protected ArticleFiles processUrl(CachedUrl cu, Matcher mat) {
ArticleFiles af = new ArticleFiles();
af.setFullTextCu(cu);
af.setRoleCu(ArticleFiles.ROLE_ABSTRACT, cu);
// XXX Full text PDF link embedded in page, cannot guess URL
return af;
}
public void testFunctionalFromTarHierarchy() throws Exception {
log.debug3("in testFromTarHierarchy");
// load the tarballs
InputStream file_input = null;
try {
file_input = getResourceAsStream(realTARFile_A);
// UrlCacher uc = au.makeUrlCacher(TAR_A_BASE);
// uc.storeContent(file_input, tarHeader);
UrlCacher uc = tarAu.makeUrlCacher(new UrlData(file_input, tarHeader, TAR_A_BASE));
uc.storeContent();
IOUtil.safeClose(file_input);
file_input = getResourceAsStream(realTARFile_B);
// uc = au.makeUrlCacher(TAR_B_BASE);
// uc.storeContent(file_input, tarHeader);
uc = tarAu.makeUrlCacher(new UrlData(file_input, tarHeader, TAR_B_BASE));
uc.storeContent();
IOUtil.safeClose(file_input);
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} finally {
IOUtil.safeClose(file_input);
}
CachedUrlSet cus = tarAu.getAuCachedUrlSet();
for (CachedUrl cu : cus.getCuIterable()) {
log.debug3("AU - cu is: " + cu.getUrl());
cu.release();
}
// We need to start from the level of the ArticleMetadataExtractor
MyListEmitter emitter = new MyListEmitter();
ArticleMetadataExtractor amEx =
new ElsevierDeferredArticleMetadataExtractor(ArticleFiles.ROLE_ARTICLE_METADATA);
Iterator<ArticleFiles> it = tarAu.getArticleIterator(MetadataTarget.Any());
while (it.hasNext()) {
ArticleFiles af = it.next();
log.debug3("Metadata test - articlefiles " + af.toString());
// CachedUrl cu = af.getFullTextCu();
CachedUrl cu = af.getRoleCu(ArticleFiles.ROLE_ARTICLE_METADATA);
log.debug3("metadata cu is " + cu.getUrl());
// List<ArticleMetadata> mdlist = mle.extract(MetadataTarget.Any(), cu);
amEx.extract(MetadataTarget.Any(), af, emitter);
List<ArticleMetadata> returnList = emitter.getAmList();
assertNotNull(returnList);
log.debug3("size of returnList is " + returnList.size());
Iterator<ArticleMetadata> mdIt = returnList.iterator();
ArticleMetadata mdRecord = null;
while (mdIt.hasNext()) {
mdRecord = (ArticleMetadata) mdIt.next();
validateCompleteMetadataRecord(mdRecord);
}
}
}
public void testArticleCountAndType() throws Exception {
int expCount = 28;
PluginTestUtil.crawlSimAu(sau);
String pat1 = "branch(\\d+)/(\\d+file\\.html)";
String rep1 = "aps/journal/v123/n$1/full/$2";
PluginTestUtil.copyAu(sau, nau, ".*[^.][^p][^d][^f]$", pat1, rep1);
String pat2 = "branch(\\d+)/(\\d+file\\.pdf)";
String rep2 = "aps/journal/v123/n$1/pdf/$2";
PluginTestUtil.copyAu(sau, nau, ".*\\.pdf$", pat2, rep2);
// Remove some URLs
int deleted = 0;
for (Iterator it = nau.getAuCachedUrlSet().contentHashIterator(); it.hasNext(); ) {
CachedUrlSetNode cusn = (CachedUrlSetNode) it.next();
if (cusn instanceof CachedUrl) {
CachedUrl cu = (CachedUrl) cusn;
String url = cu.getUrl();
if (url.contains("/journal/")
&& (url.endsWith("1file.html") || url.endsWith("2file.pdf"))) {
deleteBlock(cu);
++deleted;
}
}
}
assertEquals(8, deleted);
Iterator<ArticleFiles> it = nau.getArticleIterator();
int count = 0;
int countHtmlOnly = 0;
int countPdfOnly = 0;
while (it.hasNext()) {
ArticleFiles af = it.next();
log.info(af.toString());
CachedUrl cu = af.getFullTextCu();
String url = cu.getUrl();
assertNotNull(cu);
String contentType = cu.getContentType();
log.debug("count " + count + " url " + url + " " + contentType);
count++;
if (af.getRoleUrl(ArticleFiles.ROLE_FULL_TEXT_PDF) == null) {
++countHtmlOnly;
}
if (af.getRoleUrl(ArticleFiles.ROLE_FULL_TEXT_PDF) == url) {
++countPdfOnly;
}
}
log.debug("Article count is " + count);
assertEquals(expCount, count);
assertEquals(4, countHtmlOnly);
assertEquals(4, countPdfOnly);
}
protected ArticleFiles processFullTextPdf(CachedUrl cu, Matcher mat) {
ArticleFiles af = new ArticleFiles();
af.setRoleCu(ArticleFiles.ROLE_FULL_TEXT_PDF, cu);
String abstractUrl = mat.replaceAll("$1aspx");
CachedUrl abstractCu = cu.getArchivalUnit().makeCachedUrl(abstractUrl);
if (abstractCu.hasContent()) af.setFullTextCu(abstractCu);
else af.setFullTextCu(cu);
log.debug3("returning full text: " + af.getFullTextUrl());
return af;
}
public void testCreateArticleFiles() throws Exception {
PluginTestUtil.crawlSimAu(sau);
String pat0 = "001file[.]html";
String rep0 = "52/1/S1";
PluginTestUtil.copyAu(sau, au, ".*[.]html$", pat0, rep0);
String pat1 = "001file[.]pdf";
String rep1 = "52/1/S1.pdf";
PluginTestUtil.copyAu(sau, au, ".*[.]pdf$", pat1, rep1);
String pdfurl = "http://pediatrics.aappublications.org/cgi/reprint/52/1/S1.pdf";
String url = "http://pediatrics.aappublications.org/cgi/reprint/52/1/S1";
au.makeCachedUrl(url);
CachedUrl cu = au.makeCachedUrl(pdfurl);
assertNotNull(cu);
SubTreeArticleIterator artIter = createSubTreeIter();
assertNotNull(artIter);
ArticleFiles af = artIter.next();
assertNotNull(af);
System.out.println("article files::" + af);
assertEquals(url, af.getRoleCu(ArticleFiles.ROLE_FULL_TEXT_PDF_LANDING_PAGE).getUrl());
assertEquals(pdfurl, af.getRoleCu(ArticleFiles.ROLE_FULL_TEXT_PDF).getUrl());
}
public void testCreateArticleFiles() throws Exception {
// create urls to store in UrlCacher
String[] au_urls = {
BASE_URL + "pc/doifinder/10.1057/9780123456789",
BASE_URL + "pc/browse/citationExport?doi=10.1057/9780123456789",
BASE_URL + "pc/doifinder/download/10.1057/9780123456789",
BASE_URL + "pc/doifinder/download/10.1057/9780123456789.epub",
BASE_URL + "pc/doifinder/10.1057/9781234567890",
BASE_URL + "pc/browse/citationExport?doi=10.1057/9781234567890",
BASE_URL + "pc/doifinder/download/10.1057/9781234567890",
BASE_URL + "pc/doifinder/download/10.1057/9781234567890.epub",
BASE_URL + "pc/doifinder/10.1057/9782345678901",
BASE_URL + "pc/browse/citationExport?doi=10.1057/9782345678901",
BASE_URL + "pc/doifinder/download/10.1057/9782345678901",
BASE_URL + "pc/doifinder/download/10.1057/9782345678901.epub"
};
/*
// get cached url content type and properties from simulated contents
// for UrclCacher.storeContent()
CachedUrl cuPdf = null;
CachedUrl cuHtml = null;
CachedUrl cuEpub = null;
for (CachedUrl cu : AuUtil.getCuIterable(sau)) {
if (cuPdf == null
&& cu.getContentType().toLowerCase().startsWith(Constants.MIME_TYPE_PDF)) {
log.info("pdf contenttype: " + cu.getContentType());
cuPdf = cu;
} else if (cuHtml == null
&& cu.getContentType().toLowerCase().startsWith(Constants.MIME_TYPE_HTML)) {
log.info("html contenttype: " + cu.getContentType());
cuHtml = cu;
} else if (cuEpub == null
&& cu.getContentType().toLowerCase().startsWith("application/epub")) {
log.info("epub contenttype: " + cu.getContentType());
cuEpub = cu;
}
if (cuPdf != null && cuHtml != null && cuEpub != null) {
break;
}
}
*/
CachedUrl cu;
// store content using cached url content type and properties
for (String url : au_urls) {
if (url.contains("download") && !url.endsWith(".epub")) {
storeContent(random_content_stream, pdfHeader, url);
} else if (url.contains("download")) { // epub
storeContent(random_content_stream, epubHeader, url);
} else {
storeContent(random_content_stream, textHeader, url);
}
}
// book 9780123456789
ArticleFiles af1 = new ArticleFiles();
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_PDF, BASE_URL + "pc/doifinder/download/10.1057/9780123456789");
af1.setRoleString(
ArticleFiles.ROLE_ARTICLE_METADATA,
BASE_URL + "pc/browse/citationExport?doi=10.1057/9780123456789");
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_EPUB,
BASE_URL + "pc/doifinder/download/10.1057/9780123456789.epub");
// book 9780123456789
ArticleFiles af2 = new ArticleFiles();
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_PDF, BASE_URL + "pc/doifinder/download/10.1057/9781234567890");
af1.setRoleString(
ArticleFiles.ROLE_ARTICLE_METADATA,
BASE_URL + "pc/browse/citationExport?doi=10.1057/9781234567890");
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_EPUB,
BASE_URL + "pc/doifinder/download/10.1057/9781234567890.epub");
// book 9780123456789
ArticleFiles af3 = new ArticleFiles();
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_PDF, BASE_URL + "pc/doifinder/download/10.1057/9782345678901");
af1.setRoleString(
ArticleFiles.ROLE_ARTICLE_METADATA,
BASE_URL + "pc/browse/citationExport?doi=10.1057/9782345678901");
af1.setRoleString(
ArticleFiles.ROLE_FULL_TEXT_EPUB,
BASE_URL + "pc/doifinder/download/10.1057/9782345678901.epub");
// key the expected content to the fullTextUrl for the ArticleFiles
HashMap<String, ArticleFiles> fullUrlToAF = new HashMap<String, ArticleFiles>();
fullUrlToAF.put(BASE_URL + "pc/doifinder/download/10.1057/9780123456789", af1);
fullUrlToAF.put(BASE_URL + "pc/doifinder/download/10.1057/9781234567890", af2);
fullUrlToAF.put(BASE_URL + "pc/doifinder/download/10.1057/9782345678901", af3);
// get article iterator, get article files and the appropriate urls according
// to their roles.
String[] expectedUrls = {
EXPECTED_FULL_TEXT_URL, EXPECTED_PDF_URL,
};
for (SubTreeArticleIterator artIter = createSubTreeIter(); artIter.hasNext(); ) {
ArticleFiles af = artIter.next();
String[] actualUrls = {
af.getFullTextUrl(), af.getRoleUrl(ArticleFiles.ROLE_FULL_TEXT_PDF),
// af.getRoleUrl(ArticleFiles.ROLE_FULL_TEXT_PDF_LANDING_PAGE)
};
log.info("actualUrls: " + actualUrls.length);
for (int i = 0; i < actualUrls.length; i++) {
log.info("e_url: " + expectedUrls[i]);
log.info("url: " + actualUrls[i]);
// assertEquals(expectedUrls[i], actualUrls[i]);
}
}
}
/*
* 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;
}