I have a Delphi5 application which exports a file (.pdf) and a very small metadata file to a network location. The intention is that these 2 files should be processed, and then removed, by a polling .NET application.
My approach is to
Write the metadata file with the extension '.part'
Generate the .pdf
Rename the .part file to .dat
The .NET process is looking for files with the extension '.dat' only, so I would expect there to be no conflict between the 2 reader/writers. However, the .NET process is occasionally logging the following error ...
System.IO.IOException: The process cannot access the file '\\server\Path\FileName.dat' because it is being used by another process.
(I say occasionally - we are currently testing, so when volumes increase this may become much more of an issue)
The Delphi code looks like this :
AssignFile(FTextFile, Format('%s\%s.part', [DMSPath, FullFileName]));
try
try
ReWrite(FTextFile);
Writeln(FTextFile, MetaDataString);
finally
CloseFile(FTextFile);
end;
except
raise ELogFileException.Create( LOGFILEWRITEFAILURE );
end;
Then there is a separate method which performs the following lines of code
if FindFirst(Format('%s\*.part',[DMSPath]), faAnyFile, SearchRec) = 0 then begin
repeat
OldName := Format('%s\%s',[DMSPath, SearchRec.Name]);
NewName := Format('%s\%s',[DMSPath, ChangeFileExt(SearchRec.Name, '.dat')]);
RenameFile(OldName, NewName);
until FindNext(SearchRec) <> 0;
FindClose(SearchRec);
end;
I cannot see anything inherently wrong with this code and we have a couple of remedies in mind, but I have 2 questions
Should I try a different technique to more reliably protect the '.dat' file until it is fully ready
What circumstances could be causing this?
So far there has been one suggested cause - Antivirus software.
Any suggestions as to how the file might be produced differently? Note that my application is Delphi5; I wondered if there was a newer, more 'atomic' version of the 'MoveFileA' WinApi call I could use.
In the past, we had a problem with file being locked like that. Investigation pointed to Windows' Prefetch. The file being affected were not on a network directory though.
As far as I know, Prefetch only work on process startup and/or while booting (Controlled by a registry key), so it might not apply to your current situation.
You can check the "C:\"Windows"\Prefetch\" directory. If prefetch is active, it should contains multiple *.pf files. If there is one with your executable filename, it might be worth investigating.
Personally speaking, because there are multiple files involved, I'd create a separate lock file (eg, myfile.lck) that you write first. If the polling app sees it in the folder, then it stops looking for other files. Once that file is gone, then it dives deeper. I don't know if this would solve the problem you're encountering or not, but I'd give it a try. (Files with .dat extensions are frequently created by malicious evildoers, so they can raise spurious issues through other sources, like AV software. A lock file with 0 bytes in it is generally harmless and disregarded.)
Related
I start using a TFileStream and TStreamWriter to write simple text logfiles (instead of old Writeln(T,....)). And I have multiple applicatiosn writing to the same logfile.
Each appplication has its own TFileStream of course and they each open the file like this
FFileStream:=TFileStream.Create(LogName, fmOpenReadWrite+fmShareDenyNone)
FExporter:=TStreamWriter.Create(FFilestream, TEncoding.UTF8);
FExporter.NewLine:=#$0A;
FExporter.AutoFlush:=TRUE;
and write to the file with
FExporter.BaseStream.Seek(0, soFromEnd);
FExporter.Write('['+DateToStr(Now, FDateTimeFormat)+'] ['+TimeToStr(Now, FDateTimeFormat)+'] [#'+Lead0(GetCurrentThreadId, 5)+']: '+EntryText);
FExporter.WriteLine;
the result is somewhat "unsatisfactory" as the lines are displaced, empty lines in between and does not seem to work.
HOW would I do that correctly?
Writing multiples lines at the same time in multiples process may result in unexpected continue, because parallels execution.
You should assure that you are writing a block continually so WriteLine shoud be send inside the write using lineBreak at the end.
So the way you can write should be:
FExporter.BaseStream.Seek(0, soFromEnd);
FExporter.Write('['+DateToStr(Now, FDateTimeFormat)+'] ['+TimeToStr(Now, FDateTimeFormat)+'] [#'+Lead0(GetCurrentThreadId, 5)+']: '+EntryText + System.slineBreak);
//FExporter.WriteLine;
Update1:
As the link Oliver posted, sometime it can not work if the message size to be written is bigger than the OS file sector and, at that very moment, other process also try to write a message. Thus in this case the result content might be mixed.
So doing what I first purpose you would increase the probability to have the desired result, but may not be the solution in 100% of the cases.
To be 100% sure of writing continuous log in a single file, using multiples process, you should create a log process to receive a message from the others and to be the only responsible for writing synchronized log throughout threads.
One of my users at a large university (with, I imagine, the aggressive security settings that university IT departments general have on their computers) is getting an empty string returned by Windows XP for CSIDL_COMMON_APPDATA or CSIDL_PERSONAL. (I'm not sure which of these is returning the empty string, because I haven't yet examined his computer to see how he's installed the software, but I'm pretty sure it's the COMMON_APPDATA...)
Has anyone encountered this or have suggestions on how to deal with this?
Here's the Delphi code I'm using to retrieve the value:
Function GetSpecialFolder( FolderID: Integer):String;
var
PIDL: PItemIDList;
Path: array[0..MAX_PATH] of Char;
begin
SHGetSpecialFolderLocation(Application.Handle, FolderID, PIDL);
SHGetPathFromIDList(PIDL, Path);
Result := Path;
end; { GetSpecialFolder }
ShowMessage(GetSpecialFolder(CSIDL_COMMON_APPDATA)); <--- This is an empty string
Edit:
Figuring out this API made me feel like I was chasing my tail - I went in circles trying to find the right call. This method and others similar to it are said to be deprecated by Microsoft (as well as by a earlier poster to this question (#TLama?) who subsequently deleted the post.) But, it seems like most of us, including me, regularly and safely ignore that status.
In my searches, I found a good answer here on SO from some time ago, including sample code for the non-deprecated way of doing this: what causes this error 'Unable to write to application file.ini'.
If you want to find out why an API call is failing you need to check the return values. That's what is missing in this code.
You need to treat each function on its own merits. Read the documentation on MSDN. In the case of SHGetSpecialFolderLocation, the return value is an HRESULT. For SHGetPathFromIDList you get back a BOOL. If that is FALSE then the call failed.
The likely culprit here is SHGetSpecialFolderLocation, the code that receives the CSIDL, but you must check for errors whenever you call Windows API functions.
Taking a look at the documentation for CSIDL we see this:
CSIDL_COMMON_APPDATA
Version 5.0. The file system directory that contains application data for all users. A typical path is C:\Documents and Settings\All
Users\Application Data. This folder is used for application data that
is not user specific. For example, an application can store a
spell-check dictionary, a database of clip art, or a log file in the
CSIDL_COMMON_APPDATA folder. This information will not roam and is
available to anyone using the computer.
If the machine has a shell version lower than 5.0, then this CSIDL value is not supported. That's the only documented failure mode for this CSIDL value. I don't think that applies to your situation, so you'll just have to see what the HRESULT status code has to say.
I am debugging a DirectShow filter I created with the DSPACK code library using Delphi 6 Pro. When a breakpoint I set is hit in one particular unit named BaseClass.pas, and I begin tracing, the Execution Point jumps to strange places in the source code. This usually indicates that the source code being traced does not match the source code that was compiled into one of the packages being used by the Delphi application. Oddly enough it is only the BaseClass unit since I have traced other units belonging to the DSPACK code library and they do not exhibit this problem. I am not using run-time packages.
I scanned my disk and found only one copy of BaseClass.dcu with a modification date equal to the last time I built the program. I have not modified the source for that unit or any other belonging to DSPACK. Since my Filter is part of the main application this indicates that BaseClass.pas would be subject to a dual use situation since it is used to build the DSPACK component package (dpk), and is also referenced by my main application directly via the TBCSource object my Filter descends from. Note, I did try adding the unit PAS file directly to my Project but that didn't fix anything.
I also went back and re-opened each of the DSPACK package files and did a full re-build. None of this helped. Is there something else I can try to get the source synchronized with the compiled image of the BaseClass unit? Or is a different problem altogether and if so, what is it and how can I fix it?
Sometimes this happens when code is copied/pasted from web pages or other sources, and the lines don't end with CR/LF pairs (#13#10 or 0x0D0A, standard for Windows) but end in only LF (#10 or 0x0A, typically the line ending in *nix systems) or CR (#13 or 0x0D, typical with Mac OSX/iOS). The incorrect line terminators confuse the debugger - this has been an issue for the past several Delphi versions.
You can sometimes fix this by opening the source file using a text editor like Notepad, making a small meaningless change (insert and then delete a blank line, for instance), and then save the file.
I had same problem and made a similar utility. Fixed it.
Basically, just this:
procedure adjustCRLF(filename : String);
var
strList : TStringList;
begin
strList := TStringList.Create;
try
strList.LoadFromFile(filename);
strList.Text := AdjustLineBreaks(strList.Text);
strList.SaveToFile(filename);
finally
strList.Free;
end;
end;
There is another way this can happen: if the IDE erroneously opens another source file with the same name (but different, such as an earlier version) then all the debug points will be incorrect, and the debugger will even allow you to step through the incorrect file.
I've seen Delphi 7 do this once.
Make sure that when you rebuild it, that in the compiler options for your project that you have "Debug Information" turned on. In fact, most of the options under Debugging should be set in your project's Compiler options.
Also, if you haven't already, restart Delphi.
The archives I'm having trouble with were all created by merging a working archive with a non-existant archive, thereby effectively copying the contents of one into the other. It's part of a merging process we do. Like this...
ZipDestination := TZipForge.Create(nil);
if FileExists(DestinationZipFileName) then
ZipDestination.OpenArchive(fmOpenReadWrite + fmShareDenyWrite)
else
ZipDestination.OpenArchive(fmCreate);
ZipDestination.Zip64Mode := zmAuto;
ZipDestination.MergeWith(SourceZipFileName);
ZipDestination.CloseArchive;
and this is the code that gets a blob from the archive, uncompresses it, and makes it ready for the viewer.
CompressedStream := TMemoryStream.Create;
UnCompressedStream := TMemoryStream.Create;
GetCompressedStream(CompressedStream); // this fetches the blob from the zipfile
ZipForge.InMemory := True;
// Native Error 00035 on next line (sometimes)
ZipForge.OpenArchive(CompressedStream, False);
ZipForge.FindFirst('*.*', ArchiveItem, faAnyFile - faDirectory);
sZipFileName := ArchiveItem.FileName;
sZipPath := ArchiveItem.StoredPath;
ZipForge.ExtractToStream(sZipPath + sZipFileName, UnCompressedStream);
ZipForge.CloseArchive;
but I'm encountering "Native error 00035" sometimes.
Now the strange thing is that I'm getting these errors when I try to view the first blob within the merged archive (ie. trying to view other blobs within the merged archive doesn't raise any exception)
It could be something about ZipForge.MergeWith that I haven't catered for, or it could be a bug in my GetCompressedStream (but if I switch the order of blobs within the archive, it always happens to the first one only). Look like it's time for a test project to see what's really going on.
EDIT
Original question was simply asking for guidance on these Native Errors, for which I'm satisfied with the answer I've chosen. As for my problem, well I'm convinced it's an issue with the CompressedStream I'm passing into OpenArchive.
Native error 00035 is "Invalid archive file". It occurs when ZipForge can't find either the local or central directory headers (that is, when you try to open a file that isn't a zip).
I don't think they're documented in the help, but the translation tables for native error to error code occur in ZFConst.pas. There's a NativeToErrorCode table that converts from the "native" error into an index in the error string array. If that isn't enough to tell you what the problem is just look through ZipForge.pas for the error code in a raise statement. They consistently use the full 5-digit code, so you can search for 00035 instead of just 35 to avoid spurious results.
Free support offered from the ZipForge vendor http://componentace.com/help/zf_guide/gettinghelpfromtechnicalsu.htm
What tools would you recommend for unit testing in Delphi.
I've used FastMM4 for memoryleak testing.
And MadExcept, both rule, but it doesn't help me test the logic in my programs.
I would like some alternatives, so don't all rush to suggest DUnit :-).
Any suggestions?
The most commonly used is DUnit. It's actually included in modern versions of Delphi but if your version doesn't come with it you can download it from Sourceforge.
DUnit comes with D2007.
-->File -->New -->Other
Select Unit Test from the dialogue that pops up.
There are a couple of good demo videos on it's use, I'll see if I can dig one up.
This is a pretty good one and others come up on the right:
http://www.youtube.com/watch?v=nyZnfxDqThE
You could take a look at the unit testing classes available in our SynCommons open source unit. It's used in our Open-Source framework for all regression tests. It's perhaps not the best, but it's worth taking a look at it. See http://blog.synopse.info/post/2010/07/23/Unit-Testing-light-in-Delphi
In the up-to-come 1.13 version, there is also a new logging mechanism with stack trace of any raised exception and such, just like MadExcept, using .map file content as source.
It's now used by the unit testing classes, so that any failure will create an entry in the log with the source line, and stack trace:
C:\Dev\lib\SQLite3\exe\TestSQL3.exe 0.0.0.0 (2011-04-13)
Host=Laptop User=MyName CPU=2*0-15-1027 OS=2.3=5.1.2600 Wow64=0 Freq=3579545
TSynLogTest 1.13 2011-04-13 05:40:25
20110413 05402559 fail TTestLowLevelCommon(00B31D70) Low level common: TDynArray "" stack trace 0002FE0B SynCommons.TDynArray.Init (15148) 00036736 SynCommons.Test64K (18206) 0003682F SynCommons.TTestLowLevelCommon._TDynArray (18214) 000E9C94 TestSQL3 (163)
The difference between a test suit without logging and a test suit with logging is only this:
procedure TSynTestsLogged.Failed(const msg: string; aTest: TSynTestCase);
begin
inherited;
with TestCase[fCurrentMethod] do
fLogFile.Log(sllFail,'%: % "%"',
[Ident,TestName[fCurrentMethodIndex],msg],aTest);
end;
The logging mechanism can be used to trace recursive calls. It can use an interface-based mechanism to log when you enter and leave any method:
procedure TMyDB.SQLExecute(const SQL: RawUTF8);
var ILog: ISynLog;
begin
ILog := TSynLogDB.Enter(self,'SQLExecute');
// do some stuff
ILog.Log(sllInfo,'SQL=%',[SQL]);
end; // when you leave the method, it will write the corresponding event to the log
It will be logged as such:
20110325 19325801 + MyDBUnit.TMyDB(004E11F4).SQLExecute
20110325 19325801 info SQL=SELECT * FROM Table;
20110325 19325801 - MyDBUnit.TMyDB(004E11F4).SQLExecute
Here the method name is set in the code ('SQLExecute'). But if you have an associated .map file, the logging mechanism is able to read this symbol information, and write the exact line number of the event.
Note that by default you have time and date written to the log, but it's also possible to replace this timing with high-resolution timestamps. With this, you'll be able to profile your application with data coming from the customer side, on its real computer. Via the Enter method (and its auto-Leave feature), you have all information needed for this. ;)
Like this:
0000000000000B56 + TTestCompression(00AB3570).000E6C79 SynSelfTests.TTestCompression.TestLog (376)
0000000000001785 - TTestCompression(00AB3570).000E6D09 SynSelfTests.TTestCompression.TestLog (385)
I still need to write some tool to compute the profiling, but there is already a dedicated TSynLogFile class able to read the .log file, and recognize its content.
The first time the .map file is read, a .mab file is created, and will contain all symbol information needed. You can send the .mab file with the .exe to your client, or even embed its content to the .exe. This .mab file is optimized: a .map of 927,984 bytes compresses into a 71,943 .mab file.
So this unit could be recognized as the natural child of DUnit and MadExcept wedding, in pure OpenSource. :)
Additional information is available on our forum. Feel free to ask. Feedback and feature requests are welcome! Works from Delphi 6 up to XE.