From 04e9704b9e77c57bb1b0a06876977c1b255376ed Mon Sep 17 00:00:00 2001
From: Teodor Sigaev <teodor@sigaev.ru>
Date: Fri, 9 Jun 2006 13:25:59 +0000
Subject: [PATCH] Now ispell dictionary can eat dictionaries in MySpell format,
used by OpenOffice. Dictionaries are placed at
http://lingucomponent.openoffice.org/spell_dic.html Dictionary automatically
recognizes format of files.
Warning. MySpell's format has limitation with compound
word support: it's impossible to mark affix as
compound-only affix. So for norwegian, german etc
languages it's recommended to use original ispell format.
For that reason I don't want to remove my2ispell
scripts, it's has workaround at least for norwegian language.
---
contrib/tsearch2/ispell/spell.c | 94 +++++++++++++++++++++++++++++++--
contrib/tsearch2/ispell/spell.h | 1 +
2 files changed, 92 insertions(+), 3 deletions(-)
diff --git a/contrib/tsearch2/ispell/spell.c b/contrib/tsearch2/ispell/spell.c
index 223ae4a9ada..28f38eefd32 100644
--- a/contrib/tsearch2/ispell/spell.c
+++ b/contrib/tsearch2/ispell/spell.c
@@ -391,6 +391,7 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
char flagflags = 0;
FILE *affix;
int line=0;
+ int oldformat = 0;
if (!(affix = fopen(filename, "r")))
return (1);
@@ -412,6 +413,7 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
while (*s && t_isspace(s)) s++;
if ( *s && pg_mblen(s) == 1 )
Conf->compoundcontrol = *s;
+ oldformat++;
continue;
}
}
@@ -419,12 +421,14 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
{
suffixes = 1;
prefixes = 0;
+ oldformat++;
continue;
}
if (STRNCMP(tmpstr, "prefixes") == 0)
{
suffixes = 0;
prefixes = 1;
+ oldformat++;
continue;
}
if (STRNCMP(tmpstr, "flag") == 0)
@@ -433,10 +437,11 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
flagflags = 0;
while (*s && t_isspace(s)) s++;
+ oldformat++;
/* allow only single-encoded flags */
- if ( pg_mblen(s) != 1 )
- continue;
+ if ( pg_mblen(s) != 1 )
+ elog(ERROR,"Multiencoded flag at line %d: %s", line, s);
if (*s == '*')
{
@@ -455,12 +460,22 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
/* allow only single-encoded flags */
if ( pg_mblen(s) != 1 ) {
flagflags = 0;
- continue;
+ elog(ERROR,"Multiencoded flag at line %d: %s", line, s);
}
flag = (unsigned char) *s;
continue;
}
+ if ( STRNCMP(str, "COMPOUNDFLAG") == 0 || STRNCMP(str, "COMPOUNDMIN") == 0 ||
+ STRNCMP(str, "PFX")==0 || STRNCMP(str, "SFX")==0 ) {
+
+ if ( oldformat )
+ elog(ERROR,"Wrong affix file format");
+
+ fclose(affix);
+ return NIImportOOAffixes(Conf, filename);
+
+ }
if ((!suffixes) && (!prefixes))
continue;
@@ -475,6 +490,79 @@ NIImportAffixes(IspellDict * Conf, const char *filename)
return (0);
}
+int
+NIImportOOAffixes(IspellDict * Conf, const char *filename) {
+ char str[BUFSIZ];
+ char type[BUFSIZ];
+ char sflag[BUFSIZ];
+ char mask[BUFSIZ];
+ char find[BUFSIZ];
+ char repl[BUFSIZ];
+ bool isSuffix = false;
+ int flag = 0;
+ char flagflags = 0;
+ FILE *affix;
+ int line=0;
+ int scanread = 0;
+ char scanbuf[BUFSIZ];
+
+ sprintf(scanbuf,"%%6s %%%ds %%%ds %%%ds %%%ds", BUFSIZ/5, BUFSIZ/5, BUFSIZ/5, BUFSIZ/5);
+
+ if (!(affix = fopen(filename, "r")))
+ return (1);
+ Conf->compoundcontrol = '\t';
+
+ while (fgets(str, sizeof(str), affix))
+ {
+ line++;
+ if ( *str == '\0' || t_isspace(str) || t_iseq(str,'#') )
+ continue;
+ pg_verifymbstr( str, strlen(str), false);
+
+ if ( STRNCMP(str, "COMPOUNDFLAG")==0 ) {
+ char *s = str+strlen("COMPOUNDFLAG");
+ while (*s && t_isspace(s)) s++;
+ if ( *s && pg_mblen(s) == 1 )
+ Conf->compoundcontrol = *s;
+ continue;
+ }
+
+ scanread = sscanf(str, scanbuf, type, sflag, find, repl, mask);
+
+ lowerstr(type);
+ if ( scanread<4 || (STRNCMP(type,"sfx") && STRNCMP(type,"pfx")) )
+ continue;
+
+ if ( scanread == 4 ) {
+ if ( strlen(sflag) != 1 )
+ continue;
+ flag = *sflag;
+ isSuffix = (STRNCMP(type,"sfx")==0) ? true : false;
+ lowerstr(find);
+ if ( t_iseq(find,'y') )
+ flagflags |= FF_CROSSPRODUCT;
+ else
+ flagflags = 0;
+ } else {
+ if ( strlen(sflag) != 1 || flag != *sflag || flag==0 )
+ continue;
+ lowerstr(repl);
+ lowerstr(find);
+ lowerstr(mask);
+ if ( t_iseq(find,'0') )
+ *find = '\0';
+ if ( t_iseq(repl,'0') )
+ *repl = '\0';
+
+ NIAddAffix(Conf, flag, flagflags, mask, find, repl, isSuffix ? FF_SUFFIX : FF_PREFIX);
+ }
+ }
+
+ fclose(affix);
+
+ return 0;
+}
+
static int
MergeAffix(IspellDict * Conf, int a1, int a2)
{
diff --git a/contrib/tsearch2/ispell/spell.h b/contrib/tsearch2/ispell/spell.h
index fc3240a1d8b..fe79888bf3e 100644
--- a/contrib/tsearch2/ispell/spell.h
+++ b/contrib/tsearch2/ispell/spell.h
@@ -121,6 +121,7 @@ typedef struct
TSLexeme *NINormalizeWord(IspellDict * Conf, char *word);
int NIImportAffixes(IspellDict * Conf, const char *filename);
+int NIImportOOAffixes(IspellDict * Conf, const char *filename);
int NIImportDictionary(IspellDict * Conf, const char *filename);
int NIAddSpell(IspellDict * Conf, const char *word, const char *flag);
--
GitLab