Talk:String length: Difference between revisions
→Incorrect: please provide a reference file
(Component Pascal is UTF-16 only.) |
Walterpachl (talk | contribs) (→Incorrect: please provide a reference file) |
||
| (4 intermediate revisions by 3 users not shown) | |||
Line 14:
The example for character length does not deal with utf-8 and as much as I understand also fails with Non-BMP code points.
== Incorrect ==
The byte length calculations for unicode appear generally incorrect. They're only valid for codepoints which are in the Basic Multilingual Plane, but not for the Supplemental planes. I.e. 🀁 wouldn't fit within a single wide character; it would be represented in UTF-16 as 0xD38C and 0xDC01 (if I've done the math right). --[[User:Short Circuit|Michael Mol]] 18:24, 15 March 2012 (UTC)
: If you want to be completely general, there exist [https://en.wikipedia.org/wiki/Unicode_normalization other issues to consider.] Note, in particular, that not all combining forms have codepoints. --[[User:Rdm|Rdm]] 18:31, 15 March 2012 (UTC)
:: could we see a text file that contains the various BYTE strings and the expected length results? ..[[User:Walterpachl|Walterpachl]] ([[User talk:Walterpachl|talk]]) 08:03, 30 August 2014 (UTC)
== PL/I error ==
the last line ( put skip list ('Byte length=', length(trim(SM)); )
<br>is syntactically incorrect (a closing parenthesis is missing)
<br>I tried to add it and get an error message:
<br>IBM1569I S 9.0 SIZE argument must be a CONNECTED reference.
<br>--[[User:Walterpachl|Walterpachl]] ([[User talk:Walterpachl|talk]]) 18:43, 22 October 2013 (UTC)
| |||