public static MemoryStream RenderPage_AsMemoryStream(string pdf_filename, int page_number, int dpi, string device, ProcessPriorityClass priority_class)
{
// STDOUT/STDERR
string ghostscript_parameters = GhostscriptBinaries.GenerateGhostscriptParameters(pdf_filename, device, dpi, page_number, 0, @"-");
using (Process process = GhostscriptBinaries.StartGhostscriptProcess(ghostscript_parameters, priority_class))
{
Logging.Info("Process started!");
// Read image from stdout
using (ProcessOutputReader process_output_reader = new ProcessOutputReader(process, stdout_is_binary: true))
{
using (StreamReader sr = process.StandardOutput)
{
using (FileStream fs = (FileStream)sr.BaseStream)
{
MemoryStream ms = new MemoryStream(128 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, ms);
Logging.Debug特("Image size was {0} for PDF file {1}, page {2} @ dpi {3}", total_size, pdf_filename, page_number, dpi);
// Check that the process has exited properly
process.WaitForExit(1000);
if (!process.HasExited)
{
Logging.Error("Ghostscript process did not terminate.\n{0}", process_output_reader.GetOutputsDumpString());
}
return(ms);
}
}
}
}
}
private static MemoryStream ReadEntireStandardOutput(string process_parameters, ProcessPriorityClass priority_class)
{
Process process = ProcessSpawning.SpawnChildProcess("pdfdraw.exe", process_parameters, priority_class);
process.ErrorDataReceived += (sender, e) => { };
process.BeginErrorReadLine();
// Read image from stdout
StreamReader sr = process.StandardOutput;
FileStream fs = (FileStream)sr.BaseStream;
MemoryStream ms = new MemoryStream(128 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, ms);
// Check that the process has exited properly
process.WaitForExit(1000);
if (!process.HasExited)
{
Logging.Error("PDFRenderer process did not terminate, so killing it");
try
{
Logging.Info("Killing PDFRenderer process");
process.Kill();
Logging.Info("Killed PDFRenderer process");
}
catch (Exception)
{
Logging.Error("These was an exception while trying to kill the PDFRenderer process");
}
}
return(ms);
}
private static MemoryStream ReadEntireStandardOutput(string process_parameters, ProcessPriorityClass priority_class)
{
Stopwatch clk = Stopwatch.StartNew();
// STDOUT/STDERR
Logging.Debug("PDFDRAW :: ReadEntireStandardOutput command: pdfdraw.exe {0}", process_parameters);
using (Process process = ProcessSpawning.SpawnChildProcess("pdfdraw.exe", process_parameters, priority_class, stdout_is_binary: true))
{
using (ProcessOutputReader process_output_reader = new ProcessOutputReader(process, stdout_is_binary: true))
{
// Read image from stdout
using (StreamReader sr = process.StandardOutput)
{
using (FileStream fs = (FileStream)sr.BaseStream)
{
long elapsed = clk.ElapsedMilliseconds;
Logging.Debug("PDFDRAW :: ReadEntireStandardOutput setup time: {0} ms for parameters:\n {1}", elapsed, process_parameters);
MemoryStream ms = new MemoryStream(256 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, ms);
long elapsed2 = clk.ElapsedMilliseconds;
Logging.Debug("PDFDRAW image output {0} bytes in {1} ms (output copy took {2} ms) for command:\n pdfdraw.exe {3}", total_size, elapsed2, elapsed2 - elapsed, process_parameters);
// Check that the process has exited properly
process.WaitForExit(1000);
if (!process.HasExited)
{
Logging.Debug("PDFRenderer process did not terminate, so killing it.\n{0}", process_output_reader.GetOutputsDumpString());
try
{
process.Kill();
// wait for the completion signal; this also helps to collect all STDERR output of the application (even while it was KILLED)
process.WaitForExit(1000);
}
catch (Exception ex)
{
Logging.Error(ex, "There was a problem killing the PDFRenderer process after timeout ({0} ms)", elapsed2 + 1000);
}
Logging.Error("PDFRenderer process did not terminate, so killed it. Commandline:\n {0}\n{1}", process_parameters, process_output_reader.GetOutputsDumpString());
throw new ApplicationException($"PDFRenderer process did not terminate, so killed it.\n Commandline: pdfdraw.exe {process_parameters}");
}
else if (process.ExitCode != 0)
{
Logging.Error("PDFDRAW did fail with exit code {0} for commandline:\n {1}\n{2}", process.ExitCode, process_parameters, process_output_reader.GetOutputsDumpString());
throw new ApplicationException($"PDFRenderer::PDFDRAW did fail with exit code {process.ExitCode}.\n Commandline: pdfdraw.exe {process_parameters}");
}
return(ms);
}
}
}
}
}
public static MemoryStream RenderPage_AsMemoryStream(string pdf_filename, int page_number, int dpi, string device, ProcessPriorityClass priority_class)
{
// STDOUT/STDERR
string ghostscript_parameters = GhostscriptBinaries.GenerateGhostscriptParameters(pdf_filename, device, dpi, page_number, 0, @"-");
using (Process process = GhostscriptBinaries.StartGhostscriptProcess(ghostscript_parameters, priority_class))
{
Logging.Info("Process started!");
// Read image from stdout
using (ProcessOutputReader process_output_reader = new ProcessOutputReader(process, stdout_is_binary: true))
{
using (StreamReader sr = process.StandardOutput)
{
using (FileStream fs = (FileStream)sr.BaseStream)
{
MemoryStream ms = new MemoryStream(128 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, ms);
Logging.Debug特("Image size was {0} for PDF file {1}, page {2} @ dpi {3}", total_size, pdf_filename, page_number, dpi);
// Check that the process has exited properly
process.WaitForExit(1000);
bool has_exited = process.HasExited;
if (!has_exited)
{
try
{
process.Kill();
// wait for the completion signal; this also helps to collect all STDERR output of the application (even while it was KILLED)
process.WaitForExit(1000);
}
catch (Exception ex)
{
Logging.Error(ex, "There was a problem killing the GhostScript process after timeout");
}
}
// Check that we had a clean exit
if (!has_exited || 0 != process.ExitCode)
{
Logging.Error("Ghostscript process did not terminate.\n{0}", process_output_reader.GetOutputsDumpString());
}
return(ms);
}
}
}
}
}
public static MemoryStream RenderPage_AsMemoryStream(string pdf_filename, int page_number, int dpi, string device, ProcessPriorityClass priority_class)
{
string ghostscript_parameters = GhostscriptBinaries.GenerateGhostscriptParameters(pdf_filename, device, dpi, page_number, 0, @"-");
Process process = GhostscriptBinaries.StartGhostscriptProcess(ghostscript_parameters, priority_class);
Logging.Info("Process started!");
// Read image from stdout
StreamReader sr = process.StandardOutput;
FileStream fs = (FileStream)sr.BaseStream;
MemoryStream ms = new MemoryStream(128 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, ms);
Logging.Debug("Image size was {0}", total_size);
// Check that the process has exited properly
process.WaitForExit(1000);
if (!process.HasExited)
{
Logging.Error("Ghostscript process did not terminate");
}
return(ms);
}
public static void AddNewDocumentToLibraryFromInternet_SYNCHRONOUS(Library library, string download_url)
{
StatusManager.Instance.UpdateStatus(LIBRARY_DOWNLOAD, String.Format("Downloading {0}", download_url));
try
{
HttpWebRequest web_request = (HttpWebRequest)HttpWebRequest.Create(new Uri(download_url));
web_request.Proxy = ConfigurationManager.Instance.Proxy;
web_request.Method = "GET";
web_request.AllowAutoRedirect = true;
// https://stackoverflow.com/questions/21728773/the-underlying-connection-was-closed-an-unexpected-error-occurred-on-a-receiv
// also: https://stackoverflow.com/questions/21481682/httpwebrequest-the-underlying-connection-was-closed-the-connection-was-closed
web_request.KeepAlive = false;
// Allow ALL protocols
ServicePointManager.SecurityProtocol = SecurityProtocolType.Tls | SecurityProtocolType.Tls11 | SecurityProtocolType.Tls12 | SecurityProtocolType.Ssl3;
// same headers as sent by modern Chrome.
// Gentlemen, start your prayer wheels!
web_request.Headers.Add("Cache-Control", "no-cache");
web_request.Headers.Add("Pragma", "no-cache");
web_request.UserAgent = ConfigurationManager.Instance.ConfigurationRecord.GetWebUserAgent();
using (HttpWebResponse web_response = (HttpWebResponse)web_request.GetResponse())
{
// is this a 302/30x Response Code (Forwarded)?
// if so, then grab the forward reference URI and go grab that one.
if (web_response.StatusCode == HttpStatusCode.MovedPermanently ||
web_response.StatusCode == HttpStatusCode.Moved ||
web_response.StatusCode == HttpStatusCode.Redirect ||
web_response.StatusCode == HttpStatusCode.Found ||
web_response.StatusCode == HttpStatusCode.SeeOther ||
web_response.StatusCode == HttpStatusCode.RedirectKeepVerb ||
web_response.StatusCode == HttpStatusCode.TemporaryRedirect ||
(uint)web_response.StatusCode == 308)
{
string fwd_uri_str = web_response.GetResponseHeader("Location");
Uri fwd_uri = new Uri(web_response.ResponseUri, fwd_uri_str);
// fetch the PDF!
//
// Warning: Do NOT get into a download loop due to badly configured or nasty webservers:
if (fwd_uri.AbsoluteUri != web_request.RequestUri.AbsoluteUri)
{
AddNewDocumentToLibraryFromInternet_ASYNCHRONOUS(library, fwd_uri.AbsoluteUri);
}
else
{
MessageBoxes.Info("Looks like the webserver is throwing you into an infinite redirection loop at URI {0}.", web_request.RequestUri.AbsoluteUri);
}
}
else
{
using (Stream response_stream = web_response.GetResponseStream())
{
string content_type = web_response.GetResponseHeader("Content-Type");
// See also: https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Content-Disposition
string content_disposition = web_response.GetResponseHeader("Content-Disposition");
string original_filename = null;
try
{
if (!String.IsNullOrEmpty(content_disposition))
{
ContentDisposition contentDisposition = new ContentDisposition(content_disposition);
original_filename = contentDisposition.FileName;
//StringDictionary parameters = contentDisposition.Parameters;
}
else
{
Logging.Warn("AddNewDocumentToLibraryFromInternet: no or empty Content-Disposition header received from {1}?\n Headers:\n{0}", web_response.Headers, download_url);
// fallback: derive the filename from the URL:
original_filename = web_response.ResponseUri.LocalPath;
}
}
catch (Exception ex)
{
Logging.Error(ex, "AddNewDocumentToLibraryFromInternet: no Content-Disposition header received from {1}?\n Headers:\n{0}", web_response.Headers, download_url);
// fallback: derive the filename from the URL:
original_filename = web_response.ResponseUri.LocalPath;
}
// extract the type from the Content-Type header value:
try
{
if (!String.IsNullOrEmpty(content_type))
{
ContentType ct = new ContentType(content_type);
content_type = ct.MediaType.ToLower(CultureInfo.CurrentCulture);
}
else
{
Logging.Warn("AddNewDocumentToLibraryFromInternet: no or empty Content-Type header '{2}' received from {1}?\n Headers:\n{0}", web_response.Headers, download_url, content_type);
content_type = "text/html";
}
}
catch (Exception ex)
{
Logging.Error(ex, "AddNewDocumentToLibraryFromInternet: no or invalid Content-Type header '{2}' received from {1}?\n Headers:\n{0}", web_response.Headers, download_url, content_type);
content_type = "text/html";
}
bool is_acceptable_content_type = false;
if (content_type.EndsWith("pdf"))
{
is_acceptable_content_type = true;
}
if (content_type.StartsWith("application/octet-stream"))
{
is_acceptable_content_type = true;
}
if (is_acceptable_content_type)
{
string filename = TempFile.GenerateTempFilename("pdf");
using (FileStream fs = File.OpenWrite(filename))
{
int total_bytes = StreamToFile.CopyStreamToStream(response_stream, fs);
Logging.Info("Saved {0} bytes to {1}", total_bytes, filename);
//fs.Close(); -- autoclosed by `using` statement
}
PDFDocument pdf_document = library.AddNewDocumentToLibrary_SYNCHRONOUS(filename, original_filename, download_url, null, null, null, false, false);
File.Delete(filename);
// make sure we open every PDF fetched off the Internet: the user may need to review
// their metadata.
MainWindowServiceDispatcher.Instance.MainWindow.Dispatcher.InvokeAsync
(
new Action(() =>
{
Documents.PDF.PDFControls.PDFReadingControl pdf_reading_control = MainWindowServiceDispatcher.Instance.OpenDocument(pdf_document);
pdf_reading_control.EnableGuestMoveNotification(null);
}),
DispatcherPriority.Background
);
}
else
{
string html = "";
if (content_type.EndsWith("html"))
{
using (StreamReader sr = new StreamReader(response_stream))
{
html = sr.ReadToEnd();
Logging.Warn("Got this HTML instead of a PDF for URI {1}: {0}", html, download_url);
}
}
// TODO: check these conditions; they are meant to be pretty tight but MAYBE I still let some
// nasty websites' embedded PDF or other trickery slip through unnoticed.
bool tolerate_type = false;
foreach (string t in content_types_to_tolerate)
{
if (content_type.Contains(t))
{
tolerate_type = true;
}
}
if (!tolerate_type || web_response.StatusCode != HttpStatusCode.OK || html.Contains("<embed"))
{
MessageBoxes.Info("The document library supports only PDF files at the moment. You are trying to download something of type {0} / response code {1} at URI {2}.", content_type, (uint)web_response.StatusCode, download_url);
}
}
}
}
}
}
catch (Exception ex)
{
Logging.Error(ex, "There was a problem adding the downloaded PDF to the library for URI {0}.", download_url);
}
StatusManager.Instance.UpdateStatus(LIBRARY_DOWNLOAD, String.Format("Downloaded {0}", download_url));
}
public static void AddNewDocumentToLibraryFromInternet_SYNCHRONOUS(Library library, object download_url_obj)
{
string download_url = (string)download_url_obj;
StatusManager.Instance.UpdateStatus(LIBRARY_DOWNLOAD, String.Format("Downloading {0}", download_url));
try
{
HttpWebRequest web_request = (HttpWebRequest)HttpWebRequest.Create(download_url);
web_request.Proxy = ConfigurationManager.Instance.Proxy;
web_request.Method = "GET";
web_request.AllowAutoRedirect = true;
using (HttpWebResponse web_response = (HttpWebResponse)web_request.GetResponse())
{
if (false)
{
}
if (HttpStatusCode.Redirect == web_response.StatusCode)
{
string redirect_url = web_response.Headers["Location"];
}
else
{
Stream response_stream = web_response.GetResponseStream();
string content_type = web_response.GetResponseHeader("Content-Type");
bool is_acceptable_content_type = false;
if (content_type.ToLower(CultureInfo.CurrentCulture).EndsWith("pdf"))
{
is_acceptable_content_type = true;
}
if (content_type.ToLower(CultureInfo.CurrentCulture).StartsWith("application/octet-stream"))
{
is_acceptable_content_type = true;
}
if (is_acceptable_content_type)
{
string filename = TempFile.GenerateTempFilename("pdf");
using (FileStream fs = File.OpenWrite(filename))
{
int total_bytes = StreamToFile.CopyStreamToStream(response_stream, fs);
Logging.Info("Saved {0} bytes to {1}", total_bytes, filename);
fs.Close();
}
library.AddNewDocumentToLibrary_SYNCHRONOUS(filename, download_url, null, null, null, false, false);
File.Delete(filename);
}
else
{
if (content_type.ToLower(CultureInfo.CurrentCulture).EndsWith("html"))
{
StreamReader sr = new StreamReader(response_stream);
string html = sr.ReadToEnd();
Logging.Warn("Got this HTML instead of a PDF: {0}", html);
}
MessageBoxes.Info("The document library supports only PDF files at the moment. You are trying to download something of type {0}.", content_type);
}
}
}
}
catch (Exception ex)
{
Logging.Error(ex, "There was a problem adding the downloaded PDF to the library.");
}
StatusManager.Instance.UpdateStatus(LIBRARY_DOWNLOAD, String.Format("Downloaded {0}", download_url));
}
private static ExecResultAggregate ReadEntireStandardOutput(string pdfDrawExe, string process_parameters, bool binary_output, ProcessPriorityClass priority_class)
{
WPFDoEvents.AssertThisCodeIs_NOT_RunningInTheUIThread();
ExecResultAggregate rv = new ExecResultAggregate
{
executable = pdfDrawExe,
process_parameters = process_parameters,
stdoutIsBinary = binary_output
};
Stopwatch clk = Stopwatch.StartNew();
// STDOUT/STDERR
Logging.Debug("PDFDRAW :: ReadEntireStandardOutput command: pdfdraw.exe {0}", process_parameters);
using (Process process = ProcessSpawning.SpawnChildProcess(pdfDrawExe, process_parameters, priority_class, stdout_is_binary: true))
{
using (ProcessOutputReader process_output_reader = new ProcessOutputReader(process, stdout_is_binary: true))
{
// Read image from stdout
using (StreamReader sr = process.StandardOutput)
{
using (FileStream fs = (FileStream)sr.BaseStream)
{
long elapsed = clk.ElapsedMilliseconds;
Logging.Debug("PDFDRAW :: ReadEntireStandardOutput setup time: {0} ms for parameters:\n {1}", elapsed, process_parameters);
rv.stdoutStream = new MemoryStream(1024 * 1024);
int total_size = StreamToFile.CopyStreamToStream(fs, rv.stdoutStream);
long elapsed2 = clk.ElapsedMilliseconds;
Logging.Debug("PDFDRAW image output {0} bytes in {1} ms (output copy took {2} ms) for command:\n {4} {3}", total_size, elapsed2, elapsed2 - elapsed, process_parameters, pdfDrawExe);
// Check that the process has exited properly
process.WaitForExit(1000);
if (!process.HasExited)
{
Logging.Debug("PDFRenderer process did not terminate, so killing it.\n{0}", process_output_reader.GetOutputsDumpStrings().stderr);
try
{
process.Kill();
// wait for the completion signal; this also helps to collect all STDERR output of the application (even while it was KILLED)
process.WaitForExit(3000);
}
catch (Exception ex)
{
Logging.Error(ex, "There was a problem killing the PDFRenderer process after timeout ({0} ms)", elapsed2 + 1000);
}
// grab stderr output for successful runs and log it anyway: MuPDF diagnostics, etc. come this way:
var outs = process_output_reader.GetOutputsDumpStrings();
rv.errOutputDump = outs;
Logging.Error($"PDFRenderer process did not terminate, so killed it. Commandline:\n {pdfDrawExe} {process_parameters}\n{outs.stderr}");
rv.error = new ApplicationException($"PDFRenderer process did not terminate, so killed it.\n Commandline: {pdfDrawExe} {process_parameters}");
rv.exitCode = 0;
if (process.HasExited)
{
rv.exitCode = process.ExitCode;
}
if (rv.exitCode == 0)
{
rv.exitCode = -666; // timeout
}
}
else if (process.ExitCode != 0)
{
// grab stderr output for successful runs and log it anyway: MuPDF diagnostics, etc. come this way:
var outs = process_output_reader.GetOutputsDumpStrings();
rv.errOutputDump = outs;
Logging.Error("PDFDRAW did fail with exit code {0} for commandline:\n {3} {1}\n{2}", process.ExitCode, process_parameters, outs.stderr, pdfDrawExe);
rv.error = new ApplicationException($"PDFRenderer::PDFDRAW did fail with exit code {process.ExitCode}.\n Commandline: {pdfDrawExe} {process_parameters}");
rv.exitCode = process.ExitCode;
}
else
{
// grab stderr output for successful runs and log it anyway: MuPDF diagnostics, etc. come this way:
var outs = process_output_reader.GetOutputsDumpStrings();
rv.errOutputDump = outs;
Logging.Error("PDFDRAW did SUCCEED with exit code {0} for commandline:\n {3} {1}\n{2}", process.ExitCode, process_parameters, outs.stderr, pdfDrawExe);
rv.error = null;
rv.exitCode = process.ExitCode;
}
return(rv);
}
}
}
}
}
