/// <summary>
/// Run the algorithm on all non-backup files in the source directory.
/// </summary>
public void Run()
{
DateTime fileDate;
if (m_tablePaths == null)
{
// Use the default
m_tablePaths = new string[] {
// remove spaces at end of line. They can end up between end of sentence and note marker.
Utils.GetUtilityFile("cleanup_EOL_spaces.cct"),
// removes btX fields, which (when converted to \note: btX....\note* by current OW_to_PT cause
// Nathan's USFM->OSIS to drop the rest of the verse after a footnote.
//Utils.GetUtilityFile("remove_bt_from_OW.cct"),
"bt",
// removes \ov fields, which otherwise tend to result in a newline before various notes,
// which becomes an unwanted space after following stages.
//Utils.GetUtilityFile("remove_ov_from_OW.cct"),
"ov",
// Strip all the \ntX fields. JohnD's file makes of all these markers that don't translate into \note.
// The OSIS converter discards them, but the resulting blank lines mess up spacing of note markers.
//Utils.GetUtilityFile("remove_nt_from_OW.cct"),
"nt",
// Several more fields that are not wanted and not handled by the OW_to_PT.
"al ",
"e",
"chk2",
"bnvt",
"nq",
// Consider using this...if so, must be BEFORE we convert angle brackets to quotes!
//Utils.GetUtilityFile("fix_glottal.cct"),
// The main tables don't do well with multiple foonotes in a block. Now we've got rid of the \bts, we can break those up.
"footnotes.process",
// OW_to_PT.cct does not seem to get the quotes quite right. Kupang makes use of <<< and >>> which
// are ambiguous; OW_to_PT converts << and >> and < and >, but >>> is therefore interpreted as >> >
// and closes the double first, which is (usually) wrong.
// This version removes any space in >> > etc, and interprets >>> as > >>.
// This change may be redundant with the latest version of JohnD's OW_to_PT.cct
//Utils.GetUtilityFile("fix_quotes.cct"),
// Main conversion by John Duerkson implemented in these two files.
Utils.GetUtilityFile("OW_to_PT.cct"),
Utils.GetUtilityFile("move_footnote_to_fn.cct"),
// Strip all the \note fields that JohnD's file makes of all the markers that don't translate.
// The OSIS converter discards them, but the resulting blank lines mess up spacing of note markers.
// Didn't work...strips the whole file content after the first \note. Also not needed, now have
// \nt being deleted properly.
//Utils.GetUtilityFile("remove_note_from_USFM.cct"),
// Final cleanup strips remnants of s2 markers at end of field, and puts cross-ref notes inline so
// we don't get a spurious space before the <note> in the OSIS and beyond.
Utils.GetUtilityFile("cleanup_OW_to_USFM.cct")
};
}
// Reset problem records, in case ever used repeatedly.
m_ProblemMarkers.Clear();
m_problemReports.Remove(0, m_problemReports.Length);
m_seriousErrorCount = 0;
// Input is typically db files from OurWord, sfm from TE, or ptx from ParaText.
// It may also be from Bibledit or another USFM tool. Many archived input files
// do not have consistent file extensions. Some are named for the language as a
// suffix. Some use a book abbreviation as an extension.
// Bibledit uses .usfm on its exports.
string[] inputFileNames = Directory.GetFiles(m_inputDirName);
if (inputFileNames.Length == 0)
{
MessageBox.Show("No files found in input directory " + m_inputDirName);
return;
}
// Progress status = new Progress(inputFileNames.Length);
// status.Show();
Utils.DeleteDirectory(m_outputDirName);
Utils.EnsureDirectory(m_outputDirName);
int count = 0;
foreach (string inputFile in inputFileNames)
{
string filename = Path.GetFileName(inputFile);
string fileType = Path.GetExtension(filename).ToUpper();
if ((fileType != ".BAK") && (fileType != ".LDS") &&
(fileType != ".SSF") && (fileType != ".DBG") &&
(fileType != ".WDL") && (fileType != ".STY") &&
(!inputFile.EndsWith("~")))
{
// status.File = filename;
fileDate = File.GetLastWriteTimeUtc(inputFile);
if (fileDate > sourceDate)
{
sourceDate = fileDate;
}
if (!Convert(inputFile))
{
break;
}
}
count++;
// status.Value = count;
}
// status.Close();
if (m_ProblemMarkers.Count > 0 || m_seriousErrorCount > 0)
{
StringBuilder report = new StringBuilder();
if (m_seriousErrorCount > 0)
{
report.Append("Serious problems occurred while processing at least one file, indicated by lines beginning **** ERROR in the output\n");
report.Append("These errors result in the output being an incomplete conversion of the input\n\n");
}
if (m_ProblemMarkers.Count > 0)
{
List <string> keys = new List <string>(m_ProblemMarkers.Keys);
keys.Sort();
report.Append("The conversion process encountered some markers it could not handle.\n");
report.Append("These are indicated in the output files by lines starting ***\\.\n");
report.Append("The problem markers are these: ");
bool fFirst = true;
foreach (string key in keys)
{
if (fFirst)
{
fFirst = false;
}
else
{
report.Append("; ");
}
report.Append(String.Format("{0} ({1})", key, m_ProblemMarkers[key].ToString()));
}
report.Append(".\n\n");
}
report.Append("Here are the lines where the problems were found:\n");
report.Append(m_problemReports.ToString());
ProblemReport reportDlg = new ProblemReport();
reportDlg.ReportContents = report.ToString();
reportDlg.ShowDialog();
}
}
/// <summary>
/// Run the algorithm (on all files).
/// </summary>
public void Run(IList files)
{
Init();
m_toolPath = ToolPath;
if (CheckToolPresent && !File.Exists(m_toolPath))
{
MessageBox.Show("The program needed for this conversion was not found: " + m_toolPath, "Error");
return;
}
Utils.EnsureDirectory(m_outputDirName);
string extension = null;
foreach (string pattern in Extensions)
{
if (Directory.GetFiles(m_inputDirName, pattern).Length != 0)
{
extension = Path.GetExtension(pattern);
break;
}
}
if (extension == null)
{
MessageBox.Show("No matching input files");
}
List <string> inputFilePaths = new List <string>();
foreach (string patternFile in files)
{
string mainFilePath = Path.Combine(m_inputDirName, Path.ChangeExtension(patternFile, extension));
foreach (string filepath in GetActualFileNames(mainFilePath))
{
if (!CheckFileExists(filepath))
{
return;
}
inputFilePaths.Add(filepath);
}
}
int count = 0;
m_reportPath = Path.Combine(m_inputDirName, "ConversionReports.txt");
File.Delete(m_reportPath); // get rid of any old reports
m_errorInfo = null;
m_outputWriter = new StreamWriter(m_reportPath, true, Encoding.UTF8);
Progress status = new Progress(inputFilePaths.Count * 2);
status.Show();
foreach (string inputFile in inputFilePaths)
{
PreProcess(inputFile);
}
foreach (string inputFile in inputFilePaths)
{
string filename = Path.GetFileName(inputFile);
status.File = filename;
ConvertFile(inputFile);
count++;
status.Value = count;
}
// PostProcessing is done separately after all the files are created. This is important for OSIS_to_HTML
// in chapter-per-file mode, which needs to look ahead to the next book to determine what the first chapter
// file should be to link to.
foreach (string inputFile in inputFilePaths)
{
string filename = Path.GetFileName(inputFile);
status.File = filename;
PostProcess(inputFile);
count++;
status.Value = count;
}
status.Close();
m_outputWriter.Close();
if (m_errorInfo != null)
{
ProblemReport reportDlg = new ProblemReport();
reportDlg.ReportContents = m_errorInfo.ToString();
reportDlg.Show();
}
}
