| 1 |
>>>>> On Wed, 22 Nov 2017, Michał Górny wrote: |
| 2 |
|
| 3 |
> Path and filename encoding |
| 4 |
> -------------------------- |
| 5 |
|
| 6 |
> The path fields in the Manifest file must consist of characters |
| 7 |
> corresponding to valid UTF-8 code points excluding the NULL character |
| 8 |
> (``U+0000``), the backwards slash (``\``) and characters classified |
| 9 |
> as whitespace in the current version of the Unicode standard |
| 10 |
> [#UNICODE]_. |
| 11 |
|
| 12 |
As I said before, all C0 and C1 control characters and DEL should be |
| 13 |
excluded as well, i.e. 0x00 to 0x1f, 0x7f, and 0x80 to 0x9f. Allowing |
| 14 |
such characters in what is basically a text file is only asking for |
| 15 |
trouble. |
| 16 |
|
| 17 |
> Any of the excluded characters that are present in path must be encoded |
| 18 |
> using one of the following escape sequences: |
| 19 |
|
| 20 |
> - characters in the ``U+0000`` to ``U+007F`` range can be encoded |
| 21 |
> as ``\xHH`` where ``HH`` specifies the zero-padded, hexadecimal |
| 22 |
> character code, |
| 23 |
|
| 24 |
> - characters in the ``U+0000`` to ``U+FFFF`` range can be encoded |
| 25 |
> as ``\uHHHH`` where ``HHHH`` specifies the zero-padded, hexadecimal |
| 26 |
> character code, |
| 27 |
|
| 28 |
> - characters in the UCS-4 range can be encoded as ``\UHHHHHHHH`` |
| 29 |
> where ``HHHHHHHH`` specifies the zero-padded, hexadecimal character |
| 30 |
> code. |
| 31 |
|
| 32 |
> It is invalid for backwards slash to be used in any other context, |
| 33 |
> and a backwards slash present in filename must be encoded. Backwards |
| 34 |
> slash used as path component separator should be replaced by forward |
| 35 |
> slash instead. |
| 36 |
|
| 37 |
This entire section about the escape mechanism should be clearly |
| 38 |
labelled as being purely optional, as it is not relevant for Gentoo |
| 39 |
(and would break backwards compatibility with existing package |
| 40 |
manager implementations). Maybe add a reference to GLEP 31 too? |
| 41 |
|
| 42 |
> The encoding can be used for other characters as well. In particular, |
| 43 |
> escaping control characters is recommended to ensure that the file |
| 44 |
> works correctly in text editors. |
| 45 |
|
| 46 |
See above, this should not be "recommended", but literal control chars |
| 47 |
should be strictly forbidden. |
| 48 |
|
| 49 |
Ulrich |
Archives