public void testGetResourcesByMediaType1()
{
Resources resources = new Resources();
resources.add(new Resource( System.Text.Encoding.UTF8.GetBytes("foo"), MediatypeService.XHTML));
resources.add(new Resource(System.Text.Encoding.UTF8.GetBytes("bar"), MediatypeService.XHTML));
Assert.AreEqual(0, resources.getResourcesByMediaType(MediatypeService.PNG).Count);
Assert.AreEqual(2, resources.getResourcesByMediaType(MediatypeService.XHTML).Count);
Assert.AreEqual(2, resources.getResourcesByMediaTypes(new MediaType[] { MediatypeService.XHTML }).Count);
}
///
/// <param name="packageResource"></param>
/// <param name="epubReader"></param>
/// <param name="book"></param>
/// <param name="resources"></param>
public static void read(Resource packageResource, EpubReader epubReader, Book book, Resources resources)
{
XElement packageDocument = XElement.Load(packageResource.getInputStream());
String packageHref = packageResource.getHref();
resources = fixHrefs(packageHref, resources);
readGuide(packageDocument, epubReader, book, resources);
System.Collections.Generic.Dictionary<string, string> idMapping = new Dictionary<string, string>();
resources = readManifest(packageDocument, packageHref, epubReader, resources, idMapping);
book.setResources(resources);
readCover(packageDocument, book);
// book.setMetadata(PackageDocumentMetadataReader.readMetadata(packageDocument, book.getResources()));
book.setSpine(readSpine(packageDocument, epubReader, book.getResources(), idMapping));
// if we did not find a cover page then we make the first page of the book the cover page
if (book.getCoverPage() == null && book.getSpine().size() > 0)
{
book.setCoverPage(book.getSpine().getResource(0));
}
}
/// <summary>
/// Reads the manifest containing the resource ids, hrefs and mediatypes.
/// </summary>
/// <param>book</param>
/// <param>resourcesByHref</param>
/// <param>a Map with resources, with their id's as key.</param>
/// <param name="packageDocument"></param>
/// <param name="packageHref"></param>
/// <param name="epubReader"></param>
/// <param name="resources"></param>
/// <param name="idMapping"></param>
private static Resources readManifest(XElement packageDocument, String packageHref, EpubReader epubReader, Resources resources, Dictionary<String, String> idMapping)
{
XElement manifestElement = DOMUtil.getFirstElementByTagNameNS(packageDocument, NAMESPACE_OPF, OPFTags.manifest);
Resources result = new Resources();
if (manifestElement == null)
{
return result;
}
var itemElements = packageDocument.Elements(NAMESPACE_OPF + OPFTags.item).Elements<XElement>();
foreach (XElement itemElement in (from e in itemElements where e.Value.Trim() != string.Empty select e))
{
String id = DOMUtil.getAttribute(itemElement, OPFAttributes.id);
String href = DOMUtil.getAttribute(itemElement, OPFAttributes.href);
try
{
href = System.Web.HttpUtility.UrlDecode(href, System.Text.Encoding.GetEncoding(Constants.ENCODING));
}
catch (Exception e)
{
//log.error(e.getMessage());
}
String mediaTypeName = DOMUtil.getAttribute(itemElement, OPFAttributes.media_type);
Resource resource = resources.remove(href);
if (resource == null)
{
//log.error("resource with href '" + href + "' not found");
continue;
}
resource.setId(id);
MediaType mediaType = MediatypeService.getMediaTypeByName(mediaTypeName);
if (mediaType != null)
{
resource.setMediaType(mediaType);
}
result.add(resource);
idMapping.Add(id, resource.getId());
}
return result;
}
/// <summary>
/// Reads the document's spine, containing all sections in reading order.
/// </summary>
/// <param>book</param>
/// <param>resourcesById</param>
/// <param name="packageDocument"></param>
/// <param name="epubReader"></param>
/// <param name="resources"></param>
/// <param name="idMapping"></param>
private static Spine readSpine(XElement packageDocument, EpubReader epubReader, Resources resources, Dictionary<String, String> idMapping)
{
XElement spineElement = DOMUtil.getFirstElementByTagNameNS(packageDocument, NAMESPACE_OPF, OPFTags.spine);
if (spineElement == null)
{
//log.error("Element " + OPFTags.spine + " not found in package document, generating one automatically");
return generateSpineFromResources(resources);
}
Spine result = new Spine();
result.setTocResource(findTableOfContentsResource(spineElement, resources));
var spineNodes = packageDocument.Elements(NAMESPACE_OPF + OPFTags.itemref).Elements<XElement>();
IEnumerator spineNode = spineNodes.GetEnumerator();
List<SpineReference> spineReferences = new List<SpineReference>();
while (spineNode.MoveNext())
{
XElement spineItem = (XElement)spineNode.Current;
String itemref = DOMUtil.getAttribute(spineItem, OPFAttributes.idref);
if (StringUtil.isBlank(itemref))
{
//log.error("itemref with missing or empty idref"); // XXX
continue;
}
String id = idMapping[itemref];
if (id == null)
{
id = itemref;
}
Resource resource = resources.getByIdOrHref(id);
if (resource == null)
{
//log.error("resource with id \'" + id + "\' not found");
continue;
}
SpineReference spineReference = new SpineReference(resource);
if (OPFValues.no.Equals(DOMUtil.getAttribute(spineItem, OPFAttributes.linear)))
{
spineReference.setLinear(false);
}
spineReferences.Add(spineReference);
}
result.setSpineReferences(spineReferences);
return result;
}
/// <summary>
/// Creates a spine out of all resources in the resources. The generated spine
/// consists of all XHTML pages in order of their href.
/// </summary>
/// <param name="resources"></param>
private static Spine generateSpineFromResources(Resources resources)
{
Spine result = new Spine();
List<String> resourceHrefs = new List<String>(resources.getAllHrefs());
foreach (String resourceHref in resourceHrefs)
{
Resource resource = resources.getByHref(resourceHref);
if (resource.getMediaType() == MediatypeService.NCX)
{
result.setTocResource(resource);
}
else if (resource.getMediaType() == MediatypeService.XHTML)
{
result.addSpineReference(new SpineReference(resource));
}
}
return result;
}
/// <summary>
/// Reads the book's guide. Here some more attempts are made at finding the cover
/// page.
/// </summary>
/// <param name="packageDocument"></param>
/// <param name="epubReader"></param>
/// <param name="book"></param>
/// <param name="resources">resources</param>
private static void readGuide(XElement packageDocument, EpubReader epubReader, Book book, Resources resources)
{
XElement guideElement = DOMUtil.getFirstElementByTagNameNS(packageDocument, NAMESPACE_OPF, OPFTags.guide);
if (guideElement == null)
{
return;
}
Guide guide = book.getGuide();
var guideReferences = packageDocument.Elements(NAMESPACE_OPF + OPFTags.reference).Elements<XElement>();
foreach (XElement referenceElement in (from e in guideReferences where e.Value.Trim() != string.Empty select e))
{
String resourceHref = DOMUtil.getAttribute(referenceElement, OPFAttributes.href);
if (StringUtil.isBlank(resourceHref))
{
continue;
}
Resource resource = resources.getByHref(StringUtil.substringBefore(resourceHref, Constants.FRAGMENT_SEPARATOR_CHAR));
if (resource == null)
{
//log.error("Guide is referencing resource with href " + resourceHref + " which could not be found");
continue;
}
String type = DOMUtil.getAttribute(referenceElement, OPFAttributes.type);
if (StringUtil.isBlank(type))
{
//log.error("Guide is referencing resource with href " + resourceHref + " which is missing the 'type' attribute");
continue;
}
String title = DOMUtil.getAttribute(referenceElement, OPFAttributes.title);
if (GuideReference.COVER.Equals(type))
{
continue; // cover is handled elsewhere
}
GuideReference reference = new GuideReference(resource, type, title, StringUtil.substringAfter(resourceHref, Constants.FRAGMENT_SEPARATOR_CHAR));
guide.addReference(reference);
}
}
/// <summary>
/// The spine tag should contain a 'toc' attribute with as value the resource id of
/// the table of contents resource. Here we try several ways of finding this table
/// of contents resource. We try the given attribute value, some often-used ones
/// and finally look through all resources for the first resource with the table of
/// contents mimetype.
/// </summary>
/// <param>resourcesById</param>
/// <param name="spineElement"></param>
/// <param name="resources"></param>
private static Resource findTableOfContentsResource(XElement spineElement, Resources resources)
{
String tocResourceId = DOMUtil.getAttribute(spineElement, OPFAttributes.toc);
Resource tocResource = null;
if (StringUtil.isNotBlank(tocResourceId))
{
tocResource = resources.getByIdOrHref(tocResourceId);
}
if (tocResource != null)
{
return tocResource;
}
for (int i = 0; i < POSSIBLE_NCX_ITEM_IDS.Length; i++)
{
tocResource = resources.getByIdOrHref(POSSIBLE_NCX_ITEM_IDS[i]);
if (tocResource != null)
{
return tocResource;
}
tocResource = resources.getByIdOrHref(POSSIBLE_NCX_ITEM_IDS[i].ToUpper());
if (tocResource != null)
{
return tocResource;
}
}
// get the first resource with the NCX mediatype
tocResource = resources.findFirstResourceByMediaType(MediatypeService.NCX);
if (tocResource == null)
{
// log.error("Could not find table of contents resource. Tried resource with id '" + tocResourceId + "', " + Constants.DEFAULT_TOC_ID + ", " + Constants.DEFAULT_TOC_ID.toUpperCase() + " and any NCX resource.");
}
return tocResource;
}
/// <summary>
/// Strips off the package prefixes up to the href of the packageHref. Example: If
/// the packageHref is "OEBPS/content.opf" then a resource href like "OEBPS/foo/bar.
/// html" will be turned into "foo/bar.html"
/// </summary>
/// <param name="packageHref"></param>
/// <param name="resourcesByHref"></param>
private static Resources fixHrefs(String packageHref, Resources resourcesByHref)
{
int lastSlashPos = packageHref.IndexOf('/');
if (lastSlashPos < 0)
{
return resourcesByHref;
}
Resources result = new Resources();
foreach (Resource resource in resourcesByHref.getAll())
{
if (StringUtil.isNotBlank(resource.getHref())
|| resource.getHref().Length > lastSlashPos)
{
resource.setHref(resource.getHref().Substring(lastSlashPos + 1));
}
result.add(resource);
}
return result;
}
///
/// <param name="in"></param>
/// <param name="defaultHtmlEncoding"></param>
private Resources readResources(ZipInputStream zipInputStream, String defaultHtmlEncoding)
{
Resources result = new Resources();
for (ZipEntry zipEntry = zipInputStream.GetNextEntry(); zipEntry != null; zipEntry = zipInputStream.GetNextEntry())
{
if (zipEntry.IsDirectory)
{
continue;
}
Resource resource = ResourceUtil.createResource(zipEntry, zipInputStream);
if (resource.getMediaType() == MediatypeService.XHTML)
{
resource.setInputEncoding(defaultHtmlEncoding);
}
result.add(resource);
}
return result;
}
///
/// <param name="packageResourceHref"></param>
/// <param name="book"></param>
/// <param name="resources"></param>
private Resource processPackageResource(String packageResourceHref, Book book, Resources resources)
{
Resource packageResource = resources.remove(packageResourceHref);
try
{
PackageDocumentReader.read(packageResource, this, book, resources);
}
catch (Exception e)
{
//log.error(e.getMessage(), e);
}
return packageResource;
}
///
/// <param name="result"></param>
/// <param name="resources"></param>
private void handleMimeType(Book result, Resources resources)
{
resources.remove("mimetype");
}
///
/// <param name="resources"></param>
private String getPackageResourceHref(Resources resources)
{
String defaultResult = "OEBPS/content.opf";
String result = defaultResult;
Resource containerResource = resources.remove("META-INF/container.xml");
if (containerResource == null)
{
return result;
}
try
{
XElement xElement = XElement.Load(containerResource.getInputStream());
XNamespace ns = (xElement.Attribute("xmlns") != null) ? xElement.Attribute("xmlns").Value : XNamespace.None;
return xElement.Descendants(ns + "rootfile").FirstOrDefault((XElement p) => p.Attribute("media-type") != null && p.Attribute("media-type").Value.Equals("application/oebps-package+xml", System.StringComparison.InvariantCultureIgnoreCase)).Attribute("full-path").Value;
}
catch (Exception e)
{
//log.error(e.getMessage(), e);
}
if (StringUtil.isBlank(result))
{
result = defaultResult;
}
return result;
return "";
}
///
/// <param name="resources"></param>
public void setResources(Resources resources)
{
this.resources = resources;
}
