Lindenii Project Forge
git2d: Use extern inline for utf8_decode
# SPDX-License-Identifier: AGPL-3.0-only # SPDX-FileCopyrightText: Copyright (c) 2025 Runxi Yu <https://runxiyu.org> .PHONY: clean version.go man source.tar.gz CFLAGS = -Wall -Wextra -pedantic -std=c99 -D_GNU_SOURCE MAN_PAGES = lindenii-forge.5 lindenii-forge-hookc.1 lindenii-forge.1 lindenii-forge-mail.5 forge: source.tar.gz version.go hookc/*.c hookc/hookc man # TODO go build . man: $(MAN_PAGES:%=man/%.html) $(MAN_PAGES:%=man/%.txt) man/%.html: man/% mandoc -Thtml -O style=./mandoc.css $< > $@ man/%.txt: man/% utils/colb mandoc $< | ./utils/colb > $@ utils/colb: hookc/hookc:
git2d/git2d: git2d/main.o git2d/bare.o $(CC) $(CFLAGS) `pkg-config --cflags --libs libgit2` -lpthread -o git2d/git2d $<
git2d/git2d: git2d/main.o git2d/bare.o git2d/utf8.c $(CC) $(CFLAGS) `pkg-config --cflags --libs libgit2` -lpthread -o git2d/git2d $^
version.go: printf 'package main\n\nconst VERSION = "%s"\n' `git describe --tags --always --dirty` > $@ clean: $(RM) forge version.go vendor source.tar.gz: rm -f source.tar.gz go mod vendor git ls-files -z | xargs -0 tar -czf source.tar.gz vendor
#include "utf8.h" extern inline void *utf8_decode(void *buf, uint32_t *c, int *e);
/*- * SPDX-License-Identifier: Unlicense * SPDX-FileContributor: Chris Wellons <wellons@nullprogram.com> * * From: https://nullprogram.com/blog/2017/10/06/ */ #ifndef UTF8_H #define UTF8_H #include <stdint.h> /* * Decode the next character, C, from BUF, reporting errors in E. * * Since this is a branchless decoder, four bytes will be read from the * buffer regardless of the actual length of the next character. This * means the buffer _must_ have at least three bytes of zero padding * following the end of the data stream. * * Errors are reported in E, which will be non-zero if the parsed * character was somehow invalid: invalid byte sequence, non-canonical * encoding, or a surrogate half. * * The function returns a pointer to the next character. When an error * occurs, this pointer will be a guess that depends on the particular * error, but it will always advance at least one byte. */
static void *
inline void *
utf8_decode(void *buf, uint32_t *c, int *e) { static const char lengths[] = { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 2, 2, 3, 3, 4, 0 }; static const int masks[] = {0x00, 0x7f, 0x1f, 0x0f, 0x07}; static const uint32_t mins[] = {4194304, 0, 128, 2048, 65536}; static const int shiftc[] = {0, 18, 12, 6, 0}; static const int shifte[] = {0, 6, 4, 2, 0}; uint8_t *s = buf; int len = lengths[s[0] >> 3]; /* * Compute the pointer to the next character early so that the next * iteration can start working on the next character. Neither Clang * nor GCC figure out this reordering on their own. */ uint8_t *next = s + len + !len; /* * Assume a four-byte character and load four bytes. Unused bits are * shifted out. */ *c = (uint32_t)(s[0] & masks[len]) << 18; *c |= (uint32_t)(s[1] & 0x3f) << 12; *c |= (uint32_t)(s[2] & 0x3f) << 6; *c |= (uint32_t)(s[3] & 0x3f) << 0; *c >>= shiftc[len]; /* Accumulate the various error conditions. */ *e = (*c < mins[len]) << 6; /* non-canonical encoding */ *e |= ((*c >> 11) == 0x1b) << 7; /* surrogate half? */ *e |= (*c > 0x10FFFF) << 8; /* out of range? */ *e |= (s[1] & 0xc0) >> 2; *e |= (s[2] & 0xc0) >> 4; *e |= (s[3] ) >> 6; *e ^= 0x2a; /* top two bits of each tail byte correct? */ *e >>= shifte[len]; return next; } #endif