public WordList GetOCRText(int page, bool queue_for_ocr)
{
lock (texts)
{
// First check our cache
{
TypedWeakReference <WordList> word_list_weak;
texts.TryGetValue(page, out word_list_weak);
if (null != word_list_weak)
{
WordList word_list = word_list_weak.TypedTarget;
if (null != word_list)
{
return(word_list);
}
}
}
// Then check for an existing SINGLE file
{
string filename = pdf_render_file_layer.MakeFilename_TextSingle(page);
try
{
if (File.Exists(filename))
{
// Get this ONE page
Dictionary <int, WordList> word_lists = WordList.ReadFromFile(filename, page);
WordList word_list = word_lists[page];
texts[page] = new TypedWeakReference <WordList>(word_list);
return(word_list);
}
}
catch (Exception ex)
{
Logging.Warn(ex, "There was an error loading the OCR text for {0} page {1}.", document_fingerprint, page);
FileTools.Delete(filename);
}
}
// Then check for an existing GROUP file
{
string filename = pdf_render_file_layer.MakeFilename_TextGroup(page, TEXT_PAGES_PER_GROUP);
try
{
if (File.Exists(filename))
{
Dictionary <int, WordList> word_lists = WordList.ReadFromFile(filename);
foreach (var pair in word_lists)
{
texts[pair.Key] = new TypedWeakReference <WordList>(pair.Value);
}
TypedWeakReference <WordList> word_list_weak;
texts.TryGetValue(page, out word_list_weak);
if (null != word_list_weak)
{
WordList word_list = word_list_weak.TypedTarget;
if (null != word_list)
{
return(word_list);
}
}
}
}
catch (Exception ex)
{
Logging.Warn(ex, "There was an error loading the OCR text group for {0} page {1}.", document_fingerprint, page);
FileTools.Delete(filename);
}
}
}
// If we get this far then the text was not available so queue extraction
if (queue_for_ocr)
{
// If we have never tried the GROUP version before, queue for it
string filename = pdf_render_file_layer.MakeFilename_TextGroup(page, TEXT_PAGES_PER_GROUP);
if (!File.Exists(filename))
{
PDFTextExtractor.Instance.QueueJobGroup(new PDFTextExtractor.Job(this, page, TEXT_PAGES_PER_GROUP));
}
else
{
PDFTextExtractor.Instance.QueueJobSingle(new PDFTextExtractor.Job(this, page, TEXT_PAGES_PER_GROUP));
}
}
return(null);
}
/// <summary>
/// Returns the OCR words on the page. Null if the words are not yet available.
/// The page will be queued for OCRing if they are not available...
/// Page is 1 based...
/// </summary>
/// <param name="page"></param>
/// <returns></returns>
public WordList GetOCRText(int page, bool queue_for_ocr = true)
{
//Utilities.LockPerfTimer l1_clk = Utilities.LockPerfChecker.Start();
lock (texts_lock)
{
//l1_clk.LockPerfTimerStop();
// First check our cache
{
TypedWeakReference <WordList> word_list_weak;
texts.TryGetValue(page, out word_list_weak);
if (null != word_list_weak)
{
WordList word_list = word_list_weak.TypedTarget;
if (null != word_list)
{
return(word_list);
}
}
}
// Then check for an existing SINGLE file
{
string filename = pdf_render_file_layer.MakeFilename_TextSingle(page);
try
{
if (File.Exists(filename))
{
// Get this ONE page
Dictionary <int, WordList> word_lists = WordList.ReadFromFile(filename, page);
WordList word_list = word_lists[page];
if (null == word_list)
{
throw new Exception(String.Format("No words on page {0} in OCR file {1}", page, filename));
}
texts[page] = new TypedWeakReference <WordList>(word_list);
return(word_list);
}
}
catch (Exception ex)
{
Logging.Warn(ex, "There was an error loading the OCR text for {0} page {1}.", document_fingerprint, page);
FileTools.Delete(filename);
}
}
// Then check for an existing GROUP file
{
string filename = pdf_render_file_layer.MakeFilename_TextGroup(page);
try
{
if (File.Exists(filename))
{
Dictionary <int, WordList> word_lists = WordList.ReadFromFile(filename);
foreach (var pair in word_lists)
{
texts[pair.Key] = new TypedWeakReference <WordList>(pair.Value);
}
TypedWeakReference <WordList> word_list_weak;
texts.TryGetValue(page, out word_list_weak);
if (null != word_list_weak)
{
WordList word_list = word_list_weak.TypedTarget;
if (null != word_list)
{
return(word_list);
}
}
}
}
catch (Exception ex)
{
Logging.Warn(ex, "There was an error loading the OCR text group for {0} page {1}.", document_fingerprint, page);
FileTools.Delete(filename);
}
}
}
// If we get this far then the text was not available so queue extraction
if (queue_for_ocr)
{
// If we have never tried the GROUP version before, queue for it
string filename = pdf_render_file_layer.MakeFilename_TextGroup(page);
PDFTextExtractor.Job job = new PDFTextExtractor.Job(this, page);
if (!File.Exists(filename) && PDFTextExtractor.Instance.JobGroupHasNotFailedBefore(job))
{
PDFTextExtractor.Instance.QueueJobGroup(job);
}
else
{
PDFTextExtractor.Instance.QueueJobSingle(job);
}
}
return(null);
}