mirabilos authored 5 years ago and mirabilos committed 5 years ago

(upstream already does the protect and incsname for us)

mirabilos authored 5 years ago and mirabilos committed 5 years ago

(utf8.def is now included in the “fmt file” instead of being
 processed at compile time)

cf. https://www.overleaf.com/learn/latex/Articles/The_two_modes_of_TeX_engines:_INI_mode_and_production_mode

mirabilos authored 5 years ago and mirabilos committed 5 years ago

(it’s a fully-merged bugfix so it fits)

mirabilos authored 5 years ago and mirabilos committed 5 years ago

the fixes were added just before \UTFviii@hexcodepoint so use that

mirabilos authored 5 years ago and mirabilos committed 5 years ago

• Teχ style for hex ("00)
• Pascal style in WTF-8 for ranges (a‥b)
• UCS vs. UTF-8

mirabilos authored 5 years ago and mirabilos committed 5 years ago

• add/update comments from PR
• provide matching \UTFviii@defined even if unicodedomino.sty
  immediately replaces this with its improved version, to
  facilitate reusability not dependent on unicodedomino

mirabilos authored 5 years ago and mirabilos committed 5 years ago

• rename merged part (10FFFF upper bound in \parse@XML@charref
  and fix "F4 lead octet) ⇒ unicodedomino_kernel_fixup_f4.def
• move not merged part (better \UTFviii@checkseq) together with
  the remaining PR to unicodedomino_kernel_better_decode.def
• reorder inclusion; for now, include fixup_f4 unconditionally

mirabilos authored 5 years ago and mirabilos committed 5 years ago

according to https://en.wikibooks.org/wiki/LaTeX/Creating_Packages

mirabilos authored 6 years ago and mirabilos committed 6 years ago

use the same as for unicodedomino_kernel_fixup_f4_and_checkseq.def
since we can’t check for other added macros etc. yet… we could,
conceivably, use a version check, but given that this is not yet
released, I’m wary of that, and we submit the checkseq RSN

mirabilos authored 6 years ago and mirabilos committed 6 years ago

PR was accepted but does not use this helper macro,
and so does not define it either

mirabilos authored 6 years ago and mirabilos committed 6 years ago

further submissions are waiting on upstream testsuite bugfixes

mirabilos authored 6 years ago and mirabilos committed 6 years ago

rationale: in it, we can check for macro existence and load
other macros depending on it, while in the others we replace
upstream-provided or undefined macros sometimes

mirabilos authored 6 years ago and mirabilos committed 6 years ago

also, dual-LPPL on everything shared with upstream,
for easier inclusion (also consider copying back…)

mirabilos authored 6 years ago and mirabilos committed 6 years ago

alternatively, rewriting those to be expandable *might* help…

mirabilos authored 6 years ago and mirabilos committed 6 years ago

which we still need in production at some site