void setConfig(Configuration config) {
log.debug("config: " + config);
proxyHost = config.get(PARAM_PROXY_HOST);
proxyPort = config.getInt(PARAM_PROXY_PORT, DEFAULT_PROXY_PORT);
if (StringUtil.isNullString(proxyHost) || proxyPort <= 0) {
String http_proxy = System.getenv("http_proxy");
if (!StringUtil.isNullString(http_proxy)) {
try {
HostPortParser hpp = new HostPortParser(http_proxy);
proxyHost = hpp.getHost();
proxyPort = hpp.getPort();
} catch (HostPortParser.InvalidSpec e) {
log.warning("Can't parse http_proxy environment var, ignoring: " + http_proxy + ": " + e);
}
}
}
if (StringUtil.isNullString(proxyHost) || proxyPort <= 0) {
proxyHost = null;
} else {
log.info("Proxying through " + proxyHost + ":" + proxyPort);
}
userAgent = config.get(PARAM_USER_AGENT);
if (StringUtil.isNullString(userAgent)) {
userAgent = null;
} else {
log.debug("Setting User-Agent to " + userAgent);
}
}
void checkParamAgreement(String key, PrintfContext context) {
List<String> printfList = getElementList(key);
if (printfList == null) {
return;
}
for (String printf : printfList) {
if (StringUtil.isNullString(printf)) {
log.warning("Null printf string in " + key);
continue;
}
PrintfUtil.PrintfData p_data = PrintfUtil.stringToPrintf(printf);
Collection<String> p_args = p_data.getArguments();
for (String arg : p_args) {
ConfigParamDescr descr = findAuConfigDescr(arg);
if (descr == null) {
throw new PluginException.InvalidDefinition(
"Not a declared parameter: " + arg + " in " + printf + " in " + getPluginName());
}
// ensure range and set params used only in legal context
switch (context) {
case Regexp:
case Display:
// everything is legal in a regexp or a display string
break;
case URL:
// NUM_RANGE and SET legal because can enumerate. Can't
// enumerate RANGE
switch (descr.getType()) {
case ConfigParamDescr.TYPE_RANGE:
throw new PluginException.InvalidDefinition(
"Range parameter ("
+ arg
+ ") used in illegal context in "
+ getPluginName()
+ ": "
+ key
+ ": "
+ printf);
default:
}
}
}
}
}
/** If in testing mode FOO, copy values from FOO_override map, if any, to main map */
void processOverrides(TypedEntryMap map) {
String testMode = getTestingMode();
if (StringUtil.isNullString(testMode)) {
return;
}
Object o = map.getMapElement(testMode + DefinableArchivalUnit.SUFFIX_OVERRIDE);
if (o == null) {
return;
}
if (o instanceof Map) {
Map overrideMap = (Map) o;
for (Map.Entry entry : (Set<Map.Entry>) overrideMap.entrySet()) {
String key = (String) entry.getKey();
Object val = entry.getValue();
log.debug(getDefaultPluginName() + ": Overriding " + key + " with " + val);
map.setMapElement(key, val);
}
}
}
protected void initResultMap() throws PluginException.InvalidDefinition {
HttpResultMap hResultMap = new HttpResultMap();
// XXX Currently this only allows a CacheResultHandler class to
// initialize the result map. Instead, don't use a CacheResultMap
// directly, use either the plugin's CacheResultHandler, if specified,
// or a default one that wraps the CacheResultMap
String handler_class = null;
handler_class = definitionMap.getString(KEY_EXCEPTION_HANDLER, null);
if (handler_class != null) {
try {
resultHandler = (CacheResultHandler) newAuxClass(handler_class, CacheResultHandler.class);
resultHandler.init(hResultMap);
} catch (Exception ex) {
throw new PluginException.InvalidDefinition(
mapName + " has invalid Exception handler: " + handler_class, ex);
} catch (LinkageError le) {
throw new PluginException.InvalidDefinition(
mapName + " has invalid Exception handler: " + handler_class, le);
}
} else {
// Expect a list of mappings from either result code or exception
// name to CacheException name
Collection<String> mappings = definitionMap.getCollection(KEY_EXCEPTION_LIST, null);
if (mappings != null) {
// add each entry
for (String entry : mappings) {
if (log.isDebug2()) {
log.debug2("initMap(" + entry + ")");
}
String first;
String ceName;
try {
List<String> pair = StringUtil.breakAt(entry, '=', 2, true, true);
first = pair.get(0);
ceName = pair.get(1);
} catch (Exception ex) {
throw new PluginException.InvalidDefinition(
"Invalid syntax: " + entry + "in " + mapName);
}
Object val;
// Value should be either a CacheException or CacheResultHandler
// class name.
PluginFetchEventResponse resp =
(PluginFetchEventResponse) newAuxClass(ceName, PluginFetchEventResponse.class, null);
if (resp instanceof CacheException) {
val = resp.getClass();
} else if (resp instanceof CacheResultHandler) {
val = WrapperUtil.wrap((CacheResultHandler) resp, CacheResultHandler.class);
} else {
throw new PluginException.InvalidDefinition(
"Second arg not a "
+ "CacheException or "
+ "CacheResultHandler class: "
+ entry
+ ", in "
+ mapName);
}
try {
int code = Integer.parseInt(first);
// If parseable as an integer, it's a result code.
hResultMap.storeMapEntry(code, val);
} catch (NumberFormatException e) {
try {
Class eClass = Class.forName(first);
// If a class name, it should be an exception class
if (Exception.class.isAssignableFrom(eClass)) {
hResultMap.storeMapEntry(eClass, val);
} else {
throw new PluginException.InvalidDefinition(
"First arg not an " + "Exception class: " + entry + ", in " + mapName);
}
} catch (Exception ex) {
throw new PluginException.InvalidDefinition(
"First arg not a " + "number or class: " + entry + ", in " + mapName);
} catch (LinkageError le) {
throw new PluginException.InvalidDefinition("Can't load " + first, le);
}
}
}
}
}
resultMap = hResultMap;
}
protected void initMimeMap() throws PluginException.InvalidDefinition {
for (Iterator iter = definitionMap.entrySet().iterator(); iter.hasNext(); ) {
Map.Entry ent = (Map.Entry) iter.next();
String key = (String) ent.getKey();
Object val = ent.getValue();
if (key.endsWith(DefinableArchivalUnit.SUFFIX_LINK_EXTRACTOR_FACTORY)) {
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_LINK_EXTRACTOR_FACTORY);
if (val instanceof String) {
String factName = (String) val;
log.debug(mime + " link extractor: " + factName);
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
LinkExtractorFactory fact =
(LinkExtractorFactory) newAuxClass(factName, LinkExtractorFactory.class);
mti.setLinkExtractorFactory(fact);
}
} else if (key.endsWith(DefinableArchivalUnit.SUFFIX_CRAWL_FILTER_FACTORY)) {
// XXX This clause must precede the one for SUFFIX_HASH_FILTER_FACTORY
// XXX unless/until that key is changed to not be a terminal substring
// XXX of this one
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_CRAWL_FILTER_FACTORY);
if (val instanceof String) {
String factName = (String) val;
log.debug(mime + " crawl filter: " + factName);
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
FilterFactory fact = (FilterFactory) newAuxClass(factName, FilterFactory.class);
mti.setCrawlFilterFactory(fact);
}
} else if (key.endsWith(DefinableArchivalUnit.SUFFIX_HASH_FILTER_FACTORY)) {
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_HASH_FILTER_FACTORY);
if (val instanceof String) {
String factName = (String) val;
log.debug(mime + " filter: " + factName);
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
FilterFactory fact = (FilterFactory) newAuxClass(factName, FilterFactory.class);
mti.setHashFilterFactory(fact);
}
} else if (key.endsWith(DefinableArchivalUnit.SUFFIX_FETCH_RATE_LIMIT)) {
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_FETCH_RATE_LIMIT);
if (val instanceof String) {
String rate = (String) val;
log.debug(mime + " fetch rate: " + rate);
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
RateLimiter limit = mti.getFetchRateLimiter();
if (limit != null) {
limit.setRate(rate);
} else {
mti.setFetchRateLimiter(new RateLimiter(rate));
}
}
} else if (key.endsWith(DefinableArchivalUnit.SUFFIX_LINK_REWRITER_FACTORY)) {
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_LINK_REWRITER_FACTORY);
String factName = (String) val;
log.debug(mime + " link rewriter: " + factName);
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
LinkRewriterFactory fact =
(LinkRewriterFactory) newAuxClass(factName, LinkRewriterFactory.class);
mti.setLinkRewriterFactory(fact);
} else if (key.endsWith(DefinableArchivalUnit.SUFFIX_METADATA_EXTRACTOR_FACTORY_MAP)) {
String mime = stripSuffix(key, DefinableArchivalUnit.SUFFIX_METADATA_EXTRACTOR_FACTORY_MAP);
Map factNameMap = (Map) val;
Map factClassMap = new HashMap();
MimeTypeInfo.Mutable mti = mimeMap.modifyMimeTypeInfo(mime);
for (Iterator it = factNameMap.keySet().iterator(); it.hasNext(); ) {
String mdTypes = (String) it.next();
String factName = (String) factNameMap.get(mdTypes);
log.debug(mime + " (" + mdTypes + ") metadata extractor: " + factName);
for (String mdType : (List<String>) StringUtil.breakAt(mdTypes, ";")) {
setMdTypeFact(factClassMap, mdType, factName);
}
}
mti.setFileMetadataExtractorFactoryMap(factClassMap);
}
}
}
protected String getDefaultPluginName() {
return StringUtil.shortName(getPluginId());
}
/** Explode the archive into its constituent elements */
public void explode() throws CacheException {
CachedUrl cachedUrl = null;
int goodEntries = 0;
int badEntries = 0;
int ignoredEntries = 0;
int entriesBetweenSleep = 0;
ArchiveReader arcReader = null;
logger.info(
(storeArchive ? "Storing" : "Fetching") + " WARC file: " + origUrl + " will explode");
try {
if (storeArchive) {
UrlCacher uc = au.makeUrlCacher(new UrlData(arcStream, arcProps, fetchUrl));
BitSet bs = new BitSet();
bs.set(UrlCacher.DONT_CLOSE_INPUT_STREAM_FLAG);
uc.setFetchFlags(bs);
uc.storeContent();
archiveData.resetInputStream();
arcStream = archiveData.input;
}
// Wrap it in an ArchiveReader
logger.debug3("About to wrap stream");
arcReader = wrapStream(fetchUrl, arcStream);
logger.debug3("wrapStream() returns " + (arcReader == null ? "null" : "non-null"));
// Explode it
if (arcReader == null) {
throw new CacheException.ExploderException("no WarcReader for " + origUrl);
}
ArchivalUnit au = crawlFacade.getAu();
Set stemSet = new HashSet();
logger.debug("Exploding " + fetchUrl);
// Iterate through the elements in the WARC file, except the first
Iterator i = arcReader.iterator();
// Skip first record
for (i.next(); i.hasNext(); ) {
// XXX probably not necessary
helper.pokeWDog();
if ((++entriesBetweenSleep % sleepAfter) == 0) {
long pauseTime =
CurrentConfig.getTimeIntervalParam(PARAM_RETRY_PAUSE, DEFAULT_RETRY_PAUSE);
Deadline pause = Deadline.in(pauseTime);
logger.debug3("Sleeping for " + StringUtil.timeIntervalToString(pauseTime));
while (!pause.expired()) {
try {
pause.sleep();
} catch (InterruptedException ie) {
// no action
}
}
}
ArchiveRecord element = (ArchiveRecord) i.next();
// Each element is a URL to be cached in a suitable AU
ArchiveRecordHeader elementHeader = element.getHeader();
String elementUrl = elementHeader.getUrl();
String elementMimeType = elementHeader.getMimetype();
long elementLength = elementHeader.getLength();
logger.debug2("WARC url " + elementUrl + " mime " + elementMimeType);
if (elementUrl.startsWith("http:")) {
ArchiveEntry ae =
new ArchiveEntry(
elementUrl,
elementLength,
0, // XXX need to convert getDate string to long
element, // ArchiveRecord extends InputStream
this,
fetchUrl);
ae.setHeaderFields(makeCIProperties(elementHeader));
long bytesStored = elementLength;
logger.debug3("ArchiveEntry: " + ae.getName() + " bytes " + bytesStored);
try {
helper.process(ae);
} catch (PluginException ex) {
throw new CacheException.ExploderException("helper.process() threw", ex);
}
if (ae.getBaseUrl() != null) {
if (ae.getRestOfUrl() != null && ae.getHeaderFields() != null) {
storeEntry(ae);
handleAddText(ae);
goodEntries++;
crawlFacade.getCrawlerStatus().addContentBytesFetched(bytesStored);
} else {
ignoredEntries++;
}
} else {
badEntries++;
logger.debug2("Can't map " + elementUrl + " from " + archiveUrl);
}
}
}
} catch (IOException ex) {
throw new CacheException.ExploderException(ex);
} finally {
if (arcReader != null)
try {
arcReader.close();
arcReader = null;
} catch (IOException ex) {
throw new CacheException.ExploderException(ex);
}
if (cachedUrl != null) {
cachedUrl.release();
}
IOUtil.safeClose(arcStream);
}
if (badEntries == 0 && goodEntries > 0) {
// Make it look like a new crawl finished on each AU to which
// URLs were added.
for (Iterator it = touchedAus.iterator(); it.hasNext(); ) {
ArchivalUnit au = (ArchivalUnit) it.next();
logger.debug3(archiveUrl + " touching " + au.toString());
AuUtil.getDaemon(au).getNodeManager(au).newContentCrawlFinished();
}
} else {
ArchivalUnit au = crawlFacade.getAu();
String msg = archiveUrl + ": " + badEntries + "/" + goodEntries + " bad entries";
throw new CacheException.UnretryableException(msg);
}
}
