/src/libidn2/lib/context.c

Source
/* context.c - check contextual rule on label
   Copyright (C) 2011-2026 Simon Josefsson

   Libidn2 is free software: you can redistribute it and/or modify it
   under the terms of either:

     * the GNU Lesser General Public License as published by the Free
       Software Foundation; either version 3 of the License, or (at
       your option) any later version.

   or

     * the GNU General Public License as published by the Free
       Software Foundation; either version 2 of the License, or (at
       your option) any later version.

   or both in parallel, as here.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received copies of the GNU General Public License and
   the GNU Lesser General Public License along with this program.  If
   not, see <http://www.gnu.org/licenses/>.
*/

#include <config.h>

#include "idn2.h"
#include "tables.h"
#include <unictype.h>   /* uc_combining_class, UC_CCC_VR */
#include "context.h"

int
_idn2_contextj_rule (const uint32_t *label, size_t llen, size_t pos)
{
  uint32_t cp;

  if (llen == 0)
    return IDN2_OK;

  cp = label[pos];

  if (!_idn2_contextj_p (cp))
    return IDN2_OK;

  switch (cp)
    {
    case 0x200C:    /* ZERO WIDTH NON-JOINER */
      if (pos > 0)
  {
    /* If Canonical_Combining_Class(Before(cp)) .eq.  Virama Then True; */
    uint32_t before_cp = label[pos - 1];
    int cc = uc_combining_class (before_cp);
    if (cc == UC_CCC_VR)
      return IDN2_OK;
  }

      /* See http://permalink.gmane.org/gmane.ietf.idnabis/6980 for
         clarified rule. */

      if (pos == 0 || pos == llen - 1)
  return IDN2_CONTEXTJ;

      {
  int jt;
  size_t tmp;

  /* Search backwards. */
  for (tmp = pos - 1;; tmp--)
    {
      jt = uc_joining_type (label[tmp]);
      if (jt == UC_JOINING_TYPE_L || jt == UC_JOINING_TYPE_D)
        break;
      if (tmp == 0)
        return IDN2_CONTEXTJ;
      if (jt == UC_JOINING_TYPE_T)
        continue;
      return IDN2_CONTEXTJ;
    }

  /* Search forward. */
  for (tmp = pos + 1; tmp < llen; tmp++)
    {
      jt = uc_joining_type (label[tmp]);
      if (jt == UC_JOINING_TYPE_R || jt == UC_JOINING_TYPE_D)
        break;
      if (tmp == llen - 1)
        return IDN2_CONTEXTJ;
      if (jt == UC_JOINING_TYPE_T)
        continue;
      return IDN2_CONTEXTJ;
    }
      }

      return IDN2_OK;
      break;

    case 0x200D:    /* ZERO WIDTH JOINER */
      if (pos > 0)
  {
    uint32_t before_cp = label[pos - 1];
    int cc = uc_combining_class (before_cp);
    if (cc == UC_CCC_VR)
      return IDN2_OK;
  }
      return IDN2_CONTEXTJ;
    }

  return IDN2_CONTEXTJ_NO_RULE;
}

static const char *
_uc_script_name (ucs4_t uc)
{
  const uc_script_t *ucs = uc_script (uc);

  if (!ucs)
    return "";

  return ucs->name;
}

int
_idn2_contexto_rule (const uint32_t *label, size_t llen, size_t pos)
{
  uint32_t cp = label[pos];

  if (!_idn2_contexto_p (cp))
    return IDN2_OK;

  switch (cp)
    {
    case 0x00B7:
      /* MIDDLE DOT */
      if (llen < 3)
  return IDN2_CONTEXTO;
      if (pos == 0 || pos == llen - 1)
  return IDN2_CONTEXTO;
      if (label[pos - 1] == 0x006C && label[pos + 1] == 0x006C)
  return IDN2_OK;
      return IDN2_CONTEXTO;
      break;

    case 0x0375:
      /* GREEK LOWER NUMERAL SIGN (KERAIA) */
      if (pos == llen - 1)
  return IDN2_CONTEXTO;
      if (strcmp (_uc_script_name (label[pos + 1]), "Greek") == 0)
  return IDN2_OK;
      return IDN2_CONTEXTO;
      break;

    case 0x05F3:
      /* HEBREW PUNCTUATION GERESH */
    case 0x05F4:
      /* HEBREW PUNCTUATION GERSHAYIM */
      if (pos == 0)
  return IDN2_CONTEXTO;
      if (strcmp (_uc_script_name (label[pos - 1]), "Hebrew") == 0)
  return IDN2_OK;
      return IDN2_CONTEXTO;
      break;

    case 0x0660:
    case 0x0661:
    case 0x0662:
    case 0x0663:
    case 0x0664:
    case 0x0665:
    case 0x0666:
    case 0x0667:
    case 0x0668:
    case 0x0669:
      {
  /* ARABIC-INDIC DIGITS */
  size_t i;
  for (i = 0; i < llen; i++)
    if (label[i] >= 0x6F0 && label[i] <= 0x06F9)
      return IDN2_CONTEXTO;
  return IDN2_OK;
  break;
      }

    case 0x06F0:
    case 0x06F1:
    case 0x06F2:
    case 0x06F3:
    case 0x06F4:
    case 0x06F5:
    case 0x06F6:
    case 0x06F7:
    case 0x06F8:
    case 0x06F9:
      {
  /* EXTENDED ARABIC-INDIC DIGITS */
  size_t i;
  for (i = 0; i < llen; i++)
    if (label[i] >= 0x660 && label[i] <= 0x0669)
      return IDN2_CONTEXTO;
  return IDN2_OK;
  break;
      }
    case 0x30FB:
      {
  /* KATAKANA MIDDLE DOT */
  size_t i;
  bool script_ok = false;

  for (i = 0; !script_ok && i < llen; i++)
    if (strcmp (_uc_script_name (label[i]), "Hiragana") == 0
        || strcmp (_uc_script_name (label[i]), "Katakana") == 0
        || strcmp (_uc_script_name (label[i]), "Han") == 0)
      script_ok = true;

  if (script_ok)
    return IDN2_OK;
  return IDN2_CONTEXTO;
  break;
      }
    }

  return IDN2_CONTEXTO_NO_RULE;
}

bool
_idn2_contexto_with_rule (uint32_t cp)
{
  switch (cp)
    {
    case 0x00B7:
      /* MIDDLE DOT */
    case 0x0375:
      /* GREEK LOWER NUMERAL SIGN (KERAIA) */
    case 0x05F3:
      /* HEBREW PUNCTUATION GERESH */
    case 0x05F4:
      /* HEBREW PUNCTUATION GERSHAYIM */
    case 0x0660:
    case 0x0661:
    case 0x0662:
    case 0x0663:
    case 0x0664:
    case 0x0665:
    case 0x0666:
    case 0x0667:
    case 0x0668:
    case 0x0669:
      /* ARABIC-INDIC DIGITS */
    case 0x06F0:
    case 0x06F1:
    case 0x06F2:
    case 0x06F3:
    case 0x06F4:
    case 0x06F5:
    case 0x06F6:
    case 0x06F7:
    case 0x06F8:
    case 0x06F9:
      /* EXTENDED ARABIC-INDIC DIGITS */
    case 0x30FB:
      /* KATAKANA MIDDLE DOT */
      return true;
      break;
    }

  return false;
}

Coverage Report

Created: 2026-08-13 07:05

Line	Count	Source
1		/* context.c - check contextual rule on label
2		Copyright (C) 2011-2026 Simon Josefsson
3
4		Libidn2 is free software: you can redistribute it and/or modify it
5		under the terms of either:
6
7		* the GNU Lesser General Public License as published by the Free
8		Software Foundation; either version 3 of the License, or (at
9		your option) any later version.
10
11		or
12
13		* the GNU General Public License as published by the Free
14		Software Foundation; either version 2 of the License, or (at
15		your option) any later version.
16
17		or both in parallel, as here.
18
19		This program is distributed in the hope that it will be useful,
20		but WITHOUT ANY WARRANTY; without even the implied warranty of
21		MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
22		GNU General Public License for more details.
23
24		You should have received copies of the GNU General Public License and
25		the GNU Lesser General Public License along with this program. If
26		not, see <http://www.gnu.org/licenses/>.
27		*/
28
29		#include <config.h>
30
31		#include "idn2.h"
32		#include "tables.h"
33		#include <unictype.h> /* uc_combining_class, UC_CCC_VR */
34		#include "context.h"
35
36		int
37		_idn2_contextj_rule (const uint32_t *label, size_t llen, size_t pos)
38	7.49k	{
39	7.49k	uint32_t cp;
40
41	7.49k	if (llen == 0)
42	0	return IDN2_OK;
43
44	7.49k	cp = label[pos];
45
46	7.49k	if (!_idn2_contextj_p (cp))
47	7.22k	return IDN2_OK;
48
49	270	switch (cp)
50	270	{
51	229	case 0x200C: /* ZERO WIDTH NON-JOINER */
52	229	if (pos > 0)
53	228	{
54		/* If Canonical_Combining_Class(Before(cp)) .eq. Virama Then True; */
55	228	uint32_t before_cp = label[pos - 1];
56	228	int cc = uc_combining_class (before_cp);
57	228	if (cc == UC_CCC_VR)
58	35	return IDN2_OK;
59	228	}
60
61		/* See http://permalink.gmane.org/gmane.ietf.idnabis/6980 for
62		clarified rule. */
63
64	194	if (pos == 0 \|\| pos == llen - 1)
65	15	return IDN2_CONTEXTJ;
66
67	179	{
68	179	int jt;
69	179	size_t tmp;
70
71		/* Search backwards. */
72	183	for (tmp = pos - 1;; tmp--)
73	362	{
74	362	jt = uc_joining_type (label[tmp]);
75	362	if (jt == UC_JOINING_TYPE_L \|\| jt == UC_JOINING_TYPE_D)
76	111	break;
77	251	if (tmp == 0)
78	55	return IDN2_CONTEXTJ;
79	196	if (jt == UC_JOINING_TYPE_T)
80	183	continue;
81	13	return IDN2_CONTEXTJ;
82	196	}
83
84		/* Search forward. */
85	261	for (tmp = pos + 1; tmp < llen; tmp++)
86	261	{
87	261	jt = uc_joining_type (label[tmp]);
88	261	if (jt == UC_JOINING_TYPE_R \|\| jt == UC_JOINING_TYPE_D)
89	90	break;
90	171	if (tmp == llen - 1)
91	17	return IDN2_CONTEXTJ;
92	154	if (jt == UC_JOINING_TYPE_T)
93	150	continue;
94	4	return IDN2_CONTEXTJ;
95	154	}
96	111	}
97
98	90	return IDN2_OK;
99	0	break;
100
101	41	case 0x200D: /* ZERO WIDTH JOINER */
102	41	if (pos > 0)
103	40	{
104	40	uint32_t before_cp = label[pos - 1];
105	40	int cc = uc_combining_class (before_cp);
106	40	if (cc == UC_CCC_VR)
107	34	return IDN2_OK;
108	40	}
109	7	return IDN2_CONTEXTJ;
110	270	}
111
112	0	return IDN2_CONTEXTJ_NO_RULE;
113	270	}
114
115		static const char *
116		_uc_script_name (ucs4_t uc)
117	18.2k	{
118	18.2k	const uc_script_t *ucs = uc_script (uc);
119
120	18.2k	if (!ucs)
121	1.94k	return "";
122
123	16.3k	return ucs->name;
124	18.2k	}
125
126		int
127		_idn2_contexto_rule (const uint32_t *label, size_t llen, size_t pos)
128	6.63k	{
129	6.63k	uint32_t cp = label[pos];
130
131	6.63k	if (!_idn2_contexto_p (cp))
132	3.98k	return IDN2_OK;
133
134	2.65k	switch (cp)
135	2.65k	{
136	101	case 0x00B7:
137		/* MIDDLE DOT */
138	101	if (llen < 3)
139	1	return IDN2_CONTEXTO;
140	100	if (pos == 0 \|\| pos == llen - 1)
141	10	return IDN2_CONTEXTO;
142	90	if (label[pos - 1] == 0x006C && label[pos + 1] == 0x006C)
143	46	return IDN2_OK;
144	44	return IDN2_CONTEXTO;
145	0	break;
146
147	91	case 0x0375:
148		/* GREEK LOWER NUMERAL SIGN (KERAIA) */
149	91	if (pos == llen - 1)
150	8	return IDN2_CONTEXTO;
151	83	if (strcmp (_uc_script_name (label[pos + 1]), "Greek") == 0)
152	67	return IDN2_OK;
153	16	return IDN2_CONTEXTO;
154	0	break;
155
156	138	case 0x05F3:
157		/* HEBREW PUNCTUATION GERESH */
158	214	case 0x05F4:
159		/* HEBREW PUNCTUATION GERSHAYIM */
160	214	if (pos == 0)
161	2	return IDN2_CONTEXTO;
162	212	if (strcmp (_uc_script_name (label[pos - 1]), "Hebrew") == 0)
163	200	return IDN2_OK;
164	12	return IDN2_CONTEXTO;
165	0	break;
166
167	81	case 0x0660:
168	192	case 0x0661:
169	286	case 0x0662:
170	395	case 0x0663:
171	487	case 0x0664:
172	583	case 0x0665:
173	661	case 0x0666:
174	742	case 0x0667:
175	823	case 0x0668:
176	900	case 0x0669:
177	900	{
178		/* ARABIC-INDIC DIGITS */
179	900	size_t i;
180	20.2k	for (i = 0; i < llen; i++)
181	19.3k	if (label[i] >= 0x6F0 && label[i] <= 0x06F9)
182	5	return IDN2_CONTEXTO;
183	895	return IDN2_OK;
184	0	break;
185	900	}
186
187	134	case 0x06F0:
188	253	case 0x06F1:
189	358	case 0x06F2:
190	487	case 0x06F3:
191	576	case 0x06F4:
192	647	case 0x06F5:
193	724	case 0x06F6:
194	803	case 0x06F7:
195	874	case 0x06F8:
196	946	case 0x06F9:
197	946	{
198		/* EXTENDED ARABIC-INDIC DIGITS */
199	946	size_t i;
200	19.5k	for (i = 0; i < llen; i++)
201	18.6k	if (label[i] >= 0x660 && label[i] <= 0x0669)
202	4	return IDN2_CONTEXTO;
203	942	return IDN2_OK;
204	0	break;
205	946	}
206	402	case 0x30FB:
207	402	{
208		/* KATAKANA MIDDLE DOT */
209	402	size_t i;
210	402	bool script_ok = false;
211
212	6.50k	for (i = 0; !script_ok && i < llen; i++)
213	6.10k	if (strcmp (_uc_script_name (label[i]), "Hiragana") == 0
214	6.03k	\|\| strcmp (_uc_script_name (label[i]), "Katakana") == 0
215	5.86k	\|\| strcmp (_uc_script_name (label[i]), "Han") == 0)
216	358	script_ok = true;
217
218	402	if (script_ok)
219	358	return IDN2_OK;
220	44	return IDN2_CONTEXTO;
221	0	break;
222	402	}
223	2.65k	}
224
225	0	return IDN2_CONTEXTO_NO_RULE;
226	2.65k	}
227
228		bool
229		_idn2_contexto_with_rule (uint32_t cp)
230	0	{
231	0	switch (cp)
232	0	{
233	0	case 0x00B7:
234		/* MIDDLE DOT */
235	0	case 0x0375:
236		/* GREEK LOWER NUMERAL SIGN (KERAIA) */
237	0	case 0x05F3:
238		/* HEBREW PUNCTUATION GERESH */
239	0	case 0x05F4:
240		/* HEBREW PUNCTUATION GERSHAYIM */
241	0	case 0x0660:
242	0	case 0x0661:
243	0	case 0x0662:
244	0	case 0x0663:
245	0	case 0x0664:
246	0	case 0x0665:
247	0	case 0x0666:
248	0	case 0x0667:
249	0	case 0x0668:
250	0	case 0x0669:
251		/* ARABIC-INDIC DIGITS */
252	0	case 0x06F0:
253	0	case 0x06F1:
254	0	case 0x06F2:
255	0	case 0x06F3:
256	0	case 0x06F4:
257	0	case 0x06F5:
258	0	case 0x06F6:
259	0	case 0x06F7:
260	0	case 0x06F8:
261	0	case 0x06F9:
262		/* EXTENDED ARABIC-INDIC DIGITS */
263	0	case 0x30FB:
264		/* KATAKANA MIDDLE DOT */
265	0	return true;
266	0	break;
267	0	}
268
269	0	return false;
270	0	}