Changeset 21403 for main/trunk/greenstone2/common-src/src/jdbmedit/Txt2Jdb.java

Timestamp:

2010-01-04T17:40:03+13:00 (14 years ago)

Author:

davidb

Message:

Code was working for Ascii characters, but not for Unicode values > 128. More careful prescription of utf-8 used in code

File:

• main/trunk/greenstone2/common-src/src/jdbmedit/Txt2Jdb.java (modified) (9 diffs)

main/trunk/greenstone2/common-src/src/jdbmedit/Txt2Jdb.java

@@ -26 +26 @@
 import java.io.BufferedInputStream;
 import java.io.InputStream;
+import java.io.IOException;
+import java.io.InputStreamReader;
+import java.io.BufferedReader;
+import java.util.Properties;
 
 import jdbm.RecordManager;
@@ -32 +36 @@
 import jdbm.htree.HTree;
 
-import java.io.IOException;
-import java.util.Properties;
-
 
 public class Txt2Jdb
@@ -47 +48 @@
     throws IOException
     {
+    if (db_filename.endsWith(".jdb")) {
+        // remove file extension as JDBM does not expect it
+        db_filename = db_filename.substring(0,db_filename.length()-4);
+    }
+
         // create or open a record manager
         Properties props = new Properties();
@@ -75 +81 @@
     // Assumes InputStream is a file or standard-in  
     // (no some other form of IO mapped device)
-    public boolean eof(InputStream is) 
+    //public boolean eof(InputStream is) 
+    public boolean eof(int c) 
     throws IOException
     {
-    return (is.available() == 0);
+    return (c == -1);
     }
 
@@ -95 +102 @@
     boolean delkey = false;
 
-    BufferedInputStream bis = new BufferedInputStream(System.in);
-    
+
+    InputStreamReader isr = new InputStreamReader(System.in,"UTF-8");
+    BufferedReader bis = new BufferedReader(isr);
+
     int c = bis.read();
-    while (!eof(bis)) {
+    while (!eof(c)) {
         num_dashes = 0;
         key = new StringBuffer();
@@ -106 +115 @@
         
         // scan for first occurrence of [
-        while (!eof(bis) && c != '[') {
-        c = bis.read();
-        }
-
-        
-        if (!eof(bis)) {
+        while (!eof(c) && c != '[') {
+        c = bis.read();
+        }
+
+        
+        if (!eof(c)) {
         c = bis.read(); // skip [
         }
@@ -117 +126 @@
         
         // now look for closing ], building up 'key' as we go
-        while (!eof(bis) && c != ']') {
+        while (!eof(c) && c != ']') {
         key.append ((char)c);
         c = bis.read();
         }
         
-        if (!eof(bis)) {
+        if (!eof(c)) {
         // most likely an eol char, but if '-', then signifies record
         // is to be deleted, not added
@@ -134 +143 @@
         }
 
-        while (!eof(bis) && (c == '\n' || c == '\r')) {
+        while (!eof(c) && (c == '\n' || c == '\r')) {
         c = bis.read();
         }
@@ -140 +149 @@
         // look for 70 dashes
         tmp = new StringBuffer();
-        while (!eof(bis) && (num_dashes < 70)) {
+        while (!eof(c) && (num_dashes < 70)) {
         if (c == '\n') {
             tmp.append((char)c);