Added encoding mapping via configuration file.

diff --git a/Makefile b/Makefile
index 5b3e9b149afe395319d9623d52c29314ccc30164..125b82e0b281410b448c600f6205c72dd60d8cd6 100644 (file)
--- a/Makefile
+++ b/Makefile
@@ -4,7 +4,7 @@
 YACC=bison
 LEX=flex
 
-CFLAGS=-g -Wall -pedantic
+CFLAGS=-g -W -Wall -pedantic
 YFLAGS=--debug --defines
 LFLAGS=-8
 
@@ -13,6 +13,8 @@ gedcom_parse: standalone.o lex.gedcom_1byte_.o lex.gedcom_hilo_.o \
                encoding.o
        $(CC) $(LDFLAGS) $^ $(LOADLIBES) $(LDLIBS) -o $@
 
+libgedcom.so:
+
 lex.gedcom_1byte_.c:   gedcom_1byte.lex gedcom.tab.h gedcom.h multilex.h
        $(LEX) $(LFLAGS) -Pgedcom_1byte_ gedcom_1byte.lex
 
@@ -29,7 +31,7 @@ clean:
        rm -f core gedcom_parse test_* *.o lex.gedcom_* \
         gedcom.tab.* gedcom.output
 
-# Test programs
+# Lexer test programs
 
 test_1byte:    lex.gedcom_1byte_.test.o message.o encoding.o
        $(CC) $(LDFLAGS) $^ $(LOADLIBES) $(LDLIBS) -o $@
@@ -48,3 +50,11 @@ test_lohi:   lex.gedcom_lohi_.test.o message.o encoding.o
 
 lex.gedcom_lohi_.test.o:       lex.gedcom_lohi_.c
        $(CC) -DLEXER_TEST -c $(CPPFLAGS) $(CFLAGS) $^ -o $@
+
+# Test of parser
+
+test:  gedcom_parse
+       @for file in t/*.ged; do \
+         echo "=== testing $$file"; \
+         ./gedcom_parse $$file; \
+       done

diff --git a/encoding.c b/encoding.c
index 08f69dbb370f0fcd7cd49cf6af7378aceb5319f9..ba5aa1d73395778815776000c01a48889aff7b46 100644 (file)
--- a/encoding.c
+++ b/encoding.c
@@ -1,24 +1,107 @@
 #include <string.h>
 #include <iconv.h>
+#include <search.h>
+#include <stdio.h>
 #include "gedcom.h"
 #include "encoding.h"
 
 #define INTERNAL_ENCODING "UTF8"
+#define ENCODING_CONF_FILE "gedcom.enc"
+#define MAXBUF 255
 
 static iconv_t cd_to_internal = (iconv_t) -1;
 static char int_buf[MAXGEDCLINELEN*2];
+static void *encoding_mapping = NULL;
+
+struct node {
+  char *gedcom_name;
+  char *iconv_name;
+};
+
+int node_compare(const void *node1, const void *node2)
+{
+  return strcmp(((const struct node *) node1)->gedcom_name,
+               ((const struct node *) node2)->gedcom_name);
+}
+
+void add_encoding(char *gedcom_n, char *iconv_n)
+{
+  void **datum;
+  struct node *nodeptr = (struct node *) malloc(sizeof *nodeptr);
+  nodeptr->gedcom_name = (char *) malloc(strlen(gedcom_n) + 1);
+  nodeptr->iconv_name  = (char *) malloc(strlen(iconv_n) + 1);
+  strcpy(nodeptr->gedcom_name, gedcom_n);
+  strcpy(nodeptr->iconv_name, iconv_n);
+  datum = tsearch(nodeptr, &encoding_mapping, node_compare);
+  if ((datum == NULL) || (*datum != nodeptr)) {
+    gedcom_warning("Duplicate entry found for encoding '%s', ignoring",
+                  gedcom_n);
+  }
+}
+
+char* get_encoding(char* gedcom_n)
+{
+  void **datum;
+  struct node search_node;
+  search_node.gedcom_name = gedcom_n;
+  datum = tfind(&search_node, &encoding_mapping, node_compare);
+  if (datum == NULL) {
+    gedcom_error("No encoding found for '%s'", gedcom_n);
+    return NULL;
+  }
+  else {
+    return ((const struct node *) *datum)->iconv_name;
+  }
+}
+
+void init_encodings()
+{
+  if (encoding_mapping == NULL) {
+    FILE *in;
+    char buffer[MAXBUF + 1];
+    char gedcom_n[MAXBUF + 1];
+    char iconv_n[MAXBUF + 1];
+    in = fopen(ENCODING_CONF_FILE, "r");
+    if (in != NULL) {
+      while (fgets(buffer, sizeof(buffer), in) != NULL) {
+       if (buffer[strlen(buffer) - 1] != '\n') {
+         gedcom_error("Line too long in encoding configuration file '%s'",
+                      ENCODING_CONF_FILE);
+         return;
+       }
+       else if (buffer[0] != '#') {
+         if (sscanf(buffer, "%s %s", gedcom_n, iconv_n) == 2) {
+           add_encoding(gedcom_n, iconv_n);
+         }
+       }
+      }
+      fclose(in);
+    }
+    else {
+      gedcom_warning("Could not open encoding configuration file '%s'",
+                    ENCODING_CONF_FILE);
+    }
+  }
+}
 
 int open_conv_to_internal(char* fromcode)
 {
+  char *encoding = get_encoding(fromcode);
   if (cd_to_internal != (iconv_t) -1)
     iconv_close(cd_to_internal);
-  cd_to_internal = iconv_open(INTERNAL_ENCODING, fromcode);
+  if (encoding == NULL) {
+    cd_to_internal = (iconv_t) -1;
+  }
+  else {
+    cd_to_internal = iconv_open(INTERNAL_ENCODING, encoding);
+  }
   return (cd_to_internal != (iconv_t) -1);  
 }
 
 void close_conv_to_internal()
 {
   iconv_close(cd_to_internal);
+  cd_to_internal = (iconv_t) -1;
 }
 
 char* to_internal(char* str, size_t len)

diff --git a/encoding.h b/encoding.h
index c4b70ee58a3247f6f5245f75137727c0e84e769d..287e478e9bd636747debf02f06d736e713227b7a 100644 (file)
--- a/encoding.h
+++ b/encoding.h
@@ -1,3 +1,5 @@
 int open_conv_to_internal(char* fromcode);
 void close_conv_to_internal();
 char* to_internal(char* str, size_t len);
+void init_encodings();
+char* get_encoding(char* gedcom_name);

diff --git a/gedcom.enc b/gedcom.enc
new file mode 100644 (file)
index 0000000..8c2a56b
--- /dev/null
+++ b/gedcom.enc
@@ -0,0 +1,8 @@
+# Mapping of charsets for gedcom parsing
+# Each line contains the gedcom name, appended with (LOHI) or (HILO)
+# for 2 byte encodings, and the iconv name of the charset, separated
+# by whitespace
+UNICODE(LOHI)  UTF16LE
+UNICODE(HILO)  UTF16BE
+ASCII          ASCII
+ANSI           CP1252

diff --git a/gedcom.y b/gedcom.y
index 7b8b2ef421440c6e4c36d4f0ea44dd76b6788729..8aad4fd9f02ed3726ac82eaf78c800fbc3568f65 100644 (file)
--- a/gedcom.y
+++ b/gedcom.y
@@ -2110,7 +2110,7 @@ opt_line_item : /* empty */ { }
               | DELIM line_item { }
               ;
 
-line_item   : anychar  { int i;
+line_item   : anychar  { size_t i;
                         CLEAR_BUFFER(string_buf);
                          string_buf_ptr = string_buf;
                         /* The following also takes care of '@@' */
@@ -2132,7 +2132,7 @@ line_item   : anychar  { int i;
                      YYERROR;
                    }
                    else {
-                     int i;
+                     size_t i;
                      /* The following also takes care of '@@' */
                      if (!strncmp($2, "@@", 3))
                        *string_buf_ptr++ = '@';

diff --git a/gedcom_lohi.lex b/gedcom_lohi.lex
index 6d88b435ed979f3daa309bc10e96417fd0cf354a..bc8546851b26851c6d2ee1f3aa2e7523187adae2 100644 (file)
--- a/gedcom_lohi.lex
+++ b/gedcom_lohi.lex
@@ -9,6 +9,8 @@
 #include "gedcom.h"
 #include "multilex.h"
 #include "encoding.h"
+
+#define YY_NO_UNPUT
 %}
 
 %s NORMAL
@@ -328,8 +330,9 @@ int yywrap()
 
 int main()
 {
-  int tok;
-  int res = open_conv_to_internal("UTF16LE");
+  int tok, res;
+  init_encodings();
+  res = open_conv_to_internal("UNICODE(LOHI)");
   if (!res) {
     gedcom_error("Unable to open conversion context: %s",
                 strerror(errno));

diff --git a/standalone.c b/standalone.c
index 452c1db394e1d88bc2996edbd2b3ae26313c36c2..cedfca21a154245e26c92ddb16c175477f79bfca 100644 (file)
--- a/standalone.c
+++ b/standalone.c
@@ -66,7 +66,7 @@ int main(int argc, char* argv[])
   gedcom_set_compat_handling(compat_enabled);
   gedcom_set_error_handling(mech);
   
-  if (gedcom_xxx_parse(file_name) == 0) {
+  if (gedcom_parse_file(file_name) == 0) {
     printf("Parse succeeded\n");
     return 0;
   }