private static DictionaryBasedObject ReadFromStream_BINARY(byte[] data)
{
DictionaryBasedObject dbo = (DictionaryBasedObject)SerializeFile.LoadFromByteArray(data);
FeatureTrackingManager.Instance.UseFeature(Features.Legacy_Metadata_Binary);
return(dbo);
}
public string MetadataAsString()
{
// keep the unrecognized data around so we may fix it later...
string str = null;
try
{
try
{
DictionaryBasedObject dictionary = PDFMetadataSerializer.ReadFromStream(data);
str = JsonConvert.SerializeObject(dictionary.Attributes, Formatting.Indented);
}
catch (Exception ex)
{
Logging.Error(ex);
str = StringTools.HumanReadableASCIIAndHexStr(data);
}
}
catch (Exception ex)
{
Logging.Error(ex);
}
return(str);
}
private static DictionaryBasedObject ReadFromDisk_BINARY(string filename)
{
DictionaryBasedObject dbo = (DictionaryBasedObject)SerializeFile.LoadRedundant(filename);
FeatureTrackingManager.Instance.UseFeature(Features.Legacy_Metadata_Binary);
return(dbo);
}
public static PDFDocument LoadFromMetaData(Library library, byte[] data, Dictionary <string, byte[]> /* can be null */ library_items_annotations_cache)
{
DictionaryBasedObject dictionary = PDFMetadataSerializer.ReadFromStream(data);
PDFDocument pdf_document = new PDFDocument(library, dictionary);
pdf_document.GetAnnotations(library_items_annotations_cache);
return(pdf_document);
}
/// <summary>
/// If new (e.g. cloning), generate a new guid.
/// </summary>
public PDFAnnotation(DictionaryBasedObject dictionary, bool is_new)
{
this.dictionary = dictionary;
if (is_new)
{
Guid = System.Guid.NewGuid();
}
}
/// <summary>
/// Throws exception when metadata could not be converted to a valid PDFDocument instance.
/// </summary>
/// <param name="library"></param>
/// <param name="data"></param>
/// <param name="library_items_annotations_cache"></param>
/// <returns></returns>
public static PDFDocument LoadFromMetaData(Library library, byte[] data, Dictionary <string, byte[]> /* can be null */ library_items_annotations_cache)
{
DictionaryBasedObject dictionary = PDFMetadataSerializer.ReadFromStream(data);
LockObject _lock = new LockObject();
PDFDocument pdf_document = new PDFDocument(_lock, library, dictionary);
// thread-UNSAFE access is permitted as the PDF has just been created so there's no thread-safety risk yet.
pdf_document.doc.GetAnnotations(library_items_annotations_cache);
return(pdf_document);
}
/// <summary>
/// Throws exception when metadata could not be converted to a valid PDFDocument instance.
/// </summary>
/// <param name="library"></param>
/// <param name="data"></param>
/// <param name="library_items_annotations_cache"></param>
/// <returns></returns>
public static PDFDocument LoadFromMetaData(WebLibraryDetail web_library_detail, string fingerprint, byte[] data, PDFAnnotationList prefetched_annotations_for_document = null)
{
WPFDoEvents.AssertThisCodeIs_NOT_RunningInTheUIThread();
ASSERT.Test(!String.IsNullOrEmpty(fingerprint));
DictionaryBasedObject dictionary = PDFMetadataSerializer.ReadFromStream(data);
// Recover partially damaged record: if there's no fingerprint in the dictionary, add the given one:
string id = dictionary["Fingerprint"] as string;
if (String.IsNullOrEmpty(id))
{
Logging.Error("Library '{1}' corrupted? Sanity check: the document record for document fingerprint {0} lacked a fingerprint as part of the mandatory metadata. This has been fixed now.", fingerprint, web_library_detail.DescriptiveTitle);
dictionary["Fingerprint"] = fingerprint;
id = fingerprint;
}
// extra verification / sanity check: this will catch some very obscure DB corruption, or rather **manual editing**! ;-)
if (id != fingerprint || true)
{
// see which of them makes the most sense... And DO remember that Qiqqa fingeerprint hashes are variable-length, due to an old systemic bug!
Regex re = new Regex(@"^[a-zA-Z0-9]{20,}(?:_REF)?$");
bool id_is_possibly_legal = re.IsMatch(id);
bool key_is_possibly_legal = re.IsMatch(fingerprint);
// if the entry in the record itself is legal, run with that one. Otherwise, fall back to using the DB record KEY.
if (id_is_possibly_legal)
{
Logging.Error("Library '{2}' corrupted? Sanity check: given fingerprint '{0}' does not match the fingerprint '{1}' obtained from the DB metadata record. Running with the fingerprint specified in the metadata record.", fingerprint, id, web_library_detail.DescriptiveTitle);
}
else
{
Logging.Error("Library '{2}' corrupted? Sanity check: given fingerprint '{0}' does not match the fingerprint '{1}' obtained from the DB metadata record. Running with the fingerprint specified as database record KEY, since the other fingerprint does not look like a LEGAL fingerprint.", fingerprint, id, web_library_detail.DescriptiveTitle);
dictionary["Fingerprint"] = fingerprint;
id = fingerprint;
}
}
PDFDocument pdf_document = new PDFDocument(web_library_detail, dictionary, prefetched_annotations_for_document);
// thread-UNSAFE access is permitted as the PDF has just been created and
// is NOT known to the outside world (i.e. beyond the scope of this function)
// so there's no thread-safety risk yet.
_ = pdf_document.GetAnnotations();
return(pdf_document);
}
internal PDFDocument(WebLibraryDetail web_library_detail, DictionaryBasedObject dictionary, PDFAnnotationList prefetched_annotations_for_document = null)
{
this.library = new TypedWeakReference <WebLibraryDetail>(web_library_detail);
this.dictionary = dictionary;
// process any prefetched annotations that we may have as usual:
if (prefetched_annotations_for_document != null)
{
annotations = prefetched_annotations_for_document;
lock (access_lock)
{
dirtyNeedsReindexing = (prefetched_annotations_for_document.Count > 0);
}
}
}
/// <summary>
/// NB: only call this as part of document creation.
/// </summary>
public void CloneMetaData(PDFDocument existing_pdf_document)
{
bindable = null;
Logging.Info("Cloning metadata from {0}", existing_pdf_document.Title);
dictionary = (DictionaryBasedObject)existing_pdf_document.dictionary.Clone();
annotations = (PDFAnnotationList)existing_pdf_document.Annotations.Clone();
highlights = (PDFHightlightList)existing_pdf_document.Highlights.Clone();
inks = (PDFInkList)existing_pdf_document.Inks.Clone();
SaveToMetaData();
// Copy the citations
PDFDocumentCitationManager.CloneFrom(existing_pdf_document.PDFDocumentCitationManager);
// Now clear out the references for the annotations and highlights, so that when they are reloaded the events are resubscribed
annotations = null;
highlights = null;
inks = null;
}
internal PDFDocument_ThreadUnsafe(Library library, DictionaryBasedObject dictionary)
{
this.library = new TypedWeakReference <Library>(library);
this.dictionary = dictionary;
}
private PDFDocument(Library library, DictionaryBasedObject dictionary)
{
this.library = library;
this.dictionary = dictionary;
}
private PDFDocument(Library library)
{
this.library = library;
this.dictionary = new DictionaryBasedObject();
}
private PDFDocument(LockObject _lock, Library library, DictionaryBasedObject dictionary)
{
access_lock = _lock;
doc = new PDFDocument_ThreadUnsafe(library, dictionary);
}
internal PDFDocument(WebLibraryDetail web_library_detail)
{
this.library = new TypedWeakReference <WebLibraryDetail>(web_library_detail);
dictionary = new DictionaryBasedObject();
}
