Remove the corpus statistics code

It is obsolete and unsupported.
Per discussion at PR opencog#971.

ChangeLog

@@ -1,6 +1,7 @@
 Version 5.7.0 (XXX 2019)
  * Minor efficiency improvements to the SQL-backed dictionary.
  * Incompatible change to the Exp traversal API.
+ * Remove the obsolete and unsupported "corpus statistics" code.
 
 Version 5.6.2 (24 June 2019)
  * Bug-fix the SQL-backed dictionary.

README.md

@@ -126,7 +126,6 @@ Contents
 | ------------- |-------------|
 | LICENSE     | The license describing terms of use |
 | link-grammar/*.c | The program.  (Written in ANSI-C) |
-| link-grammar/corpus/*.c | Optional corpus statistics database. |
 | link-grammar/minisat/ | Optional SAT Solver. (Written in C++) |
 | link-grammar/sat-solver/ | Optional SAT Solver. (Written in C++) |
 |  |  |

bindings/java-jni/Makefile.am

@@ -37,10 +37,6 @@ liblink_grammar_java_la_LIBADD  += ${MINISAT_LIBS}
 endif !LIBMINISAT_BUNDLED
 endif
 
-if WITH_CORPUS
-liblink_grammar_java_la_LIBADD  += $(top_builddir)/link-grammar/corpus/liblink-corpus.la ${SQLITE3_LIBS}
-endif
-
 # Math libraries are needed for floorf, etc.
 liblink_grammar_java_la_LIBADD  += -lm
 

bindings/java-jni/jni-client.c

@@ -18,7 +18,6 @@
 #endif /* HAVE_STDATOMIC_H */
 
 #include "link-grammar/api-structures.h"
-#include "link-grammar/corpus/corpus.h"
 #include "link-grammar/error.h"
 #include "jni-client.h"
 #include "link-grammar/utilities.h"
@@ -479,55 +478,6 @@ Java_org_linkgrammar_LinkGrammar_getLinkageDisjunct(JNIEnv *env, jclass cls, jin
 	return j;
 }
 
-JNIEXPORT jstring JNICALL
-Java_org_linkgrammar_LinkGrammar_getLinkageSense(JNIEnv *env,
-                    jclass cls, jint i, jint j)
-{
-	per_thread_data *ptd = get_ptd(env, cls);
-	Linkage lkg = ptd->linkage;
-	Sense *sns;
-	const char * w = NULL;
-	jstring js;
-
-	if (!lkg) return NULL;
-	lg_corpus_linkage_senses(lkg);
-	sns = lg_get_word_sense(lkg, i);
-	while ((0 < j) && sns)
-	{
-		sns = lg_sense_next(sns);
-		j--;
-	}
-
-	/* does not need to be freed, points into data structures */
-	if (sns) w = lg_sense_get_sense(sns);
-
-	if (w) js = (*env)->NewStringUTF(env, w);
-	else js = NULL;
-	return js;
-}
-
-JNIEXPORT jdouble JNICALL
-Java_org_linkgrammar_LinkGrammar_getLinkageSenseScore(JNIEnv *env,
-                    jclass cls, jint i, jint j)
-{
-	per_thread_data *ptd = get_ptd(env, cls);
-	Linkage lkg = ptd->linkage;
-	Sense *sns;
-	double score = 0.0;
-
-	if (!lkg) return 0.0;
-	sns = lg_get_word_sense(lkg, i);
-	while ((0 < j) && sns)
-	{
-		sns = lg_sense_next(sns);
-		j--;
-	}
-
-	if (sns) score = lg_sense_get_score(sns);
-
-	return score;
-}
-
 /*
  * Class:      LinkGrammar
  * Method:     numSkippedWords

configure.ac

@@ -398,21 +398,6 @@ fi
 AM_CONDITIONAL(LIBMINISAT_BUNDLED, test -n "$use_minisat_bundled_library")
 AM_CONDITIONAL(WITH_SAT_SOLVER, test "x$enable_sat_solver" = xyes)
 
-# ====================================================================
-# Corpus stats disabled by default
-
-AC_ARG_ENABLE( corpus-stats,
-  [  --enable-corpus-stats   use corpus statistics],
-  [],
-  [enable_corpus_stats=no]
-)
-if test "x$enable_corpus_stats" = "xyes"
-then
-	AC_DEFINE(USE_CORPUS, 1, [Define for compilation])
-fi
-
-AM_CONDITIONAL(WITH_CORPUS, test x${enable_corpus_stats} = xyes)
-
 # ====================================================================
 
 AC_ARG_ENABLE( wordgraph_display,
@@ -1021,7 +1006,6 @@ Makefile
 link-grammar.pc
 link-grammar.spec
 link-grammar/Makefile
-link-grammar/corpus/Makefile
 link-grammar/minisat/Makefile
 link-grammar/sat-solver/Makefile
 link-grammar/link-features.h
@@ -1089,7 +1073,6 @@ $PACKAGE-$VERSION  build configuration settings
 	HunSpell dictionary location:   ${HunSpellDictDir}${HunSpellDictDir_status}
 	Boolean SAT parser:             ${enable_sat_solver}${use_minisat_bundled_library}
 	SQLite-backed dictionary:       ${SQLiteFound}
-	Corpus statistics database:     ${enable_corpus_stats}
 	RegEx tokenizer:                ${enable_regex_tokenizer}
 	Definitions:                    ${DEFS}
 	Libraries:                      ${lglibs}

link-grammar/Makefile.am

@@ -12,10 +12,6 @@ if WITH_SAT_SOLVER
 SUBDIRS += $(MINISAT_DIR) sat-solver
 endif
 
-if WITH_CORPUS
-SUBDIRS += corpus
-endif
-
 if OS_WIN32
 install-libtool-import-lib:
 	$(INSTALL) .libs/liblink-grammar.dll.a $(DESTDIR)$(libdir)
@@ -82,10 +78,6 @@ liblink_grammar_la_LIBADD  += ${MINISAT_LIBS}
 endif !LIBMINISAT_BUNDLED
 endif WITH_SAT_SOLVER
 
-if WITH_CORPUS
-liblink_grammar_la_LIBADD  += $(top_builddir)/link-grammar/corpus/liblink-corpus.la ${SQLITE3_LIBS}
-endif
-
 # Math libraries are needed for floorf, etc.
 liblink_grammar_la_LIBADD  += -lm
 
@@ -126,7 +118,6 @@ liblink_grammar_la_SOURCES =        \
 	post-process/pp_lexer.l          \
 	post-process/pp_linkset.c        \
 	prepare/build-disjuncts.c        \
-	prepare/expand.c                 \
 	prepare/exprune.c                \
 	print/print.c                    \
 	print/print-util.c               \
@@ -186,7 +177,6 @@ liblink_grammar_la_SOURCES =        \
 	post-process/pp_linkset.h        \
 	post-process/pp-structures.h     \
 	prepare/build-disjuncts.h        \
-	prepare/expand.h                 \
 	prepare/exprune.h                \
 	print/print.h                    \
 	print/print-util.h               \

link-grammar/README.md

@@ -23,8 +23,6 @@ Listed in rough processing order.
  * post-process - Constraint rules applied after parsing.
                   Also, rules for constituent-tree construction.
 
- * corpus       - word-sense disambiguation data from an SQL file.
-
  * print        - functions that convert parses into
                   human-readable strings and printed output.
 

link-grammar/api-structures.h

@@ -45,7 +45,6 @@
 
 #include "api-types.h"
 #include "tracon-set.h"
-#include "corpus/corpus.h"
 #include "memory-pool.h"
 #include "string-set.h"
 

link-grammar/api-types.h

@@ -41,10 +41,6 @@ typedef struct pp_knowledge_s pp_knowledge;
 typedef struct pp_linkset_s pp_linkset;
 typedef struct PP_domains_s PP_domains;
 
-typedef struct corpus_s Corpus;
-typedef struct sense_s Sense;
-typedef struct cluster_s Cluster;
-
 typedef struct Wordgraph_pathpos_s Wordgraph_pathpos;
 
 #endif