JIS plane 1, part 1

Note: Compare behaviour of GB 01-04.

Note: U+FFE3 (￣) is the fullwidth counterpart of both U+00AF (¯) and U+203E (‾); the latter is the one typically used when mapping JIS C 6220 / JIS X 0201. Mapping the double-byte character to U+203E in contexts where 0x7E is mapped to U+007E (~) is done by OSF in ASCII-based EUC-JP only, and by JIS X 0213 (2000 JIS / 2004 JIS) in EUC-JP only (and then not by all encoders). Compare GB 00-03-94.

Note: U+2014 (em dash) and U+2015 (horizontal bar) both correspond to the same JIS character. UTC's mappings generally favoured U+2015, and Microsoft and consequently WHATWG (HTML5) follow this. JIS (JIS X 0221, JIS X 0213) consider it to be U+2014, and Apple follows suit; Python apparently considers it to be U+2015 even in its JIS X 0213 codecs. OSF maps it as U+2015 in their MS-based version and as U+2014 otherwise. Compare behaviour of GB 00-01-10.

The codepoint selected to display in the NEC column is arbitrary.

Note: U+005C (backslash) is sometimes rendered the same as U+00A5 (¥), especially when it's used to map the 7-bit code 0x5C (backslash in ASCII, yen sign in JIS C 6220 / JIS X 0201). Generally speaking, the double byte character is only mapped to U+005C if is is not used as the mapping for 0x5C: OSF map it as U+005C in JIS-Roman-based EUC but not the other versions, and JIS X 0213 (2000 JIS / 2004 JIS) maps the double byte character to U+005C in Shift_JIS but to U+FF3C in EUC.

Note: Various mappings of various legacy CJK sets map this character to either U+223C (tilde operator), U+301C (wave dash), or U+FF5E (fullwidth tilde). Of these: U+301C was allocated specifically for the JIS character but was displayed in the Unicode charts with curvature inverted relative to the JIS charts for a considerable time (and is still displayed as such by some older fonts such as MS PMincho). U+223C is primarily intended as a mathematical operator, and usually shorter

U+FF5E is just the fullwidth version of the ASCII character and might therefore be rendered as either a tilde dash or a spacing tilde accent, of any glyph width, within a roughly 1em advance—although a wave dash is the most common by far (however, U+FF5E actually has a separate mapping at 02-02-23 in JIS X 0212 and 01-02-18 in JIS X 0213, which is shown as a spacing accent in their respective charts).

Compare, and especially contrast, the behaviour of GB 01-11 and KSC 01-01-13.

The codepoint selected to display in the NEC column is arbitrary.

Note: U+2225 is used by Microsoft-influenced mappings, U+2016 is used by others. U+2225 has a separate mapping in JIS X 0213 (01-02-52). U+2016 is necessarily straight vertical, whereas U+2225 is often shown slanted. Compare GB 01-12 and GB 01-46.

The codepoint selected to display in the NEC column is arbitrary.

Note: Apple initially mapped this to U+22EF and the single-byte Shift JIS 0xFF to U+2026. Since this apparently created interoperability issues, they (Apple) revised these mapping to U+2026 and U+2026+(F87F) respectively, where U+F87F is a private use character used for round tripping. Compare GB 00-01-13, for which Apple apparently retained the U+22EF mapping.

Note: Microsoft-influenced mappings use U+FF0D for the minus sign (making the JIS minus sign, as opposed to the JIS hyphen, the definitive fullwidth form of the ASCII hyphen-minus). Others use the definitive minus sign codepoint (U+2212). Also: WHATWG's (HTML5) encoders exceptionally treat U+2212 and U+FF0D the same (while its decoders use U+FF0D), since doing otherwise was breaking Japanese postcode forms on Macintoshes. Note a JIS X 0213 mapping for U+FF0D at 01-02-17. Compare/contrast GB 03-13.

The codepoint selected to display in the NEC column is arbitrary.

Note: Mapping the double-byte character to U+00A5 is only done when that mapping isn't already used for 0x5C, e.g. JIS X 0213 (2000 JIS / 2004 JIS) does so for EUC but not Shift_JIS. OSF use U+00A5 here in their ASCII-based EUC mappings, but not in their MS-based or JIS-Roman-based EUC mappings.