/**
* Return the archive file type corresponding to the filename extension in the URL, or null if
* none.
*/
public String getFromUrl(String url) throws MalformedURLException {
if (StringUtil.endsWithIgnoreCase(url, ".tar.gz")) {
return getExtMimeMap().get(".tar.gz");
}
String ext = UrlUtil.getFileExtension(url).toLowerCase();
return getExtMimeMap().get("." + ext);
}
/**
* A method to remove any non-canonical '..' or '.' elements in the path, as well as protecting
* against illegal path traversal.
*
* @param url the raw url
* @return String the canonicalized url
* @throws MalformedURLException
*/
public String canonicalizePath(String url) throws MalformedURLException {
String canonUrl = UrlUtil.normalizeUrl(url, UrlUtil.PATH_TRAVERSAL_ACTION_THROW);
// canonicalize "dir" and "dir/"
// XXX if these are ever two separate nodes, this is wrong
if (canonUrl.endsWith(UrlUtil.URL_PATH_SEPARATOR)) {
canonUrl = canonUrl.substring(0, canonUrl.length() - 1);
}
return canonUrl;
}
private void buildUrlSets(String url) {
try {
outputMessage("\nFetching " + url, TEST_SUMMARY_MESSAGE);
URL srcUrl = new URL(url);
// URLConnection conn = srcUrl.openConnection();
// String type = conn.getContentType();
// type = conn.getHeaderField("content-type");
// InputStream istr = conn.getInputStream();
LockssUrlConnection conn = UrlUtil.openConnection(url, connectionPool);
if (proxyHost != null) {
conn.setProxy(proxyHost, proxyPort);
}
if (userAgent != null) {
conn.setRequestProperty("user-agent", userAgent);
}
try {
conn.execute();
int resp = conn.getResponseCode();
if (resp != 200) {
outputMessage("Resp: " + resp + ": " + conn.getResponseMessage(), TEST_SUMMARY_MESSAGE);
return;
}
depth_fetched[m_curDepth - 1]++;
String cookies = conn.getResponseHeaderValue("Set-Cookie");
if (cookies != null) {
outputMessage("Cookies: " + cookies, PLAIN_MESSAGE);
}
String type = conn.getResponseContentType();
if (type == null || !type.toLowerCase().startsWith("text/html")) {
outputMessage("Type: " + type + ", not parsing", URL_SUMMARY_MESSAGE);
return;
}
outputMessage("Type: " + type + ", extracting Urls", URL_SUMMARY_MESSAGE);
InputStream istr = conn.getResponseInputStream();
InputStreamReader reader = new InputStreamReader(istr);
// MyMockCachedUrl mcu = new MyMockCachedUrl(srcUrl.toString(), reader);
GoslingHtmlLinkExtractor extractor = new GoslingHtmlLinkExtractor();
extractor.extractUrls(null, istr, null, srcUrl.toString(), new MyLinkExtractorCallback());
istr.close();
depth_parsed[m_curDepth - 1]++;
} finally {
conn.release();
}
} catch (MalformedURLException murle) {
murle.printStackTrace();
outputErrResults(url, "Malformed URL:" + murle.getMessage());
} catch (IOException ex) {
ex.printStackTrace();
outputErrResults(url, "IOException: " + ex.getMessage());
}
}
// code copied from HttpClientUrlConnect.createMethod - without normalizing
// both encoded and unencoded urls will be preserved
public String normalizeUrl(String urlString, ArchivalUnit au) throws PluginException {
String u_str = urlString;
/* if the urlString is not "readable" ascii (0x1F < x < 0x7f), then
* normalize/encode the string
*/
if (!isReadableAscii(urlString)) {
if (log.isDebug2()) log.debug2("in:" + u_str);
u_str = UrlUtil.encodeUri(urlString, Constants.ENCODING_UTF_8);
if (log.isDebug2()) log.debug2("out:" + u_str);
}
return u_str;
}
public void foundLink(String url) {
m_extracted.add(url);
try {
String normUrl = UrlUtil.normalizeUrl(url);
if (BaseCrawler.isSupportedUrlProtocol(normUrl) && m_au.shouldBeCached(normUrl)) {
m_incls.add(normUrl);
} else {
m_excls.add(normUrl);
}
} catch (MalformedURLException e) {
m_excls.add(url);
}
}
/*
* In order to find full text PDF you need to find the citation_pdf_url meta tag in the
* abstract html pull out the pdf url normalize it (reorder params...) and find the matching
* cached URL
*/
protected ArticleFiles processAbstract(CachedUrl absCu, Matcher absMat) {
NodeList nl = null;
ArticleFiles af = new ArticleFiles();
if (absCu != null && absCu.hasContent()) {
// TEMPORARY: set absCU as default full text CU in case there is
// no PDF CU with content; the current metadata manager currently
// uses only the full text CU, but this will change with the new
// metadata schema that can have multiple CUs for an article.
af.setFullTextCu(absCu);
af.setRoleCu(ArticleFiles.ROLE_ABSTRACT, absCu);
try {
InputStreamSource is =
new InputStreamSource(new Stream(absCu.getUnfilteredInputStream()));
Page pg = new Page(is);
Lexer lx = new Lexer(pg);
Parser parser = new Parser(lx);
Lexer.STRICT_REMARKS = false;
NodeFilter nf =
new NodeFilter() {
public boolean accept(Node node) {
if (!(node instanceof MetaTag)) return false;
MetaTag meta = (MetaTag) node;
if (!"citation_pdf_url".equalsIgnoreCase(meta.getMetaTagName())) return false;
return true;
}
};
nl = parser.extractAllNodesThatMatch(nf);
} catch (ParserException e) {
log.debug("Unable to parse abstract page html", e);
} catch (UnsupportedEncodingException e) {
log.debug("Bad encoding in abstact page html", e);
} finally {
absCu.release();
}
}
try {
if (nl != null) {
if (nl.size() > 0) {
// minimally encode URL to prevent URL constructor
// from stripping trailing spaces
String pdfUrlStr = ((MetaTag) nl.elementAt(0)).getMetaContent();
URL pdfUrl = new URL(UrlUtil.minimallyEncodeUrl(pdfUrlStr));
List<String> paramList = new ArrayList<String>();
paramList.add("fileType");
paramList.add("fileId");
paramList.add("fileName");
pdfUrl = reArrangeUrlParams(pdfUrl, paramList);
if (!pdfUrl.getHost().startsWith("www.")) {
pdfUrl = new URL(pdfUrl.getProtocol(), "www." + pdfUrl.getHost(), pdfUrl.getFile());
}
// note: must leave URL encoded because that's how we store URLs
CachedUrl pdfCu = au.makeCachedUrl(pdfUrl.toString());
if (pdfCu != null && pdfCu.hasContent()) {
// replace absCU with pdfCU if exists and has content
af.setFullTextCu(pdfCu);
af.setRoleCu(ArticleFiles.ROLE_FULL_TEXT_PDF, pdfCu);
}
}
}
} catch (MalformedURLException e) {
log.debug("Badly formatted pdf url link", e);
} catch (IllegalArgumentException e) {
log.debug("Badly formatted pdf url link", e);
}
return af;
}
protected String urlEncode(String param) {
return UrlUtil.encodeUrl(param);
}
/** Common request handling. */
public void service(HttpServletRequest req, HttpServletResponse resp)
throws ServletException, IOException {
resetState();
boolean success = false;
HttpSession session = req.getSession(false);
try {
this.req = req;
this.resp = resp;
if (log.isDebug()) {
logParams();
}
resp.setContentType("text/html");
if (!mayPageBeCached()) {
resp.setHeader("pragma", "no-cache");
resp.setHeader("Cache-control", "no-cache");
}
reqURL = new URL(UrlUtil.getRequestURL(req));
clientAddr = getLocalIPAddr();
// check that current user has permission to run this servlet
if (!isServletAllowed(myServletDescr())) {
displayWarningInLieuOfPage("You are not authorized to use " + myServletDescr().heading);
return;
}
// check whether servlet is disabled
String reason = ServletUtil.servletDisabledReason(myServletDescr().getServletName());
if (reason != null) {
displayWarningInLieuOfPage("This function is disabled. " + reason);
return;
}
if (session != null) {
session.setAttribute(SESSION_KEY_RUNNING_SERVLET, getHeading());
String reqHost = req.getRemoteHost();
String forw = req.getHeader(HttpFields.__XForwardedFor);
if (!StringUtil.isNullString(forw)) {
reqHost += " (proxies for " + forw + ")";
}
session.setAttribute(SESSION_KEY_REQUEST_HOST, reqHost);
}
lockssHandleRequest();
success = (errMsg == null);
} catch (ServletException e) {
log.error("Servlet threw", e);
throw e;
} catch (IOException e) {
log.error("Servlet threw", e);
throw e;
} catch (RuntimeException e) {
log.error("Servlet threw", e);
throw e;
} finally {
if (session != null) {
session.setAttribute(SESSION_KEY_RUNNING_SERVLET, null);
session.setAttribute(LockssFormAuthenticator.__J_AUTH_ACTIVITY, TimeBase.nowMs());
}
if ("please".equalsIgnoreCase(req.getHeader("X-Lockss-Result"))) {
log.debug3("X-Lockss-Result: " + (success ? "Ok" : "Fail"));
resp.setHeader("X-Lockss-Result", success ? "Ok" : "Fail");
}
resetMyLocals();
resetLocals();
}
}
