English-name examples from the Apache specification. Multi-argument formulas below use the semicolon separator in Apache’s documented
SUBSTITUTEandEXACTsyntax; the JIS call itself has just one argument. Verify your localized Function Wizard. [S1, S6–S7] - Version/platform: Apache’s original feature specification and catalogue establish documented functionality but are not a test on the installed 4.1.16 build. No Windows, macOS or Linux result was observed. [S1, S2, S8]
Reproducible worksheet: half-width to full-width
In a new Calc document, enter the literal text into cells A2:A10, including any narrow katakana and punctuation. Put =JIS(A2) in B2 and fill down through B10. The outputs are expected, independently derived from Apache’s Unicode mapping table. [S1]
| Cell | A: source text | B: expected result | Explanation |
|---|---|---|---|
| A2 | ABC123 |
ABC123 |
ASCII letters/digits become full-width |
| A3 | カタカナ |
カタカナ |
Half-width katakana becomes full-width |
| A4 | ガ |
ガ |
Base plus voiced mark combine |
| A5 | パ |
パ |
Base plus semi-voiced mark combine |
| A6 | Sales 12 |
Sales 12 |
Ordinary ASCII space is left unchanged |
| A7 | こんにちは |
こんにちは |
Hiragana is outside the documented map |
| A8 | \ |
¥ |
Backslash U+005C becomes full-width yen sign U+FFE5 |
| A9 | 「テスト」 |
「テスト」 |
Japanese brackets and katakana map |
| A10 | " |
” |
ASCII double-quote becomes right double quotation mark |
Important: The ASCII space in row A6 remains U+0020 despite the widening of adjacent Latin letters. The \ in A8 is one backslash character. A10’s result is the curly right quotation mark U+201D, not the visually similar full-width straight quote U+FF02. [S1]
Direct formula checks
=JIS("ガ")
=JIS("パ")
=JIS("\")
=JIS("")
Expected: ガ, パ, ¥, and an empty text result. The empty-input result is derived from the character-by-character transformation and has not been native-executed. [S1]
Advanced examples
Widen an ASCII space deliberately
Apache’s documented mapping does not include U+0020. To replace ordinary spaces with U+3000 after JIS conversion, use SUBSTITUTE explicitly. [S1, S6]
=SUBSTITUTE(JIS("Sales 12");" ";" ")
Expected: Sales 12 with an ideographic space U+3000 between the words. The third literal above contains U+3000. This may be useful for a mandated display format, but do not alter source records unless the destination requires it.
Compare strings after adopting a full-width standard
If A2 contains ABC123 and C2 contains ABC123, you can compare their full-width representations with Apache’s EXACT function. [S1, S7]
=EXACT(JIS(A2);JIS(C2))
Expected: TRUE. The input texts remain different Unicode strings; the formula is an intentional normalization comparison, not evidence that the originals are identical.
Voiced marks and length
Apache’s mapping combines two half-width characters (カ plus ゙) into one full-width ガ. Consequently LEN(JIS("ガ")) has an expected result of 1, even though the original ガ contains two Unicode characters. [S1, S5]
=LEN(JIS("ガ"))
Expected: 1. This is a character count, not a measurement of bytes or visual width.
Quotation-mark exception
Apache does not apply a simple ASCII-to-full-width block mapping to every symbol. It maps ASCII quotation mark U+0022 to right double quotation mark U+201D. In Calc text literals, a quotation mark inside a string is represented by two successive double quotes. Thus: [S1]
=JIS("A""B")
Expected text: A”B (with U+201D between the full-width letters). This example is documentation-derived and still requires native validation.
Troubleshooting and limitations
| Situation | What to check |
|---|---|
| Text appears unchanged | The characters may already be full-width or outside Apache’s mapping, such as hiragana. [S1] |
| Spaces remain narrow | ASCII U+0020 is not a mapped character in the spec. Use SUBSTITUTE if U+3000 is required. [S1, S6] |
| Kana length shrinks | A half-width kana plus voiced mark may become one full-width character. Compare code points. [S1, S5] |
| Backslash appears as a yen symbol | Apache maps U+005C to U+FFE5; Japanese fonts may also affect how some code points look. Do not infer identity from glyph appearance. [S1] |
| Quotes unexpectedly become curly | U+0022 maps to U+201D as a documented exception. [S1] |
| Excel import or sharing changes outputs | Microsoft documents locale-specific names/behaviour for this family (JIS or DBCS). Test a copy in the actual recipient program. [S4] |
| Error or unknown function | Check spelling, cell contents and localized formula interface. Runtime error codes are unverified here; the Apache spec does not supply a comprehensive invalid-type error matrix. |
JIS is not a universal Unicode normalization tool, and its name is not a promise to encode strings in JIS X 0208. It replaces selected characters in Calc text values according to the documented tables. [S1]
Compatibility and historical release scope
- Apache OpenOffice Calc: an Apache-authored specification identifies
JIS(Text T)as implemented and the Apache function catalogue includes JIS; the current 4.1.16 executable was not tested here. [S1, S2, S8] - Microsoft Excel: Microsoft describes JIS/DBCS-style conversion and explicitly says function names and affected characters depend on language settings. A formula that looks equivalent is not guaranteed to return every same code point. [S4]
- LibreOffice Calc: similar functions exist in its own documentation, but LibreOffice information is not proof of an Apache OpenOffice result; validate both applications if sharing workbooks. [S1]
Related functions and verified references
The companion ASC function maps the other direction. LEN, SUBSTITUTE, and EXACT support text checks and deliberate cleanup.
Official sources (accessed 2026-10-11):
- Apache’s implemented JIS and ASC specification with Unicode tables (
S1, last changed 2007-07-25). - Apache Calc text-functions/category catalogue (
S2, historical catalogue). - Apache Calc LEN (
S5), SUBSTITUTE (S6), and EXACT (S7). - Apache releases/downloads (
S8, currently lists 4.1.16). - Microsoft Excel JIS/DBCS guidance (
S4, interoperability comparison only).
Validation note
The examples show expected results checked against Apache OpenOffice documentation and independent arithmetic, text mapping, or logical reasoning. They were not run in Apache OpenOffice Calc. Platform, locale, import, and unusual error cases are not certified.