private int num_pages = -1; // signal: -3 and below: failed permanently (2 + attempts to diagnose); -2: failed permanently due to absent PDF file; -1: not yet determined; 0: empty document (rare/weird!); > 0: number of actual pages in document
/// <summary>
/// This is an approximate response: it takes a *fast* shortcut to check if the given
/// PDF has been OCR'd in the past.
///
/// The emphasis here is on NOT triggering a new OCR action! Just taking a peek, *quickly*.
///
/// The cost: one(1) I/O per check.
///
/// Implementation Note: do NOT check if DocumentExists: our pagecount cache check is sufficient and one I/O per check.
/// </summary>
public bool HasOCRdata()
{
// BasePath:
string base_path = ConfigurationManager.Instance.ConfigurationRecord.System_OverrideDirectoryForOCRs;
if (String.IsNullOrEmpty(base_path))
{
base_path = BASE_PATH_DEFAULT;
}
// string cached_count_filename = MakeFilename_PageCount(Fingerprint);
// return MakeFilenameWith2LevelIndirection("pagecount", "0", "txt");
string indirection_characters = Fingerprint.Substring(0, 2).ToUpper();
string cached_count_filename = Path.GetFullPath(Path.Combine(base_path, indirection_characters, String.Format("{0}.{1}.{2}.{3}", Fingerprint, @"pagecount", @"0", @"txt")));
return(File.Exists(cached_count_filename));
}
private string MakeFilenameWith2LevelIndirection(string file_type, object token, string extension)
{
string indirection_characters = Fingerprint.Substring(0, 2).ToUpper();
return(Path.GetFullPath(Path.Combine(BasePath, indirection_characters, String.Format("{0}.{1}.{2}.{3}", Fingerprint, file_type, token, extension))));
}
