Sync our Snowball stemmer dictionaries with current upstream

The main change is a new stemmer for Greek.  There are minor changes
in the Danish and French stemmers.

Author: Panagiotis Mavrogiorgos <pmav99@gmail.com>

doc/src/sgml/textsearch.sgml

@@ -3810,6 +3810,7 @@ Parser: "pg_catalog.default"
  pg_catalog | finnish_stem    | snowball stemmer for finnish language
  pg_catalog | french_stem     | snowball stemmer for french language
  pg_catalog | german_stem     | snowball stemmer for german language
+ pg_catalog | greek_stem      | snowball stemmer for greek language
  pg_catalog | hungarian_stem  | snowball stemmer for hungarian language
  pg_catalog | indonesian_stem | snowball stemmer for indonesian language
  pg_catalog | irish_stem      | snowball stemmer for irish language

src/backend/snowball/Makefile

@@ -41,6 +41,7 @@ OBJS= $(WIN32RES) dict_snowball.o api.o utilities.o \
 	stem_UTF_8_finnish.o \
 	stem_UTF_8_french.o \
 	stem_UTF_8_german.o \
+	stem_UTF_8_greek.o \
 	stem_UTF_8_hungarian.o \
 	stem_UTF_8_indonesian.o \
 	stem_UTF_8_irish.o \
@@ -69,6 +70,7 @@ LANGUAGES=  \
 	finnish		finnish		\
 	french		french		\
 	german		german		\
+	greek		greek		\
 	hungarian	hungarian	\
 	indonesian	indonesian	\
 	irish		irish		\

src/backend/snowball/README

@@ -29,8 +29,8 @@ We choose to include the derived files in the PostgreSQL distribution
 because most installations will not have the Snowball compiler available.
 
 We are currently synced with the Snowball git commit
-1964ce688cbeca505263c8f77e16ed923296ce7a
-of 2018-06-29.
+4456b82c26c02493e8807a66f30593a98c5d2888
+of 2019-06-24.
 
 To update the PostgreSQL sources from a new Snowball version:
 
@@ -57,7 +57,7 @@ do not require any changes.
 4. Check whether any stemmer modules have been added or removed.  If so, edit
 the OBJS list in Makefile, the list of #include's in dict_snowball.c, and the
 stemmer_modules[] table in dict_snowball.c.  You might also need to change
-the LANGUAGES list in Makefile.
+the LANGUAGES list in Makefile and tsearch_config_languages in initdb.c.
 
 5. The various stopword files in stopwords/ must be downloaded
 individually from pages on the snowballstem.org website.

src/backend/snowball/dict_snowball.c

@@ -50,6 +50,7 @@
 #include "snowball/libstemmer/stem_UTF_8_finnish.h"
 #include "snowball/libstemmer/stem_UTF_8_french.h"
 #include "snowball/libstemmer/stem_UTF_8_german.h"
+#include "snowball/libstemmer/stem_UTF_8_greek.h"
 #include "snowball/libstemmer/stem_UTF_8_hungarian.h"
 #include "snowball/libstemmer/stem_UTF_8_indonesian.h"
 #include "snowball/libstemmer/stem_UTF_8_irish.h"
@@ -115,6 +116,7 @@ static const stemmer_module stemmer_modules[] =
 	STEMMER_MODULE(finnish, PG_UTF8, UTF_8),
 	STEMMER_MODULE(french, PG_UTF8, UTF_8),
 	STEMMER_MODULE(german, PG_UTF8, UTF_8),
+	STEMMER_MODULE(greek, PG_UTF8, UTF_8),
 	STEMMER_MODULE(hungarian, PG_UTF8, UTF_8),
 	STEMMER_MODULE(indonesian, PG_UTF8, UTF_8),
 	STEMMER_MODULE(irish, PG_UTF8, UTF_8),

src/backend/snowball/libstemmer/utilities.c

@@ -59,31 +59,49 @@ extern int skip_utf8(const symbol * p, int c, int lb, int l, int n) {
 /* Code for character groupings: utf8 cases */
 
 static int get_utf8(const symbol * p, int c, int l, int * slot) {
-    int b0, b1;
+    int b0, b1, b2;
     if (c >= l) return 0;
     b0 = p[c++];
     if (b0 < 0xC0 || c == l) {   /* 1100 0000 */
-        * slot = b0; return 1;
+        *slot = b0;
+        return 1;
     }
-    b1 = p[c++];
+    b1 = p[c++] & 0x3F;
     if (b0 < 0xE0 || c == l) {   /* 1110 0000 */
-        * slot = (b0 & 0x1F) << 6 | (b1 & 0x3F); return 2;
+        *slot = (b0 & 0x1F) << 6 | b1;
+        return 2;
     }
-    * slot = (b0 & 0xF) << 12 | (b1 & 0x3F) << 6 | (p[c] & 0x3F); return 3;
+    b2 = p[c++] & 0x3F;
+    if (b0 < 0xF0 || c == l) {   /* 1111 0000 */
+        *slot = (b0 & 0xF) << 12 | b1 << 6 | b2;
+        return 3;
+    }
+    *slot = (b0 & 0xE) << 18 | b1 << 12 | b2 << 6 | (p[c] & 0x3F);
+    return 4;
 }
 
 static int get_b_utf8(const symbol * p, int c, int lb, int * slot) {
-    int b0, b1;
+    int a, b;
     if (c <= lb) return 0;
-    b0 = p[--c];
-    if (b0 < 0x80 || c == lb) {   /* 1000 0000 */
-        * slot = b0; return 1;
+    b = p[--c];
+    if (b < 0x80 || c == lb) {   /* 1000 0000 */
+        *slot = b;
+        return 1;
+    }
+    a = b & 0x3F;
+    b = p[--c];
+    if (b >= 0xC0 || c == lb) {   /* 1100 0000 */
+        *slot = (b & 0x1F) << 6 | a;
+        return 2;
     }
-    b1 = p[--c];
-    if (b1 >= 0xC0 || c == lb) {   /* 1100 0000 */
-        * slot = (b1 & 0x1F) << 6 | (b0 & 0x3F); return 2;
+    a |= (b & 0x3F) << 6;
+    b = p[--c];
+    if (b >= 0xE0 || c == lb) {   /* 1110 0000 */
+        *slot = (b & 0xF) << 12 | a;
+        return 3;
     }
-    * slot = (p[--c] & 0xF) << 12 | (b1 & 0x3F) << 6 | (b0 & 0x3F); return 3;
+    *slot = (p[--c] & 0xE) << 18 | (b & 0x3F) << 12 | a;
+    return 4;
 }
 
 extern int in_grouping_U(struct SN_env * z, const unsigned char * s, int min, int max, int repeat) {
@@ -230,8 +248,13 @@ extern int find_among(struct SN_env * z, const struct among * v, int v_size) {
                 common++;
             }
         }
-        if (diff < 0) { j = k; common_j = common; }
-                 else { i = k; common_i = common; }
+        if (diff < 0) {
+            j = k;
+            common_j = common;
+        } else {
+            i = k;
+            common_i = common;
+        }
         if (j - i <= 1) {
             if (i > 0) break; /* v->s has been inspected */
             if (j == i) break; /* only one item in v */
@@ -360,9 +383,8 @@ extern int replace_s(struct SN_env * z, int c_bra, int c_ket, int s_size, const
         z->l += adjustment;
         if (z->c >= c_ket)
             z->c += adjustment;
-        else
-            if (z->c > c_bra)
-                z->c = c_bra;
+        else if (z->c > c_bra)
+            z->c = c_bra;
     }
     if (s_size) memmove(z->p + c_bra, s, s_size * sizeof(symbol));
     if (adjptr != NULL)

src/bin/initdb/initdb.c

@@ -716,6 +716,8 @@ static const struct tsearch_config_match tsearch_config_languages[] =
 	{"french", "French"},
 	{"german", "de"},
 	{"german", "German"},
+	{"greek", "el"},
+	{"greek", "Greek"},
 	{"hungarian", "hu"},
 	{"hungarian", "Hungarian"},
 	{"indonesian", "id"},

src/include/catalog/catversion.h

@@ -53,6 +53,6 @@
  */
 
 /*							yyyymmddN */
-#define CATALOG_VERSION_NO	201906161
+#define CATALOG_VERSION_NO	201907041
 
 #endif

src/include/snowball/libstemmer/api.h

@@ -19,8 +19,15 @@ struct SN_env {
     unsigned char * B;
 };
 
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 extern struct SN_env * SN_create_env(int S_size, int I_size, int B_size);
 extern void SN_close_env(struct SN_env * z, int S_size);
 
 extern int SN_set_current(struct SN_env * z, int size, const symbol * s);
 
+#ifdef __cplusplus
+}
+#endif

src/include/snowball/libstemmer/stem_UTF_8_greek.h

+/* This file was generated automatically by the Snowball to ISO C compiler */
+/* http://snowballstem.org/ */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+extern struct SN_env * greek_UTF_8_create_env(void);
+extern void greek_UTF_8_close_env(struct SN_env * z);
+
+extern int greek_UTF_8_stem(struct SN_env * z);
+
+#ifdef __cplusplus
+}
+#endif
+