X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=src%2Ficu_chain_test.c;h=192d28cc4e2b341d29679837a058c5e43933c86d;hb=47fcaa2307337683215b398e3eaad2e06c5fe9db;hp=ddc926e1124be769922014af50869f5f574512e7;hpb=d060969f7c7f2a41142ae5dfdb945cda973c91ee;p=pazpar2-moved-to-github.git

diff --git a/src/icu_chain_test.c b/src/icu_chain_test.c
index ddc926e..192d28c 100644
--- a/src/icu_chain_test.c
+++ b/src/icu_chain_test.c
@@ -1,17 +1,39 @@
-/**
- gcc -I/usr/include/libxml2 -lxml2 -o icu-xml-convert icu-xml-convert.c
+/* $Id: icu_chain_test.c,v 1.6 2007-07-05 18:40:24 adam Exp $
+   Copyright (c) 2006-2007, Index Data.
+
+This file is part of Pazpar2.
+
+Pazpar2 is free software; you can redistribute it and/or modify it under
+the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 2, or (at your option) any later
+version.
+
+Pazpar2 is distributed in the hope that it will be useful, but WITHOUT ANY
+WARRANTY; without even the implied warranty of MERCHANTABILITY or
+FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+for more details.
+
+You should have received a copy of the GNU General Public License
+along with Pazpar2; see the file LICENSE.  If not, write to the
+Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
+02111-1307, USA.
  */
 
-#include <stdio.h>
+#if HAVE_CONFIG_H
+#include "cconfig.h"
+#endif
+
 #include <string.h>
 
-#define _GNU_SOURCE
 #include <stdio.h>
 #include <stdlib.h>
 
 //#include <yaz/xmalloc.h>
 #include <yaz/options.h>
 
+
+#ifdef HAVE_ICU
+
 #include <unicode/ucnv.h>
 #include <unicode/ustring.h>
 
@@ -31,63 +53,60 @@ static struct config_t {
   
 void print_option_error(const struct config_t *p_config)
 {  
-  fprintf(stderr, "Calling error, valid options are :\n");
-  fprintf(stderr, "icu_chain_test\n"
-          "   [-c (path/to/config/file.xml)]\n"
-          "   [-p (a|c|l|t)] print ICU info \n"
-          "   [-x] XML output\n"
-          "\n"
-          "Examples:\n"
-          "cat hugetextfile.txt | ./icu_chain_test -c config.xml \n"
-          "./icu_chain_test -p c\n"
-          "./icu_chain_test -p l -x\n"
-          "./icu_chain_test -p t -x\n"
+    fprintf(stderr, "Calling error, valid options are :\n");
+    fprintf(stderr, "icu_chain_test\n"
+            "   [-c (path/to/config/file.xml)]\n"
+            "   [-p (a|c|l|t)] print ICU info \n"
+            "   [-x] XML output\n"
+            "\n"
+            "Examples:\n"
+            "cat hugetextfile.txt | ./icu_chain_test -c config.xml \n"
+            "./icu_chain_test -p c\n"
+            "./icu_chain_test -p l -x\n"
+            "./icu_chain_test -p t -x\n"
           );
-  exit(1);
+    exit(1);
 }
 
-void read_params(int argc, char **argv, struct config_t *p_config){    
-  char *arg;
-  int ret;
-  
-  /* set default parameters */
-  p_config->conffile[0] = 0;
-  p_config->print[0] = 0;
-  p_config->xmloutput = 0;
-  p_config->chain = 0;
-  p_config->infile = stdin;
-  p_config->outfile = stdout;
-
-  /* set up command line parameters */
-  
-  while ((ret = options("c:p:x", argv, argc, &arg)) != -2)
+void read_params(int argc, char **argv, struct config_t *p_config)
+{    
+    char *arg;
+    int ret;
+    
+    /* set default parameters */
+    p_config->conffile[0] = 0;
+    p_config->print[0] = 0;
+    p_config->xmloutput = 0;
+    p_config->chain = 0;
+    p_config->infile = stdin;
+    p_config->outfile = stdout;
+    
+    /* set up command line parameters */
+    
+    while ((ret = options("c:p:x", argv, argc, &arg)) != -2)
     {
-      switch (ret)
+        switch (ret)
         {
         case 'c':
-          strcpy(p_config->conffile, arg);
-          break;
+            strcpy(p_config->conffile, arg);
+            break;
         case 'p':
-          strcpy(p_config->print, arg);
-          break;
+            strcpy(p_config->print, arg);
+            break;
         case 'x':
             p_config->xmloutput = 1;
-          break;
+            break;
         default:
-          print_option_error(p_config);
+            print_option_error(p_config);
         }
     }
     
-    //p_config->infile = fopen("/etc/passwd", "r");
-    
-
-
-  if ((!strlen(p_config->conffile)
-      && !strlen(p_config->print))
-      || !config.infile
-      || !config.outfile)
-
-    print_option_error(p_config);
+    if ((!strlen(p_config->conffile)
+         && !strlen(p_config->print))
+        || !config.infile
+        || !config.outfile)
+        
+        print_option_error(p_config);
 };
 
 
@@ -117,7 +136,7 @@ static void print_icu_converters(const struct config_t *p_config)
     count = ucnv_countAvailable();
     if (p_config->xmloutput)
         fprintf(config.outfile, "<converters count=\"%d\" default=\"%s\">\n",
-               count, ucnv_getDefaultName());
+                count, ucnv_getDefaultName());
     else {    
         fprintf(config.outfile, "Available ICU converters: %d\n", count);
         fprintf(config.outfile, "Default ICU Converter is: '%s'\n", ucnv_getDefaultName());
@@ -138,254 +157,254 @@ static void print_icu_converters(const struct config_t *p_config)
 
 static void print_icu_transliterators(const struct config_t *p_config)
 {
-  int32_t count;
-  int32_t i;
-
-  count = utrans_countAvailableIDs();
-
-  int32_t buf_cap = 128;
-  char buf[buf_cap];
-
-  if (p_config->xmloutput)
-    fprintf(config.outfile, "<transliterators count=\"%d\">\n",  count);
-   else 
-    fprintf(config.outfile, "Available ICU transliterators: %d\n", count);
-
-  for(i=0;i<count;i++)
+    int32_t count;
+    int32_t i;
+    
+    count = utrans_countAvailableIDs();
+    
+    int32_t buf_cap = 128;
+    char buf[buf_cap];
+    
+    if (p_config->xmloutput)
+        fprintf(config.outfile, "<transliterators count=\"%d\">\n",  count);
+    else 
+        fprintf(config.outfile, "Available ICU transliterators: %d\n", count);
+    
+    for(i = 0; i <count; i++)
     {
-      utrans_getAvailableID(i, buf, buf_cap);
-       if (p_config->xmloutput)
-         fprintf(config.outfile, "<transliterator id=\"%s\"/>\n", buf);
-       else
-         fprintf(config.outfile, " %s", buf);
+        utrans_getAvailableID(i, buf, buf_cap);
+        if (p_config->xmloutput)
+            fprintf(config.outfile, "<transliterator id=\"%s\"/>\n", buf);
+        else
+            fprintf(config.outfile, " %s", buf);
     }
-
-  if (p_config->xmloutput){
-    fprintf(config.outfile, "</transliterators>\n");
-  }
-  else
+    
+    if (p_config->xmloutput){
+        fprintf(config.outfile, "</transliterators>\n");
+    }
+    else
     {
-      fprintf(config.outfile, "\n\nUnicode Set Patterns:\n"
-             "   Pattern         Description\n"
-             "   Ranges          [a-z] 	The lower case letters a through z\n"
-             "   Named Chars     [abc123] The six characters a,b,c,1,2 and 3\n"
-             "   String          [abc{def}] chars a, b and c, and string 'def'\n"
-             "   Categories      [\\p{Letter}] Perl General Category 'Letter'.\n"
-             "   Categories      [:Letter:] Posix General Category 'Letter'.\n"
-             "\n"
-             "   Combination     Example\n"
-             "   Union           [[:Greek:] [:letter:]]\n"
-             "   Intersection    [[:Greek:] & [:letter:]]\n"
-             "   Set Complement  [[:Greek:] - [:letter:]]\n"
-             "   Complement      [^[:Greek:] [:letter:]]\n"
-             "\n"
+        fprintf(config.outfile, "\n\nUnicode Set Patterns:\n"
+                "   Pattern         Description\n"
+                "   Ranges          [a-z] 	The lower case letters a through z\n"
+                "   Named Chars     [abc123] The six characters a,b,c,1,2 and 3\n"
+                "   String          [abc{def}] chars a, b and c, and string 'def'\n"
+                "   Categories      [\\p{Letter}] Perl General Category 'Letter'.\n"
+                "   Categories      [:Letter:] Posix General Category 'Letter'.\n"
+                "\n"
+                "   Combination     Example\n"
+                "   Union           [[:Greek:] [:letter:]]\n"
+                "   Intersection    [[:Greek:] & [:letter:]]\n"
+                "   Set Complement  [[:Greek:] - [:letter:]]\n"
+                "   Complement      [^[:Greek:] [:letter:]]\n"
+                "\n"
              "see: http://icu.sourceforge.net/userguide/unicodeSet.html\n"
-             "\n"
-             "Examples:\n"
-             "   [:Punctuation:] Any-Remove\n"
-             "   [:Cased-Letter:] Any-Upper\n"
-             "   [:Control:] Any-Remove\n"
-             "   [:Decimal_Number:] Any-Remove\n"
-             "   [:Final_Punctuation:] Any-Remove\n"
-             "   [:Georgian:] Any-Upper\n"
-             "   [:Katakana:] Any-Remove\n"
-             "   [:Arabic:] Any-Remove\n"
-             "   [:Punctuation:] Remove\n"
-             "   [[:Punctuation:]-[.,]] Remove\n"
-             "   [:Line_Separator:] Any-Remove\n"
-             "   [:Math_Symbol:] Any-Remove\n"
-             "   Lower; [:^Letter:] Remove (word tokenization)\n"
-             "   [:^Number:] Remove (numeric tokenization)\n"
-             "   [:^Katagana:] Remove (remove everything except Katagana)\n"
-             "   Lower;[[:WhiteSpace:][:Punctuation:]] Remove (word tokenization)\n"
-             "   NFD; [:Nonspacing Mark:] Remove; NFC   (removes accents from characters)\n"
-             "   [A-Za-z]; Lower(); Latin-Katakana; Katakana-Hiragana (transforms latin and katagana to hiragana)\n"
-             "   [[:separator:][:start punctuation:][:initial punctuation:]] Remove \n"
-             "\n"
-             "see http://icu.sourceforge.net/userguide/Transform.html\n"
-             "    http://www.unicode.org/Public/UNIDATA/UCD.html\n"
-             "    http://icu.sourceforge.net/userguide/Transform.html\n"
-             "    http://icu.sourceforge.net/userguide/TransformRule.html\n"
-             );
-
-
-  fprintf(config.outfile, "\n\n");
-  
+                "\n"
+                "Examples:\n"
+                "   [:Punctuation:] Any-Remove\n"
+                "   [:Cased-Letter:] Any-Upper\n"
+                "   [:Control:] Any-Remove\n"
+                "   [:Decimal_Number:] Any-Remove\n"
+                "   [:Final_Punctuation:] Any-Remove\n"
+                "   [:Georgian:] Any-Upper\n"
+                "   [:Katakana:] Any-Remove\n"
+                "   [:Arabic:] Any-Remove\n"
+                "   [:Punctuation:] Remove\n"
+                "   [[:Punctuation:]-[.,]] Remove\n"
+                "   [:Line_Separator:] Any-Remove\n"
+                "   [:Math_Symbol:] Any-Remove\n"
+                "   Lower; [:^Letter:] Remove (word tokenization)\n"
+                "   [:^Number:] Remove (numeric tokenization)\n"
+                "   [:^Katagana:] Remove (remove everything except Katagana)\n"
+                "   Lower;[[:WhiteSpace:][:Punctuation:]] Remove (word tokenization)\n"
+                "   NFD; [:Nonspacing Mark:] Remove; NFC   (removes accents from characters)\n"
+                "   [A-Za-z]; Lower(); Latin-Katakana; Katakana-Hiragana (transforms latin and katagana to hiragana)\n"
+                "   [[:separator:][:start punctuation:][:initial punctuation:]] Remove \n"
+                "\n"
+                "see http://icu.sourceforge.net/userguide/Transform.html\n"
+                "    http://www.unicode.org/Public/UNIDATA/UCD.html\n"
+                "    http://icu.sourceforge.net/userguide/Transform.html\n"
+                "    http://icu.sourceforge.net/userguide/TransformRule.html\n"
+            );
+        
+        
+        fprintf(config.outfile, "\n\n");
+        
     }
 }
 
 static void print_icu_xml_locales(const struct config_t *p_config)
 {
-  int32_t count;
-  int32_t i;
-  UErrorCode status = U_ZERO_ERROR;
-
-  UChar keyword[64];
-  int32_t keyword_len = 0;
-  char keyword_str[128];
-  int32_t keyword_str_len = 0;
-
-  UChar language[64];
-  int32_t language_len = 0;
-  char lang_str[128];
-  int32_t lang_str_len = 0;
-
-  UChar script[64];
-  int32_t script_len = 0;
-  char script_str[128];
-  int32_t script_str_len = 0;
-
-  UChar location[64];
-  int32_t location_len = 0;
-  char location_str[128];
-  int32_t location_str_len = 0;
-
-  UChar variant[64];
-  int32_t variant_len = 0;
-  char variant_str[128];
-  int32_t variant_str_len = 0;
-
-  UChar name[64];
-  int32_t name_len = 0;
-  char name_str[128];
-  int32_t name_str_len = 0;
-
-  UChar localname[64];
-  int32_t localname_len = 0;
-  char localname_str[128];
-  int32_t localname_str_len = 0;
-
-  count = uloc_countAvailable() ;
-
-  if (p_config->xmloutput){
+    int32_t count;
+    int32_t i;
+    UErrorCode status = U_ZERO_ERROR;
     
-      fprintf(config.outfile, "<locales count=\"%d\" default=\"%s\" collations=\"%d\">\n", 
-             count, uloc_getDefault(), ucol_countAvailable());
-  }
+    UChar keyword[64];
+    int32_t keyword_len = 0;
+    char keyword_str[128];
+    int32_t keyword_str_len = 0;
+
+    UChar language[64];
+    int32_t language_len = 0;
+    char lang_str[128];
+    int32_t lang_str_len = 0;
+
+    UChar script[64];
+    int32_t script_len = 0;
+    char script_str[128];
+    int32_t script_str_len = 0;
+
+    UChar location[64];
+    int32_t location_len = 0;
+    char location_str[128];
+    int32_t location_str_len = 0;
+
+    UChar variant[64];
+    int32_t variant_len = 0;
+    char variant_str[128];
+    int32_t variant_str_len = 0;
+
+    UChar name[64];
+    int32_t name_len = 0;
+    char name_str[128];
+    int32_t name_str_len = 0;
+
+    UChar localname[64];
+    int32_t localname_len = 0;
+    char localname_str[128];
+    int32_t localname_str_len = 0;
+
+    count = uloc_countAvailable() ;
+
+    if (p_config->xmloutput){
+    
+        fprintf(config.outfile, "<locales count=\"%d\" default=\"%s\" collations=\"%d\">\n", 
+                count, uloc_getDefault(), ucol_countAvailable());
+    }
   
-  for(i=0;i<count;i++) 
-  {
+    for(i=0;i<count;i++) 
+    {
 
-    keyword_len 
-      = uloc_getDisplayKeyword(uloc_getAvailable(i), "en", 
-                                keyword, 64, 
-                                &status);
+        keyword_len 
+            = uloc_getDisplayKeyword(uloc_getAvailable(i), "en", 
+                                     keyword, 64, 
+                                     &status);
 
-    u_strToUTF8(keyword_str, 128, &keyword_str_len,
-                keyword, keyword_len,
-                &status);
+        u_strToUTF8(keyword_str, 128, &keyword_str_len,
+                    keyword, keyword_len,
+                    &status);
     
     
-    language_len 
-      = uloc_getDisplayLanguage(uloc_getAvailable(i), "en", 
-                                language, 64, 
-                                &status);
-
-    u_strToUTF8(lang_str, 128, &lang_str_len,
-                language, language_len,
-                &status);
-
-
-    script_len 
-      = uloc_getDisplayScript(uloc_getAvailable(i), "en", 
-                                script, 64, 
-                                &status);
-
-    u_strToUTF8(script_str, 128, &script_str_len,
-                script, script_len,
-                &status);
-
-    location_len 
-      = uloc_getDisplayCountry(uloc_getAvailable(i), "en", 
-                                location, 64, 
-                                &status);
-
-    u_strToUTF8(location_str, 128, &location_str_len,
-                location, location_len,
-                &status);
-
-    variant_len 
-      = uloc_getDisplayVariant(uloc_getAvailable(i), "en", 
-                                variant, 64, 
-                                &status);
-
-    u_strToUTF8(variant_str, 128, &variant_str_len,
-                variant, variant_len,
-                &status);
-
-    name_len 
-      = uloc_getDisplayName(uloc_getAvailable(i), "en", 
-                                name, 64, 
-                                &status);
-
-    u_strToUTF8(name_str, 128, &name_str_len,
-                name, name_len,
-                &status);
-
-    localname_len 
-      = uloc_getDisplayName(uloc_getAvailable(i), uloc_getAvailable(i), 
-                                localname, 64, 
-                                &status);
-
-    u_strToUTF8(localname_str, 128, &localname_str_len,
-                localname, localname_len,
-                &status);
-
-
-    if (p_config->xmloutput){
-      fprintf(config.outfile, "<locale id=\"%s\"", uloc_getAvailable(i)); 
-      /* fprintf(config.outfile, " locale=\"%s\"", uloc_getAvailable(i)); */
-      /* if (strlen(keyword_str)) */
-      /*   fprintf(config.outfile, " keyword=\"%s\"", keyword_str); */
-      /* if (ucol_getAvailable(i)) */
-      /*   fprintf(config.outfile, " collation=\"1\""); */
-      if (strlen(lang_str))
-        fprintf(config.outfile, " language=\"%s\"", lang_str);
-      if (strlen(script_str))
-        fprintf(config.outfile, " script=\"%s\"", script_str);
-      if (strlen(location_str))
-        fprintf(config.outfile, " location=\"%s\"", location_str);
-      if (strlen(variant_str))
-        fprintf(config.outfile, " variant=\"%s\"", variant_str);
-      if (strlen(name_str))
-        fprintf(config.outfile, " name=\"%s\"", name_str);
-      if (strlen(localname_str))
-        fprintf(config.outfile, " localname=\"%s\"", localname_str);
-      fprintf(config.outfile, ">");
-      if (strlen(localname_str))
-        fprintf(config.outfile, "%s", localname_str);
-      fprintf(config.outfile, "</locale>\n"); 
-    }
-    else if (1 == p_config->xmloutput){
-      fprintf(config.outfile, "%s", uloc_getAvailable(i)); 
-      fprintf(config.outfile, " | ");
-      if (strlen(name_str))
-        fprintf(config.outfile, "%s", name_str);
-      fprintf(config.outfile, " | ");
-      if (strlen(localname_str))
-        fprintf(config.outfile, "%s", localname_str);
-      fprintf(config.outfile, "\n");
+        language_len 
+            = uloc_getDisplayLanguage(uloc_getAvailable(i), "en", 
+                                      language, 64, 
+                                      &status);
+
+        u_strToUTF8(lang_str, 128, &lang_str_len,
+                    language, language_len,
+                    &status);
+
+
+        script_len 
+            = uloc_getDisplayScript(uloc_getAvailable(i), "en", 
+                                    script, 64, 
+                                    &status);
+
+        u_strToUTF8(script_str, 128, &script_str_len,
+                    script, script_len,
+                    &status);
+
+        location_len 
+            = uloc_getDisplayCountry(uloc_getAvailable(i), "en", 
+                                     location, 64, 
+                                     &status);
+
+        u_strToUTF8(location_str, 128, &location_str_len,
+                    location, location_len,
+                    &status);
+
+        variant_len 
+            = uloc_getDisplayVariant(uloc_getAvailable(i), "en", 
+                                     variant, 64, 
+                                     &status);
+
+        u_strToUTF8(variant_str, 128, &variant_str_len,
+                    variant, variant_len,
+                    &status);
+
+        name_len 
+            = uloc_getDisplayName(uloc_getAvailable(i), "en", 
+                                  name, 64, 
+                                  &status);
+
+        u_strToUTF8(name_str, 128, &name_str_len,
+                    name, name_len,
+                    &status);
+
+        localname_len 
+            = uloc_getDisplayName(uloc_getAvailable(i), uloc_getAvailable(i), 
+                                  localname, 64, 
+                                  &status);
+
+        u_strToUTF8(localname_str, 128, &localname_str_len,
+                    localname, localname_len,
+                    &status);
+
+
+        if (p_config->xmloutput){
+            fprintf(config.outfile, "<locale id=\"%s\"", uloc_getAvailable(i)); 
+            /* fprintf(config.outfile, " locale=\"%s\"", uloc_getAvailable(i)); */
+            /* if (strlen(keyword_str)) */
+            /*   fprintf(config.outfile, " keyword=\"%s\"", keyword_str); */
+            /* if (ucol_getAvailable(i)) */
+            /*   fprintf(config.outfile, " collation=\"1\""); */
+            if (strlen(lang_str))
+                fprintf(config.outfile, " language=\"%s\"", lang_str);
+            if (strlen(script_str))
+                fprintf(config.outfile, " script=\"%s\"", script_str);
+            if (strlen(location_str))
+                fprintf(config.outfile, " location=\"%s\"", location_str);
+            if (strlen(variant_str))
+                fprintf(config.outfile, " variant=\"%s\"", variant_str);
+            if (strlen(name_str))
+                fprintf(config.outfile, " name=\"%s\"", name_str);
+            if (strlen(localname_str))
+                fprintf(config.outfile, " localname=\"%s\"", localname_str);
+            fprintf(config.outfile, ">");
+            if (strlen(localname_str))
+                fprintf(config.outfile, "%s", localname_str);
+            fprintf(config.outfile, "</locale>\n"); 
+        }
+        else if (1 == p_config->xmloutput){
+            fprintf(config.outfile, "%s", uloc_getAvailable(i)); 
+            fprintf(config.outfile, " | ");
+            if (strlen(name_str))
+                fprintf(config.outfile, "%s", name_str);
+            fprintf(config.outfile, " | ");
+            if (strlen(localname_str))
+                fprintf(config.outfile, "%s", localname_str);
+            fprintf(config.outfile, "\n");
+        }
+        else
+            fprintf(config.outfile, "%s ", uloc_getAvailable(i));
     }
+    if (p_config->xmloutput)
+        fprintf(config.outfile, "</locales>\n");
     else
-      fprintf(config.outfile, "%s ", uloc_getAvailable(i));
-  }
-  if (p_config->xmloutput)
-    fprintf(config.outfile, "</locales>\n");
-  else
-    fprintf(config.outfile, "\n");
-
-  if(U_FAILURE(status)) {
-    fprintf(stderr, "ICU Error: %d %s\n", status, u_errorName(status));
-    exit(status);
-  }
+        fprintf(config.outfile, "\n");
+
+    if(U_FAILURE(status)) {
+        fprintf(stderr, "ICU Error: %d %s\n", status, u_errorName(status));
+        exit(status);
+    }
 }
 
 
 static void print_info(const struct config_t *p_config)
 {
-  if (p_config->xmloutput)
-    fprintf(config.outfile, "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
-           "<icu>\n");
+    if (p_config->xmloutput)
+        fprintf(config.outfile, "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
+                "<icu>\n");
 
     if ('c' == config.print[0])
         print_icu_converters(&config);
@@ -402,16 +421,15 @@ static void print_info(const struct config_t *p_config)
     if (p_config->xmloutput)
         fprintf(config.outfile, "</icu>\n");
 
-  exit(0);
+    exit(0);
 };
 
 
 
 static void process_text_file(const struct config_t *p_config)
 {
-    char * line = 0;
-    size_t line_cap = 0;
-    ssize_t line_len;
+    char *line = 0;
+    char linebuf[1024];
  
     xmlDoc *doc = xmlParseFile(config.conffile);  
     xmlNode *xml_node = xmlDocGetRootElement(doc);
@@ -435,7 +453,8 @@ static void process_text_file(const struct config_t *p_config)
                 "<tokens>\n");
     
     // read input lines for processing
-    while ((line_len = getline(&line, &line_cap, config.infile)) != -1) {
+    while ((line=fgets(linebuf, sizeof(linebuf)-1, config.infile)))
+    {
         success = icu_chain_assign_cstr(config.chain, line, &status);
         line_count++;
 
@@ -463,7 +482,7 @@ static void process_text_file(const struct config_t *p_config)
         
     }
 
-   if (p_config->xmloutput)
+    if (p_config->xmloutput)
         fprintf(config.outfile, 
                 "</tokens>\n"
                 "</icu>\n");
@@ -474,10 +493,14 @@ static void process_text_file(const struct config_t *p_config)
         free(line);
 };
 
+#endif // HAVE_ICU
+
 
 int main(int argc, char **argv) 
 {
 
+#ifdef HAVE_ICU
+
     read_params(argc, argv, &config);
 
     if (config.conffile && strlen(config.conffile))
@@ -485,8 +508,16 @@ int main(int argc, char **argv)
      
     if (config.print && strlen(config.print))
         print_info(&config);
-    
-    
+
+#else // HAVE_ICU
+
+    printf("ICU not available on your system.\n"
+           "Please install libicu36-dev and icu-doc or similar, "
+           "re-configure and re-compile\n");
+
+
+#endif // HAVE_ICU
+
     return(0);
 };