JCNC/JNotepad
JCNC/JNotepad
2
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

!29 fix: #I7VP0I 检测文件失败时，使用系统默认编码

Browse Source 
Merge pull request !29 from songdragon/fix-I7VP0I
This commit is contained in:
Luke 2023-08-24 15:18:09 +00:00 committed by Gitee
commit f950b42450
No known key found for this signature in database
GPG Key ID: 173E9B9CA92EEF8F
1 changed files with 18 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
src/main/java/org/jcnc/jnotepad/tool/EncodingDetector.java
View File

@ -23,6 +23,10 @@ import static org.jcnc.jnotepad.constants.TextConstants.UNKNOWN;
public class EncodingDetector { public class EncodingDetector {
private static final Logger LOG = LogUtil.getLogger(EncodingDetector.class); private static final Logger LOG = LogUtil.getLogger(EncodingDetector.class);
/**
* 编码侦测概率，阈值：50%
*/
public static final int THRESHOLD_CONFIDENCE = 50;
private EncodingDetector() { private EncodingDetector() {
@ -39,10 +43,20 @@ public class EncodingDetector {
try (BufferedInputStream inputStream = new BufferedInputStream(new FileInputStream(file.getPath()))) { try (BufferedInputStream inputStream = new BufferedInputStream(new FileInputStream(file.getPath()))) {
charsetDetector.setText(inputStream); charsetDetector.setText(inputStream);
CharsetMatch[] matchList = charsetDetector.detectAll(); CharsetMatch[] matchList = charsetDetector.detectAll();
for (CharsetMatch match : matchList) { if (matchList == null || matchList.length == 0) {
return UNKNOWN;
}
CharsetMatch maxConfidence = matchList[0];
if (maxConfidence.getConfidence() < THRESHOLD_CONFIDENCE) {
return UNKNOWN;
}
for (int i = 1; i < matchList.length; i++) {
CharsetMatch match = matchList[i];
LOG.debug("{} : {}", match.getName(), match.getConfidence()); LOG.debug("{} : {}", match.getName(), match.getConfidence());
if (match.getConfidence() > 50) { if (match.getConfidence() >= THRESHOLD_CONFIDENCE && match.getConfidence() >= maxConfidence.getConfidence()) {
return match.getName(); maxConfidence = match;
} else {
return maxConfidence.getName();
} }
} }
} catch (Exception e) { } catch (Exception e) {
@ -60,7 +74,7 @@ public class EncodingDetector {
public static Charset detectEncodingCharset(File file) { public static Charset detectEncodingCharset(File file) {
String charset = detectEncoding(file); String charset = detectEncoding(file);
if (charset.equals(UNKNOWN)) { if (charset.equals(UNKNOWN)) {
return StandardCharsets.UTF_8; return Charset.defaultCharset();
} }
return Charset.forName(charset); return Charset.forName(charset);
} }