private Set outputUrlResults(String url, Set m_inclset, Set m_exclset) {
Set new_incls = new TreeSet(CollectionUtils.subtract(m_inclset, m_reported));
Set new_excls = new TreeSet(CollectionUtils.subtract(m_exclset, m_reported));
if (!m_inclset.isEmpty()) {
outputMessage(
"\nIncluded Urls: ("
+ new_incls.size()
+ " new, "
+ (m_inclset.size() - new_incls.size())
+ " old)",
URL_SUMMARY_MESSAGE);
depth_incl[m_curDepth - 1] += new_incls.size();
}
for (Iterator it = new_incls.iterator(); it.hasNext(); ) {
outputMessage(it.next().toString(), PLAIN_MESSAGE);
}
if (!m_exclset.isEmpty()) {
outputMessage(
"\nExcluded Urls: ("
+ new_excls.size()
+ " new, "
+ (m_exclset.size() - new_excls.size())
+ " old)",
URL_SUMMARY_MESSAGE);
}
for (Iterator it = new_excls.iterator(); it.hasNext(); ) {
outputMessage(it.next().toString(), PLAIN_MESSAGE);
}
m_reported.addAll(new_incls);
m_reported.addAll(new_excls);
if (m_outWriter != null) {
try {
m_outWriter.flush();
} catch (IOException ex) {
}
}
return new_incls;
}
private void outputSummary(String baseUrl, Set fetched, Set toCrawl, long elapsedTime) {
int fetchCount = fetched.size();
outputMessage(
"\n\nSummary for starting Url: " + baseUrl + " and depth: " + m_crawlDepth,
TEST_SUMMARY_MESSAGE);
outputMessage(
"\nUrls fetched: " + fetchCount + " Urls extracted: " + m_extracted.size(),
PLAIN_MESSAGE);
outputMessage("\nDepth Fetched Parsed New URLs", PLAIN_MESSAGE);
for (int depth = 1; depth <= m_crawlDepth; depth++) {
PrintfFormat pf = new PrintfFormat("%5d %7d %6d %8d");
Integer[] args =
new Integer[] {
new Integer(depth),
new Integer(depth_fetched[depth - 1]),
new Integer(depth_parsed[depth - 1]),
new Integer(depth_incl[depth - 1]),
};
String s = pf.sprintf(args);
outputMessage(s, PLAIN_MESSAGE);
}
outputMessage("\nRemaining unfetched: " + toCrawl.size(), PLAIN_MESSAGE);
if (false) {
for (Iterator iter = toCrawl.iterator(); iter.hasNext(); ) {
String url = (String) iter.next();
outputMessage(url, PLAIN_MESSAGE);
}
}
long secs = elapsedTime / Constants.SECOND;
long fetchRate = 0;
if (secs > 0) {
fetchRate = fetchCount * 60 * Constants.SECOND / elapsedTime;
}
outputMessage(
"\nElapsed Time: " + secs + " secs." + " Fetch Rate: " + fetchRate + " p/m",
PLAIN_MESSAGE);
}
protected CIProperties makeCIProperties(ArchiveRecordHeader elementHeader) throws IOException {
CIProperties ret = new CIProperties();
Set elementHeaderFieldKeys = elementHeader.getHeaderFieldKeys();
for (Iterator i = elementHeaderFieldKeys.iterator(); i.hasNext(); ) {
String key = (String) i.next();
try {
Object valueObject = elementHeader.getHeaderValue(key);
if (valueObject == null) {
logger.warning("Ignoring null value for key '" + key + "'.");
} else {
String value = valueObject.toString();
logger.debug3(key + ": " + value);
ret.put(key, value);
}
} catch (ClassCastException ex) {
logger.error("makeCIProperties: " + key + " threw ", ex);
throw new CacheException.ExploderException(ex);
}
}
return (ret);
}
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);
}
}
}
/** 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);
}
}