Apache
/
poi
mirror of https://github.com/apache/poi.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

try to fix tests 

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1903783 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
PJ Fanning 2022-08-30 21:43:43 +00:00
parent a1cb1812dd
commit ecfb35e0d9
3 changed files with 24 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
poi/src/main/java/org/apache/poi/poifs/filesystem/Ole10Native.java
View File

@ -276,7 +276,7 @@ public class Ole10Native {
byte[] buf = new byte[MAX_STRING_LENGTH];
for (int i=0; i<buf.length; i++) {
if ((buf[i] = is.readByte()) == 0) {
return StringUtil.getFromCompressedUnicode(buf, 0, i);
return StringUtil.getFromCompressedUTF8(buf, 0, i);
}
}
throw new Ole10NativeException("AsciiZ string was not null terminated after " + MAX_STRING_LENGTH + " bytes - Exiting.");

2
poi/src/main/java/org/apache/poi/ss/formula/ptg/StringPtg.java
View File

@ -48,7 +48,7 @@ public final class StringPtg extends ScalarConstantPtg {
if (_is16bitUnicode) {
field_3_string = StringUtil.readUnicodeLE(in, nChars);
} else {
field_3_string = StringUtil.readCompressedLatinA(in, nChars);
field_3_string = StringUtil.readCompressedUnicode(in, nChars);
}
}

36
poi/src/main/java/org/apache/poi/util/StringUtil.java
View File

@ -18,6 +18,7 @@
package org.apache.poi.util;
import static java.nio.charset.StandardCharsets.ISO_8859_1;
import static java.nio.charset.StandardCharsets.UTF_8;
import java.nio.charset.Charset;
import java.nio.charset.StandardCharsets;
@ -128,33 +129,40 @@ public final class StringUtil {
* @param string byte array to read
* @param offset offset to read byte array
* @param len length to read byte array
* @return String generated String instance by reading byte array
* @return String generated String instance by reading byte array (ISO-8859-1)
*/
public static String getFromCompressedUnicode(
final byte[] string,
final int offset,
final int len) {
int len_to_use = Math.min(len, string.length - offset);
return new String(string, offset, len_to_use, UTF8);
return new String(string, offset, len_to_use, ISO_8859_1);
}
/**
* Read 8 bit data (in UTF-8 codepage) into a (unicode) Java
* String and return.
* (In Excel terms, read compressed 8 bit unicode as a string)
*
* @param string byte array to read
* @param offset offset to read byte array
* @param len length to read byte array
* @return String generated String instance by reading byte array (UTF-8)
*/
public static String getFromCompressedUTF8(
final byte[] string,
final int offset,
final int len) {
int len_to_use = Math.min(len, string.length - offset);
return new String(string, offset, len_to_use, UTF_8);
}
/**
* @param in stream,
* @param nChars number pf chars
* @return UTF-8 encoded result
* @return ISO_8859_1 encoded result
*/
public static String readCompressedUnicode(LittleEndianInput in, int nChars) {
byte[] buf = IOUtils.safelyAllocate(nChars, MAX_RECORD_LENGTH);
in.readFully(buf);
return new String(buf, UTF8);
}
/**
* @param in stream,
* @param nChars number pf chars
* @return LATIN-A (ISO-8859-1) encoded result
*/
public static String readCompressedLatinA(LittleEndianInput in, int nChars) {
byte[] buf = IOUtils.safelyAllocate(nChars, MAX_RECORD_LENGTH);
in.readFully(buf);
return new String(buf, ISO_8859_1);