/**********************************************************************
*** Processing Functions
**********************************************************************/
private void ProcessFile(string filePath)
{
_filePath = filePath;
UpdateDialogTitle();
Cursor.Current = Cursors.WaitCursor;
// read the OpenOffice document into memory; if the user has the same file still open
// in OpenOffice this operation fails with an exception; we then giv ehim the chance to
// close the document and retry.
bool retry = false;
_doc = new OpenOfficeDoc();
do
{
try
{
stsAction.Text = "Reading file";
if (!_doc.Read(filePath))
{
return;
}
break;
}
catch (Exception ex)
{
DialogResult res = MessageBox.Show("Error: Could not read file from disk. Original error: " + ex.Message,
"File Open Error", MessageBoxButtons.RetryCancel);
if (res == DialogResult.Retry)
{
retry = true;
}
else
{
stsAction.Text = "Processing aborted.";
Cursor.Current = Cursors.Default;
return;
}
}
} while (retry);
// extract the hyperlinks from the document
stsAction.Text = "Extracting hyperlinks";
ClearAllLinks();
if (!_doc.ExtractLinks(_enLinks))
{
Cursor.Current = Cursors.Default;
return;
}
// translate these hyperlinks
stsAction.Text = "Translating hyperlinks";
_lt.TranslateLinks2(_enLinks, _deLinks);
// show the links and their translations
_curIdx = 0;
ShowCurrentLink();
stsAction.Text = "Done";
Cursor.Current = Cursors.Default;
appendLinksToDocumentToolStripMenuItem.Enabled = true;
}
/// <summary>
/// Collect links and their translations from an OpenOffice Writer file.
/// We assume that the files contains first the English text, then the translated
/// German text. We build on the fact that the sequence of hypeslinks is unchanged
/// in the translation so that we can associate each link with its translated version.
/// </summary>
/// <param name="filePath">File path to the input document.</param>
private void CollectLinksFromDocument(string filePath)
{
OpenOfficeDoc doc = new OpenOfficeDoc();
if (!doc.Read(filePath))
{
return;
}
List <Hyperlink> links = new List <Hyperlink> ();
if (!doc.ExtractLinks(links))
{
return;
}
if (!doc.ContainsLangSeparator())
{
return;
}
// fold link list in half and treat the second half as translations
// of the first half
if (links.Count % 2 != 0)
{
_msgStream.WriteLine("{0}: Number of links in document is not even! Skipping this document.",
filePath);
return;
}
int splitIndex = links.Count / 2;
int remaining = links.Count - splitIndex;
List <Hyperlink> deLinks = links.GetRange(splitIndex, remaining);
links.RemoveRange(splitIndex, remaining);
// loop over both lists in sync and record potential translation candidates in
// text lines of the format
//
// <english-text>; <english-link>; <german-text>; <german-link>
//
// We only collect links to Wikipedia.org!
//
for (int i = 0; i < splitIndex; ++i)
{
Hyperlink enLink = links[i];
Hyperlink deLink = deLinks[i];
Uri uri = new Uri(enLink.uri);
if (uri.Authority == "en.wikipedia.org")
{
_outStream.WriteLine("{0};{1};{2};{3}",
enLink.text, enLink.uri, deLink.text, deLink.uri);
}
}
}