bump hunspell to 1.3.4
Change-Id: If7c4868c4296d1ca0e485dc06fdf3472e2e8653f
Reviewed-on: https://gerrit.libreoffice.org/23863
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
diff --git a/download.lst b/download.lst
index 23ec620..21c95b1 100644
--- a/download.lst
+++ b/download.lst
@@ -61,7 +61,7 @@ export GRAPHITE_TARBALL := 4311dd9ace498b57c85f611e0670df64-graphite2-minimal-1.
export HARFBUZZ_MD5SUM := 0e27e531f4c4acff601ebff0957755c2
export HARFBUZZ_TARBALL := harfbuzz-0.9.40.tar.bz2
export HSQLDB_TARBALL := 17410483b5b5f267aa18b7e00b65e6e0-hsqldb_1_8_0.zip
export HUNSPELL_TARBALL := 4967da60b23413604c9e563beacc63b4-hunspell-1.3.3.tar.gz
export HUNSPELL_TARBALL := 423cff69e68c87ac11e4aa8462951954-hunspell-1.3.4.tar.gz
export HYPHEN_TARBALL := 5ade6ae2a99bc1e9e57031ca88d36dad-hyphen-2.8.8.tar.gz
export ICU_TARBALL := c4a2d71ff56aec5ebfab2a3f059be99d-icu4c-56_1-src.tgz
export JFREEREPORT_FLOW_ENGINE_TARBALL := ba2930200c9f019c2d93a8c88c651a0f-flow-engine-0.9.4.zip
diff --git a/external/hunspell/UnpackedTarball_hunspell.mk b/external/hunspell/UnpackedTarball_hunspell.mk
index 8f14062..e70e33f 100644
--- a/external/hunspell/UnpackedTarball_hunspell.mk
+++ b/external/hunspell/UnpackedTarball_hunspell.mk
@@ -13,15 +13,7 @@ $(eval $(call gb_UnpackedTarball_set_tarball,hunspell,$(HUNSPELL_TARBALL)))
$(eval $(call gb_UnpackedTarball_add_patches,hunspell,\
external/hunspell/hunspell-solaris.patch \
external/hunspell/hunspell-1.3.2-overflow.patch \
external/hunspell/hunspell-android.patch \
external/hunspell/hunspell-1.3.2-nullptr.patch \
external/hunspell/hunspell-1.3.2-literal.patch \
external/hunspell/hunspell-fdo48017-wfopen.patch \
external/hunspell/hunspell-morph-overflow.patch \
external/hunspell/ubsan.patch.0 \
external/hunspell/hunspell-1.3.3-rhbz1261421.patch \
external/hunspell/hunspell-tdf95024-compound.patch \
external/hunspell/hunspell-windows.patch \
))
ifeq ($(COM),MSC)
diff --git a/external/hunspell/hunspell-1.3.2-literal.patch b/external/hunspell/hunspell-1.3.2-literal.patch
deleted file mode 100644
index b0a8fb9..0000000
--- a/external/hunspell/hunspell-1.3.2-literal.patch
+++ /dev/null
@@ -1,11 +0,0 @@
--- misc/hunspell-1.3.2/src/tools/hunspell.cxx 2012-08-13 12:09:06.107017665 +0200
+++ misc/build/hunspell-1.3.2/src/tools/hunspell.cxx 2012-08-13 12:14:12.233500532 +0200
@@ -20,7 +20,7 @@
#define HUNSPELL_VERSION VERSION
#define INPUTLEN 50
-#define HUNSPELL_PIPE_HEADING "@(#) International Ispell Version 3.2.06 (but really Hunspell "VERSION")\n"
+#define HUNSPELL_PIPE_HEADING "@(#) International Ispell Version 3.2.06 (but really Hunspell " VERSION ")\n"
#define HUNSPELL_HEADING "Hunspell "
#define ODF_EXT "odt|ott|odp|otp|odg|otg|ods|ots"
#define ENTITY_APOS "'"
diff --git a/external/hunspell/hunspell-1.3.2-nullptr.patch b/external/hunspell/hunspell-1.3.2-nullptr.patch
deleted file mode 100644
index 624e6ce..0000000
--- a/external/hunspell/hunspell-1.3.2-nullptr.patch
+++ /dev/null
@@ -1,20 +0,0 @@
--- misc/hunspell-1.3.2/src/hunspell/affentry.hxx 2010-04-15 13:22:08.000000000 +0200
+++ misc/build/hunspell-1.3.2/src/hunspell/affentry.hxx 2012-08-13 12:08:48.568996730 +0200
@@ -27,7 +27,7 @@
struct hentry * checkword(const char * word, int len, char in_compound,
const FLAG needflag = FLAG_NULL);
- struct hentry * check_twosfx(const char * word, int len, char in_compound, const FLAG needflag = NULL);
+ struct hentry * check_twosfx(const char * word, int len, char in_compound, const FLAG needflag = FLAG_NULL);
char * check_morph(const char * word, int len, char in_compound,
const FLAG needflag = FLAG_NULL);
@@ -90,7 +90,7 @@
// const FLAG cclass = FLAG_NULL, const FLAG needflag = FLAG_NULL, char in_compound=IN_CPD_NOT);
const FLAG cclass = FLAG_NULL, const FLAG needflag = FLAG_NULL, const FLAG badflag = 0);
- struct hentry * check_twosfx(const char * word, int len, int optflags, PfxEntry* ppfx, const FLAG needflag = NULL);
+ struct hentry * check_twosfx(const char * word, int len, int optflags, PfxEntry* ppfx, const FLAG needflag = FLAG_NULL);
char * check_twosfx_morph(const char * word, int len, int optflags,
PfxEntry* ppfx, const FLAG needflag = FLAG_NULL);
diff --git a/external/hunspell/hunspell-1.3.2-overflow.patch b/external/hunspell/hunspell-1.3.2-overflow.patch
deleted file mode 100644
index 8500b483..0000000
--- a/external/hunspell/hunspell-1.3.2-overflow.patch
+++ /dev/null
@@ -1,91 +0,0 @@
--- misc/hunspell-1.3.2/src/hunspell/affixmgr.cxx 2010-02-27 12:59:53.000000000 +0100
+++ misc/build/hunspell-1.3.2/src/hunspell/affixmgr.cxx 2011-05-18 16:29:45.919141893 +0200
@@ -6,6 +6,8 @@
#include <stdio.h>
#include <ctype.h>
+#include <limits>
+
#include <vector>
#include "affixmgr.hxx"
@@ -4000,7 +4002,10 @@
case 3: {
np++;
numents = atoi(piece);
- if (numents == 0) {
+ if ((numents <= 0) ||
+ ((::std::numeric_limits<size_t>::max()
+ / sizeof(struct affentry)) < numents))
+ {
char * err = pHMgr->encode_flag(aflag);
if (err) {
HUNSPELL_WARNING(stderr, "error: line %d: bad entry number\n",
--- misc/hunspell-1.3.2/src/tools/munch.c 2010-02-27 21:49:49.000000000 +0100
+++ misc/build/hunspell-1.3.2/src/tools/munch.c 2011-05-18 15:53:53.427072106 +0200
@@ -4,6 +4,7 @@
#include <string.h>
#include <unistd.h>
#include <stdlib.h>
+#include <stdint.h>
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
@@ -233,10 +233,19 @@
case 1: { achar = *piece; break; }
case 2: { if (*piece == 'Y') ff = XPRODUCT; break; }
case 3: { numents = atoi(piece);
- ptr = malloc(numents * sizeof(struct affent));
- ptr->achar = achar;
- ptr->xpflg = ff;
- fprintf(stderr,"parsing %c entries %d\n",achar,numents);
+ if ((numents < 0) ||
+ ((SIZE_MAX/sizeof(struct affent)) < numents))
+ {
+ fprintf(stderr,
+ "Error: too many entries: %d\n", numents);
+ numents = 0;
+ } else {
+ ptr = malloc(numents * sizeof(struct affent));
+ ptr->achar = achar;
+ ptr->xpflg = ff;
+ fprintf(stderr,"parsing %c entries %d\n",
+ achar,numents);
+ }
break;
}
default: break;
--- misc/hunspell-1.3.2/src/tools/unmunch.c 2010-02-23 15:53:29.000000000 +0100
+++ misc/build/hunspell-1.3.2/src/tools/unmunch.c 2011-05-18 20:53:43.843599726 +0200
@@ -6,6 +6,7 @@
#include <string.h>
#include <unistd.h>
#include <stdlib.h>
+#include <stdint.h>
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
@@ -158,10 +159,19 @@
case 1: { achar = *piece; break; }
case 2: { if (*piece == 'Y') ff = XPRODUCT; break; }
case 3: { numents = atoi(piece);
- ptr = malloc(numents * sizeof(struct affent));
- ptr->achar = achar;
- ptr->xpflg = ff;
- fprintf(stderr,"parsing %c entries %d\n",achar,numents);
+ if ((numents < 0) ||
+ ((SIZE_MAX/sizeof(struct affent)) < numents))
+ {
+ fprintf(stderr,
+ "Error: too many entries: %d\n", numents);
+ numents = 0;
+ } else {
+ ptr = malloc(numents * sizeof(struct affent));
+ ptr->achar = achar;
+ ptr->xpflg = ff;
+ fprintf(stderr,"parsing %c entries %d\n",
+ achar,numents);
+ }
break;
}
default: break;
diff --git a/external/hunspell/hunspell-1.3.3-rhbz1261421.patch b/external/hunspell/hunspell-1.3.3-rhbz1261421.patch
deleted file mode 100644
index 4354dd0..0000000
--- a/external/hunspell/hunspell-1.3.3-rhbz1261421.patch
+++ /dev/null
@@ -1,191 +0,0 @@
From 97e079a23d459aeb6e64435350d7710c90dbca85 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Caol=C3=A1n=20McNamara?= <caolanm@redhat.com>
Date: Fri, 11 Sep 2015 13:28:52 +0100
Subject: [PATCH] Resolves: rhbz#1261421 crash on mashing hangul korean
keyboard
---
src/hunspell/hunspell.cxx | 69 +++++++++++++++++++++++++++++++++++------------
src/hunspell/hunspell.hxx | 4 ++-
src/hunspell/replist.cxx | 18 ++++++++++---
src/hunspell/replist.hxx | 2 ++
src/tools/hunspell.cxx | 2 +-
6 files changed, 78 insertions(+), 24 deletions(-)
diff --git a/src/hunspell/hunspell.cxx b/src/hunspell/hunspell.cxx
index 7fae54b..d8ef357 100644
--- misc/hunspell-1.3.3/src/hunspell/hunspell.cxx
+++ misc/build/hunspell-1.3.3/src/hunspell/hunspell.cxx
@@ -12,6 +12,7 @@
#endif
#include "csutil.hxx"
+#include <limits>
#include <string>
Hunspell::Hunspell(const char * affpath, const char * dpath, const char * key)
@@ -349,8 +350,13 @@ int Hunspell::spell(const char * word, int * info, char ** root)
// input conversion
RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL;
- if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
- else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
+ int convstatus = rl ? rl->conv(word, wspace, MAXWORDUTF8LEN) : 0;
+ if (convstatus < 0)
+ return 0;
+ else if (convstatus > 0)
+ wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
+ else
+ wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
if (wl == 0 || maxdic == 0) return 1;
if (root) *root = NULL;
@@ -702,8 +708,13 @@ int Hunspell::suggest(char*** slst, const char * word)
// input conversion
RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL;
- if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
- else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
+ int convstatus = rl ? rl->conv(word, wspace, MAXWORDUTF8LEN) : 0;
+ if (convstatus < 0)
+ return 0;
+ else if (convstatus > 0)
+ wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
+ else
+ wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
if (wl == 0) return 0;
int ns = 0;
@@ -1020,7 +1031,7 @@ int Hunspell::suggest(char*** slst, const char * word)
// output conversion
rl = (pAMgr) ? pAMgr->get_oconvtable() : NULL;
for (int j = 0; rl && j < ns; j++) {
- if (rl->conv((*slst)[j], wspace)) {
+ if (rl->conv((*slst)[j], wspace, MAXWORDUTF8LEN) > 0) {
free((*slst)[j]);
(*slst)[j] = mystrdup(wspace);
}
@@ -1395,8 +1406,13 @@ int Hunspell::analyze(char*** slst, const char * word)
// input conversion
RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL;
- if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
- else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
+ int convstatus = rl ? rl->conv(word, wspace, MAXWORDUTF8LEN) : 0;
+ if (convstatus < 0)
+ return 0;
+ else if (convstatus > 0)
+ wl = cleanword2(cw, wspace, unicw, &nc, &captype, &abbv);
+ else
+ wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv);
if (wl == 0) {
if (abbv) {
@@ -1684,12 +1700,16 @@ int Hunspell::get_langnum() const
return langnum;
}
-int Hunspell::input_conv(const char * word, char * dest)
+int Hunspell::input_conv(const char * word, char * dest, size_t destsize)
{
RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL;
- return (rl && rl->conv(word, dest));
+ return (rl && rl->conv(word, dest, destsize) > 0);
}
+int Hunspell::input_conv(const char * word, char * dest)
+{
+ return input_conv(word, dest, std::numeric_limits<std::size_t>::max());
+}
// return the beginning of the element (attr == NULL) or the attribute
const char * Hunspell::get_xml_pos(const char * s, const char * attr)
diff --git a/src/hunspell/hunspell.hxx b/src/hunspell/hunspell.hxx
index e62f0dd..0b5ad2e 100644
--- misc/hunspell-1.3.3/src/hunspell/hunspell.hxx
+++ misc/build/hunspell-1.3.3/src/hunspell/hunspell.hxx
@@ -226,7 +226,9 @@ public:
/* need for putdic */
int input_conv(const char * word, char * dest);
-
+ // ^^-deprecated, use this-vv"
+ int input_conv(const char * word, char * dest, size_t destsize);
+
/* experimental and deprecated functions */
#ifdef HUNSPELL_EXPERIMENTAL
diff --git a/src/hunspell/replist.cxx b/src/hunspell/replist.cxx
index b9b1255..bac3e06 100644
--- misc/hunspell-1.3.3/src/hunspell/replist.cxx
+++ misc/build/hunspell-1.3.3/src/hunspell/replist.cxx
@@ -74,6 +74,7 @@
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
+#include <limits>
#include "replist.hxx"
#include "csutil.hxx"
@@ -139,19 +140,30 @@ int RepList::add(char * pat1, char * pat2) {
return 0;
}
-int RepList::conv(const char * word, char * dest) {
+int RepList::conv(const char * word, char * dest, size_t destsize) {
int stl = 0;
int change = 0;
for (size_t i = 0; i < strlen(word); i++) {
int n = near(word + i);
int l = match(word + i, n);
if (l) {
+ size_t replen = strlen(dat[n]->pattern2);
+ if (stl+replen >= destsize)
+ return -1;
strcpy(dest + stl, dat[n]->pattern2);
- stl += strlen(dat[n]->pattern2);
+ stl += replen;
i += l - 1;
change = 1;
- } else dest[stl++] = word[i];
+ } else {
+ if (stl+1 >= destsize)
+ return -1;
+ dest[stl++] = word[i];
+ }
}
dest[stl] = '\0';
return change;
}
+
+int RepList::conv(const char * word, char * dest) {
+ return conv(word, dest, std::numeric_limits<std::size_t>::max());
+}
diff --git a/src/hunspell/replist.hxx b/src/hunspell/replist.hxx
index 1e3d6e4..e418298 100644
--- misc/hunspell-1.3.3/src/hunspell/replist.hxx
+++ misc/build/hunspell-1.3.3/src/hunspell/replist.hxx
@@ -99,5 +99,7 @@ public:
int near(const char * word);
int match(const char * word, int n);
int conv(const char * word, char * dest);
+ // ^^-deprecated, use this-vv"
+ int conv(const char * word, char * dest, size_t destsize);
};
#endif
diff --git a/src/tools/hunspell.cxx b/src/tools/hunspell.cxx
index 6124ac4..1b50fe1 100644
--- misc/hunspell-1.3.3/src/tools/hunspell.cxx
+++ misc/build/hunspell-1.3.3/src/tools/hunspell.cxx
@@ -524,7 +524,7 @@ int putdic(char * word, Hunspell * pMS)
word = chenc(word, ui_enc, dic_enc[0]);
- if(pMS->input_conv(word, buf)) word = buf;
+ if(pMS->input_conv(word, buf, MAXLNLEN)) word = buf;
int ret;
--
2.4.0
diff --git a/external/hunspell/hunspell-android.patch b/external/hunspell/hunspell-android.patch
deleted file mode 100644
index 6653d63..0000000
--- a/external/hunspell/hunspell-android.patch
+++ /dev/null
@@ -1,20 +0,0 @@
--- misc/hunspell-1.3.2/config.sub 2010-02-23 10:08:48.000000000 +0100
+++ misc/build/hunspell-1.3.2/config.sub 2011-09-26 23:31:02.000000000 +0200
@@ -118,7 +118,7 @@
# Here we must recognize all the valid KERNEL-OS combinations.
maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'`
case $maybe_os in
- nto-qnx* | linux-gnu* | linux-dietlibc | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | \
+ nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | \
kfreebsd*-gnu* | knetbsd*-gnu* | netbsd*-gnu* | storm-chaos* | os2-emx* | rtmk-nova*)
os=-$maybe_os
basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`
@@ -1161,7 +1161,7 @@
| -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \
| -chorusos* | -chorusrdb* \
| -cygwin* | -pe* | -psos* | -moss* | -proelf* | -rtems* \
- | -mingw32* | -linux-gnu* | -linux-uclibc* | -uxpv* | -beos* | -mpeix* | -udk* \
+ | -mingw32* | -linux-gnu* | -linux-androideabi* | -linux-uclibc* | -uxpv* | -beos* | -mpeix* | -udk* \
| -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \
| -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \
| -storm-chaos* | -tops10* | -tenex* | -tops20* | -its* \
diff --git a/external/hunspell/hunspell-fdo48017-wfopen.patch b/external/hunspell/hunspell-fdo48017-wfopen.patch
deleted file mode 100644
index 14f1db72..0000000
--- a/external/hunspell/hunspell-fdo48017-wfopen.patch
+++ /dev/null
@@ -1,19 +0,0 @@
diff -ru hunspell/src/hunspell/csutil.cxx build/hunspell/src/hunspell/csutil.cxx
--- hunspell/src/hunspell/csutil.cxx 2011-02-02 11:35:43.000000000 +0100
+++ build/hunspell/src/hunspell/csutil.cxx 2014-04-24 19:42:01.373285409 +0200
@@ -57,9 +57,14 @@
if (strncmp(path, WIN32_LONG_PATH_PREFIX, 4) == 0) {
int len = MultiByteToWideChar(CP_UTF8, 0, path, -1, NULL, 0);
wchar_t *buff = (wchar_t *) malloc(len * sizeof(wchar_t));
+ wchar_t *buff2 = (wchar_t *) malloc(len * sizeof(wchar_t));
+ FILE * f = NULL;
MultiByteToWideChar(CP_UTF8, 0, path, -1, buff, len);
- FILE * f = _wfopen(buff, (strcmp(mode, "r") == 0) ? L"r" : L"rb");
+ if (_wfullpath( buff2, buff, len ) != NULL) {
+ f = _wfopen(buff2, (strcmp(mode, "r") == 0) ? L"r" : L"rb");
+ }
free(buff);
+ free(buff2);
return f;
}
#endif
diff --git a/external/hunspell/hunspell-morph-overflow.patch b/external/hunspell/hunspell-morph-overflow.patch
deleted file mode 100644
index fe7c4f7..0000000
--- a/external/hunspell/hunspell-morph-overflow.patch
+++ /dev/null
@@ -1,30 +0,0 @@
--- hunspell/src/hunspell/affixmgr.cxx 2014-09-24 16:11:10.750421303 +0200
+++ build/hunspell/src/hunspell/affixmgr.cxx 2014-09-26 15:25:09.448688908 +0200
@@ -2400,8 +2400,10 @@
}
mystrcat(*result, presult, MAXLNLEN);
if (m || (*m != '\0')) {
- sprintf(*result + strlen(*result), "%c%s%s%s", MSEP_FLD,
+ char m2[MAXLNLEN];
+ sprintf(m2, "%c%s%s%s", MSEP_FLD,
MORPH_PART, word + i, line_uniq_app(&m, MSEP_REC));
+ mystrcat(*result, m2, MAXLNLEN);
}
if (m) free(m);
mystrcat(*result, "\n", MAXLNLEN);
@@ -2481,11 +2483,13 @@
}
mystrcat(*result, presult, MAXLNLEN);
if (m && (*m != '\0')) {
- sprintf(*result + strlen(*result), "%c%s%s%s", MSEP_FLD,
+ char m2[MAXLNLEN];
+ sprintf(m2, "%c%s%s%s", MSEP_FLD,
MORPH_PART, word + i, line_uniq_app(&m, MSEP_REC));
+ mystrcat(*result, m2, MAXLNLEN);
}
if (m) free(m);
- sprintf(*result + strlen(*result), "%c", MSEP_REC);
+ if (strlen(*result) + 1 < MAXLNLEN) sprintf(*result + strlen(*result), "%c", MSEP_REC);
ok = 1;
}
diff --git a/external/hunspell/hunspell-tdf95024-compound.patch b/external/hunspell/hunspell-tdf95024-compound.patch
deleted file mode 100644
index 133cf4a..0000000
--- a/external/hunspell/hunspell-tdf95024-compound.patch
+++ /dev/null
@@ -1,108 +0,0 @@
From 42807f970ac2d65f0d13a7c57eb454b210e92240 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?L=C3=A1szl=C3=B3=20N=C3=A9meth?=
<laszlo.nemeth@collabora.com>
Date: Mon, 12 Oct 2015 08:43:12 +0200
Subject: [PATCH] fix compound handling for new Hungarian orthography
The frequent cases of this compound limitation are handled by
the extended dictionary, but not these ones with both derivative
and inflectional suffixes.
---
src/hunspell/affixmgr.cxx | 19 +++++++++++++++----
src/hunspell/affixmgr.hxx | 1 +
2 files changed, 16 insertions(+), 4 deletions(-)
diff --git a/src/hunspell/affixmgr.cxx b/src/hunspell/affixmgr.cxx
index 0992e6e..0950425 100644
--- misc/hunspell-1.3.3/src/hunspell/affixmgr.cxx
+++ misc/build/hunspell-1.3.3/src/hunspell/affixmgr.cxx
@@ -139,8 +139,9 @@ AffixMgr::AffixMgr(const char * affpath, HashMgr** ptr, int * md, const char * k
cpdvowels=NULL; // vowels (for calculating of Hungarian compounding limit, O(n) search! XXX)
cpdvowels_utf16=NULL; // vowels for UTF-8 encoding (bsearch instead of O(n) search)
cpdvowels_utf16_len=0; // vowels
- pfxappnd=NULL; // previous prefix for counting the syllables of prefix BUG
- sfxappnd=NULL; // previous suffix for counting a special syllables BUG
+ pfxappnd=NULL; // previous prefix for counting syllables of the prefix BUG
+ sfxappnd=NULL; // previous suffix for counting syllables of the suffix BUG
+ sfxextra=0; // modifier for syllable count of sfxappnd BUG
cpdsyllablenum=NULL; // syllable count incrementing flag
checknum=0; // checking numbers, and word with numbers
wordchars=NULL; // letters + spec. word characters
@@ -1201,6 +1202,7 @@ struct hentry * AffixMgr::prefix_check(const char * word, int len, char in_compo
pfx = NULL;
pfxappnd = NULL;
sfxappnd = NULL;
+ sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry * pe = pStart[0];
@@ -1261,6 +1263,7 @@ struct hentry * AffixMgr::prefix_check_twosfx(const char * word, int len,
pfx = NULL;
sfxappnd = NULL;
+ sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry * pe = pStart[0];
@@ -1302,6 +1305,7 @@ char * AffixMgr::prefix_check_morph(const char * word, int len, char in_compound
pfx = NULL;
sfxappnd = NULL;
+ sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry * pe = pStart[0];
@@ -1353,6 +1357,7 @@ char * AffixMgr::prefix_check_twosfx_morph(const char * word, int len,
pfx = NULL;
sfxappnd = NULL;
+ sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry * pe = pStart[0];
@@ -1993,7 +1998,7 @@ struct hentry * AffixMgr::compound_check(const char * word, int len,
// XXX only second suffix (inflections, not derivations)
if (sfxappnd) {
char * tmp = myrevstrdup(sfxappnd);
- numsyllable -= get_syllable(tmp, strlen(tmp));
+ numsyllable -= get_syllable(tmp, strlen(tmp)) + sfxextra;
free(tmp);
}
@@ -2512,7 +2517,7 @@ int AffixMgr::compound_check_morph(const char * word, int len,
// XXX only second suffix (inflections, not derivations)
if (sfxappnd) {
char * tmp = myrevstrdup(sfxappnd);
- numsyllable -= get_syllable(tmp, strlen(tmp));
+ numsyllable -= get_syllable(tmp, strlen(tmp)) + sfxextra;
free(tmp);
}
@@ -2696,6 +2701,12 @@ struct hentry * AffixMgr::suffix_check (const char * word, int len,
sfx=sptr; // BUG: sfx not stateless
sfxflag = sptr->getFlag(); // BUG: sfxflag not stateless
if (!sptr->getCont()) sfxappnd=sptr->getKey(); // BUG: sfxappnd not stateless
+// LANG_hu section: spec. Hungarian rule
+ else if (langnum == LANG_hu && sptr->getKeyLen() && sptr->getKey()[0] == 'i' &&
+ sptr->getKey()[1] != 'y' && sptr->getKey()[1] != 't') {
+ sfxextra = 1;
+ }
+// END of LANG_hu section
return rv;
}
}
diff --git a/src/hunspell/affixmgr.hxx b/src/hunspell/affixmgr.hxx
index 2679b7f..8839814 100644
--- misc/hunspell-1.3.3/src/hunspell/affixmgr.hxx
+++ misc/build/hunspell-1.3.3/src/hunspell/affixmgr.hxx
@@ -155,6 +155,7 @@ class LIBHUNSPELL_DLL_EXPORTED AffixMgr
char * cpdsyllablenum;
const char * pfxappnd; // BUG: not stateless
const char * sfxappnd; // BUG: not stateless
+ int sfxextra; // BUG: not stateless
FLAG sfxflag; // BUG: not stateless
char * derived; // BUG: not stateless
SfxEntry * sfx; // BUG: not stateless
--
1.9.1
diff --git a/external/hunspell/hunspell-windows.patch b/external/hunspell/hunspell-windows.patch
new file mode 100644
index 0000000..357272d
--- /dev/null
+++ b/external/hunspell/hunspell-windows.patch
@@ -0,0 +1,21 @@
--- /dev/null 2016-03-15 09:11:25.292954614 +0000
+++ misc/hunspell-1.3.4/src/hunspell/hunvisapi.h 2016-04-06 16:40:47.181698825 +0100
@@ -0,0 +1,18 @@
+#ifndef _HUNSPELL_VISIBILITY_H_
+#define _HUNSPELL_VISIBILITY_H_
+
+#if defined(HUNSPELL_STATIC)
+# define LIBHUNSPELL_DLL_EXPORTED
+#elif defined(_MSC_VER)
+# if defined(BUILDING_LIBHUNSPELL)
+# define LIBHUNSPELL_DLL_EXPORTED __declspec(dllexport)
+# else
+# define LIBHUNSPELL_DLL_EXPORTED __declspec(dllimport)
+# endif
+#elif defined(BUILDING_LIBHUNSPELL) && 1
+# define LIBHUNSPELL_DLL_EXPORTED __attribute__((__visibility__("default")))
+#else
+# define LIBHUNSPELL_DLL_EXPORTED
+#endif
+
+#endif
diff --git a/external/hunspell/ubsan.patch.0 b/external/hunspell/ubsan.patch.0
deleted file mode 100644
index b52802d..0000000
--- a/external/hunspell/ubsan.patch.0
+++ /dev/null
@@ -1,35 +0,0 @@
--- src/hunspell/hashmgr.cxx
+++ src/hunspell/hashmgr.cxx
@@ -479,7 +479,7 @@
int HashMgr::hash(const char * word) const
{
- long hv = 0;
+ unsigned long hv = 0;
for (int i=0; i < 4 && *word != 0; i++)
hv = (hv << 8) | (*word++);
while (*word != 0) {
--- src/hunspell/csutil.cxx
+++ src/hunspell/csutil.cxx
@@ -147,7 +147,7 @@
case 0xd0: { // 2-byte UTF-8 codes
if ((*(u8+1) & 0xc0) == 0x80) {
u2->h = (*u8 & 0x1f) >> 2;
- u2->l = (*u8 << 6) + (*(u8+1) & 0x3f);
+ u2->l = (*reinterpret_cast<unsigned char const *>(u8) << 6) + (*(u8+1) & 0x3f);
u8++;
} else {
HUNSPELL_WARNING(stderr, "UTF-8 encoding error. Missing continuation byte in %ld. character position:\n%s\n", static_cast<long>(u8 - (signed char *)src), src);
@@ -158,10 +158,10 @@
}
case 0xe0: { // 3-byte UTF-8 codes
if ((*(u8+1) & 0xc0) == 0x80) {
- u2->h = ((*u8 & 0x0f) << 4) + ((*(u8+1) & 0x3f) >> 2);
+ u2->h = ((*reinterpret_cast<unsigned char const *>(u8) & 0x0f) << 4) + ((*(u8+1) & 0x3f) >> 2);
u8++;
if ((*(u8+1) & 0xc0) == 0x80) {
- u2->l = (*u8 << 6) + (*(u8+1) & 0x3f);
+ u2->l = (*reinterpret_cast<unsigned char const *>(u8) << 6) + (*(u8+1) & 0x3f);
u8++;
} else {
HUNSPELL_WARNING(stderr, "UTF-8 encoding error. Missing continuation byte in %ld. character position:\n%s\n", static_cast<long>(u8 - (signed char *)src), src);
