Commit e20947122d1224181e62416030bcd7a2f11749c6 - halibut

Disallow the zero Unicode character in all input. Halibut works internally with standard C-style null-terminated strings (or rather wide strings), so L'\0' appearing unexpectedly in the input can cause all kinds of havoc. It would be nice to redo all the string processing using (pointer, length) pairs and become robust against that, but I don't think it's realistic without a major rewrite. Zero characters have no actual use that I can see, so a simpler fix is to just outlaw them completely. This applies to a direct \0 appearing in the input file, and also to any sneaky attempts to enter one via \u0000. Simon Tatham 2 years ago

3 changed file(s) with 23 addition(s) and 1 deletion(s). Raw diff Collapse all Expand all

-0

error.c less more

82	82	do_error(NULL, "no data in input files");
83	83	}
84	84
	85	void err_zerochar(errorstate es, const filepos fpos)
	86	{
	87	es->fatal = true;
	88	do_error(fpos, "the Unicode zero character is not permitted in input");
	89	}
	90
85	91	void err_brokencodepara(errorstate es, const filepos fpos)
86	92	{
87	93	es->fatal = true;

-0

halibut.h less more

224	224	void err_cantopen(errorstate es, const char sp);
225	225	/* no data in input files */
226	226	void err_nodata(errorstate *es);
	227	/* unexpected zero character in input file */
	228	void err_zerochar(errorstate es, const filepos fpos);
227	229	/* line in codepara didn't begin `\c' */
228	230	void err_brokencodepara(errorstate es, const filepos fpos);
229	231	/* expected `}' after keyword */

+15

-1

input.c less more

169	169	NULL, 0);
170	170	assert(p == buf+1 && inlen == 0);
171	171
	172	for (int i = 0; i < in->nwc; i++) {
	173	if (in->wc[i] == 0) {
	174	/* The zero Unicode character is never legal */
	175	err_zerochar(in->es, pos);
	176	return EOF;
	177	}
	178	}
	179
172	180	in->wcpos = 0;
173	181	}
174	182	}
175	183
176		return in->wc[in->wcpos++];
	184	wchar_t wc = in->wc[in->wcpos++];
	185
	186	return wc;
177	187
178	188	} else
179	189	return EOF;

1526	1536	}
1527	1537	case c_u:
1528	1538	uchr = t.aux;
	1539	if (uchr == 0) {
	1540	err_zerochar(in->es, &t.pos);
	1541	break;
	1542	}
1529	1543	utext[0] = uchr; utext[1] = 0;
1530	1544	wd.type = style;
1531	1545	wd.breaks = false;