boxtec
/
yacy_search_server
mirror of https://github.com/yacy/yacy_search_server
4
1
Fork 0
Code Issues Projects Releases Wiki Activity

fix for deadlocks caused by self-blocking access to TreeMap in concurrent environments. The TreeMap was replaced by a ConcurrentHashMap and additional care that the strings are compared all in lowercase

Browse Source 
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6828 6c8d7289-2bf4-0310-a012-ef5d649a1542
pull/1/head
orbiter 15 years ago
parent 027b971bde
commit 3247f0e901
1 changed files with 24 additions and 30 deletions
Show Stats Download Patch File Download Diff File

54
source/net/yacy/document/TextParser.java
Unescape View File

@ -32,16 +32,13 @@ import java.io.File;
import java.io.FileInputStream; import java.io.FileInputStream;
import java.io.IOException; import java.io.IOException;
import java.io.InputStream; import java.io.InputStream;
import java.text.Collator;
import java.util.ArrayList; import java.util.ArrayList;
import java.util.HashMap; import java.util.HashMap;
import java.util.HashSet; import java.util.HashSet;
import java.util.List; import java.util.List;
import java.util.Locale;
import java.util.Map; import java.util.Map;
import java.util.Set; import java.util.Set;
import java.util.TreeMap; import java.util.concurrent.ConcurrentHashMap;
import java.util.TreeSet;
import net.yacy.document.parser.bzipParser; import net.yacy.document.parser.bzipParser;
import net.yacy.document.parser.csvParser; import net.yacy.document.parser.csvParser;
@ -73,18 +70,11 @@ public final class TextParser {
private static final Log log = new Log("PARSER"); private static final Log log = new Log("PARSER");
// use a collator to relax when distinguishing between lowercase und uppercase letters private static final Map<String, Idiom> mime2parser = new ConcurrentHashMap<String, Idiom>();
private static final Collator insensitiveCollator = Collator.getInstance(Locale.US); private static final Map<String, Idiom> ext2parser = new ConcurrentHashMap<String, Idiom>();
static { private static final Map<String, String> ext2mime = new ConcurrentHashMap<String, String>();
insensitiveCollator.setStrength(Collator.SECONDARY); private static final Map<String, Object> denyMime = new ConcurrentHashMap<String, Object>();
insensitiveCollator.setDecomposition(Collator.NO_DECOMPOSITION); private static final Map<String, Object> denyExtensionx = new ConcurrentHashMap<String, Object>();
}
private static final Map<String, Idiom> mime2parser = new TreeMap<String, Idiom>(insensitiveCollator);
private static final Map<String, Idiom> ext2parser = new TreeMap<String, Idiom>(insensitiveCollator);
private static final Map<String, String> ext2mime = new TreeMap<String, String>(insensitiveCollator);
private static final Set<String> denyMime = new TreeSet<String>(insensitiveCollator);
private static final Set<String> denyExtension = new TreeSet<String>(insensitiveCollator);
static { static {
initParser(new bzipParser()); initParser(new bzipParser());
@ -130,6 +120,7 @@ public final class TextParser {
} }
if (prototypeMime != null) for (String ext: parser.supportedExtensions()) { if (prototypeMime != null) for (String ext: parser.supportedExtensions()) {
ext = ext.toLowerCase();
String s = ext2mime.get(ext); String s = ext2mime.get(ext);
if (s != null) log.logSevere("parser for extension '" + ext + "' was set to mime '" + s + "', overwriting with new mime '" + prototypeMime + "'."); if (s != null) log.logSevere("parser for extension '" + ext + "' was set to mime '" + s + "', overwriting with new mime '" + prototypeMime + "'.");
ext2mime.put(ext, prototypeMime); ext2mime.put(ext, prototypeMime);
@ -137,6 +128,7 @@ public final class TextParser {
for (String ext: parser.supportedExtensions()) { for (String ext: parser.supportedExtensions()) {
// process the extensions // process the extensions
ext = ext.toLowerCase();
Idiom p0 = ext2parser.get(ext); Idiom p0 = ext2parser.get(ext);
if (p0 != null) log.logSevere("parser for extension '" + ext + "' was set to '" + p0.getName() + "', overwriting with new parser '" + parser.getName() + "'."); if (p0 != null) log.logSevere("parser for extension '" + ext + "' was set to '" + p0.getName() + "', overwriting with new parser '" + parser.getName() + "'.");
ext2parser.put(ext, parser); ext2parser.put(ext, parser);
@ -318,7 +310,8 @@ public final class TextParser {
String ext = url.getFileExtension(); String ext = url.getFileExtension();
Idiom idiom; Idiom idiom;
if (ext != null && ext.length() > 0) { if (ext != null && ext.length() > 0) {
if (denyExtension.contains(ext)) throw new ParserException("file extension '" + ext + "' is denied (1)", url); ext = ext.toLowerCase();
if (denyExtensionx.containsKey(ext)) throw new ParserException("file extension '" + ext + "' is denied (1)", url);
idiom = ext2parser.get(ext); idiom = ext2parser.get(ext);
if (idiom != null) idioms.add(idiom); if (idiom != null) idioms.add(idiom);
} }
@ -326,14 +319,14 @@ public final class TextParser {
// check given mime type // check given mime type
if (mimeType1 != null) { if (mimeType1 != null) {
mimeType1 = normalizeMimeType(mimeType1); mimeType1 = normalizeMimeType(mimeType1);
if (denyMime.contains(mimeType1)) throw new ParserException("mime type '" + mimeType1 + "' is denied (1)", url); if (denyMime.containsKey(mimeType1)) throw new ParserException("mime type '" + mimeType1 + "' is denied (1)", url);
idiom = mime2parser.get(mimeType1); idiom = mime2parser.get(mimeType1);
if (idiom != null && !idioms.contains(idiom)) idioms.add(idiom); if (idiom != null && !idioms.contains(idiom)) idioms.add(idiom);
} }
// check mime type computed from extension // check mime type computed from extension
String mimeType2 = ext2mime.get(ext); String mimeType2 = ext2mime.get(ext);
if (mimeType2 == null || denyMime.contains(mimeType2)) return idioms; // in this case we are a bit more lazy if (mimeType2 == null || denyMime.containsKey(mimeType2)) return idioms; // in this case we are a bit more lazy
idiom = mime2parser.get(mimeType2); idiom = mime2parser.get(mimeType2);
if (idiom != null && !idioms.contains(idiom)) idioms.add(idiom); if (idiom != null && !idioms.contains(idiom)) idioms.add(idiom);
@ -346,15 +339,15 @@ public final class TextParser {
public static String supportsMime(String mimeType) { public static String supportsMime(String mimeType) {
if (mimeType == null) return null; if (mimeType == null) return null;
mimeType = normalizeMimeType(mimeType); mimeType = normalizeMimeType(mimeType);
if (denyMime.contains(mimeType)) return "mime type '" + mimeType + "' is denied (2)"; if (denyMime.containsKey(mimeType)) return "mime type '" + mimeType + "' is denied (2)";
if (mime2parser.get(mimeType) == null) return "no parser for mime '" + mimeType + "' available"; if (mime2parser.get(mimeType) == null) return "no parser for mime '" + mimeType + "' available";
return null; return null;
} }
public static String supportsExtension(final DigestURI url) { public static String supportsExtension(final DigestURI url) {
String ext = url.getFileExtension(); String ext = url.getFileExtension().toLowerCase();
if (ext == null || ext.length() == 0) return null; if (ext == null || ext.length() == 0) return null;
if (denyExtension.contains(ext)) return "file extension '" + ext + "' is denied (2)"; if (denyExtensionx.containsKey(ext)) return "file extension '" + ext + "' is denied (2)";
String mimeType = ext2mime.get(ext); String mimeType = ext2mime.get(ext);
if (mimeType == null) return "no parser available"; if (mimeType == null) return "no parser available";
Idiom idiom = mime2parser.get(mimeType); Idiom idiom = mime2parser.get(mimeType);
@ -368,11 +361,12 @@ public final class TextParser {
} }
public static String mimeOf(String ext) { public static String mimeOf(String ext) {
return ext2mime.get(ext); return ext2mime.get(ext.toLowerCase());
} }
private static String normalizeMimeType(String mimeType) { private static String normalizeMimeType(String mimeType) {
if (mimeType == null) return "application/octet-stream"; if (mimeType == null) return "application/octet-stream";
mimeType = mimeType.toLowerCase();
final int pos = mimeType.indexOf(';'); final int pos = mimeType.indexOf(';');
return ((pos < 0) ? mimeType.trim() : mimeType.substring(0, pos).trim()); return ((pos < 0) ? mimeType.trim() : mimeType.substring(0, pos).trim());
} }
@ -382,13 +376,13 @@ public final class TextParser {
String n; String n;
for (String s: denyList.split(",")) { for (String s: denyList.split(",")) {
n = normalizeMimeType(s); n = normalizeMimeType(s);
if (n != null && n.length() > 0) denyMime.add(n); if (n != null && n.length() > 0) denyMime.put(n, null);
} }
} }
public static String getDenyMime() { public static String getDenyMime() {
String s = ""; String s = "";
for (String d: denyMime) s += d + ","; for (String d: denyMime.keySet()) s += d + ",";
if (s.length() > 0) s = s.substring(0, s.length() - 1); if (s.length() > 0) s = s.substring(0, s.length() - 1);
return s; return s;
} }
@ -396,22 +390,22 @@ public final class TextParser {
public static void grantMime(String mime, boolean grant) { public static void grantMime(String mime, boolean grant) {
String n = normalizeMimeType(mime); String n = normalizeMimeType(mime);
if (n == null || n.length() == 0) return; if (n == null || n.length() == 0) return;
if (grant) denyMime.remove(n); else denyMime.add(n); if (grant) denyMime.remove(n); else denyMime.put(n, null);
} }
public static void setDenyExtension(String denyList) { public static void setDenyExtension(String denyList) {
denyExtension.clear(); denyExtensionx.clear();
for (String s: denyList.split(",")) denyExtension.add(s); for (String s: denyList.split(",")) denyExtensionx.put(s, null);
} }
public static String getDenyExtension() { public static String getDenyExtension() {
String s = ""; String s = "";
for (String d: denyExtension) s += d + ","; for (String d: denyExtensionx.keySet()) s += d + ",";
s = s.substring(0, s.length() - 1); s = s.substring(0, s.length() - 1);
return s; return s;
} }
public static void grantExtension(String ext, boolean grant) { public static void grantExtension(String ext, boolean grant) {
if (grant) denyExtension.remove(ext); else denyExtension.add(ext); if (grant) denyExtensionx.remove(ext); else denyExtensionx.put(ext, null);
} }
} }

Write Preview
Loading…
Cancel
Save