View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0001946 | 1003.1(2024)/Issue8 | Base Definitions and Headers | public | 2025-09-13 02:27 | 2025-09-25 16:16 |
| Reporter | calestyo | Assigned To | |||
| Priority | low | Severity | Editorial | Type | Clarification Requested |
| Status | Resolved | Resolution | Accepted As Marked | ||
| Name | Christoph Anton Mitterer | ||||
| Organization | |||||
| User Reference | |||||
| Section | 5 or 7.3 | ||||
| Page Number | 113 or 129 | ||||
| Line Number | 3513, ff. or 4157-4162 | ||||
| Interp Status | --- | ||||
| Final Accepted Text | 0001946:0007279 | ||||
| Summary | 0001946: do locale definition file global keywords allow 1-n blanks after the keyword or does it have to be 1 space | ||||
| Description | Hey. Merely a cosmetic issue: The "comment_char %c\n", <comment character> and "escape_char %c\n", <escape character> (as well as perhaps their analogues in Character Set Description File) look just like the syntax format defined in 5 File Format Notation: "<format>", [<arg1>, <arg2>,..., <argn>] where " " is one or more blanks and ∆ is exactly one space. However, it doesn't seem to me as if that format is defined to also apply to e.g. 7.3 Locale Definition So AFAICS, it's not really clear here, whether the escape_char keyword and friends need to be followed by a single space, or whether multiple blanks are allowed. PS: I don't want to open separate issues about these (as they're not really issues, IMO): 1) Maybe it should be mention that ISO/IEC TR 14652 goes beyond what POSIX specifies, not only in terms of more categories but also in adding keywords to existing ones and even extending the definitions of existing keywords (like "abday" and "day" where POSIX always starts with Sunday, but the TR not necessarily) 2) Throughout the whole standard there are symbols like: _POSIX2_LOCALEDEF, where however also a form without the leading _ is used. Maybe I got it wrong, but I thought the symbol is always *with* the leading _ and just some utilities also accept the name without? If so, there would be some places where the standard still talks about `symbol foo` where `foo` has no leading _. | ||||
| Desired Action | Clarify the above :-) | ||||
| Tags | tc1-2024 | ||||
|
|
On the leading underscore thing, you are probably thinking of getconf, which is required to accept some of these symbols without the leading underscore for backwards compatibility. |
|
|
On page 129 line 4156,4161 section 7.3, change:It shall have the following format, starting in column 1 to: It shall have the following format, interpreted as described in [xref to XBD 5], starting in column 1 |
|
|
ISO/IEC TR 14652 has been withdrawn by ISO, so it would not be appropriate to refer to it in POSIX. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2025-09-13 02:27 | calestyo | New Issue | |
| 2025-09-18 16:29 | geoffclare | Note Added: 0007270 | |
| 2025-09-25 16:15 | geoffclare | Note Added: 0007279 | |
| 2025-09-25 16:15 | geoffclare | Status | New => Resolved |
| 2025-09-25 16:15 | geoffclare | Resolution | Open => Accepted As Marked |
| 2025-09-25 16:15 | geoffclare | Interp Status | => --- |
| 2025-09-25 16:15 | geoffclare | Final Accepted Text | => 0001946:0007279 |
| 2025-09-25 16:16 | geoffclare | Tag Attached: tc1-2024 | |
| 2025-09-25 16:16 | geoffclare | Note Added: 0007280 |
