wqweto / ziparchive Goto Github PK

Hi
Please check and guide me on below code

With m_oZip
.OpenArchive "c:\temp\crdata.zip"
Set m_oExtractInMemory = m_oZip
bResult = .Extract("c:\temp", , "1111")
Set m_oExtractInMemory = Nothing
sLastError = .LastError
End With

Hi (wqweto),

Using this model :

.OpenArchive "C:\TEMP\aaa.zip"
.Extract baOutput, "mem/report.pdf"
WriteBinaryFile "C:\TEMP\report.pdf", baOutput

Is there a way to know the size of "mem/report.pdf" before writing it in "C:\TEMP" ?

Tried by search in the doc but "size" word appears too much times to makes searching revelant.

Thanks :)

hi nice to meet you I'm back after a long time.
A problem occurred while using "ZipArchive" again. :)

I have a file where the file date in the FileInfo method is about 1 second greater than the file date before compression.
Can you make the actual file and date the same?
And is it possible to keep the same date as the actual file when unpacking?

The ever-paranoid users at vbforums have raised concerns about PuTTY's licensing model as used in this project:

http://www.vbforums.com/showthread.php?848511-Is-there-a-way-to-work-around-the-Resource-Editor

I tried to keep things civil, but you know how it gets over there...

Anyway, to keep them happy, maybe it's worthwhile to include an explicit copy of PuTTY's license in the /lib/sshzlib/ folder, since PuTTY's developers are sloppy about including copyright and licensing details in each source file.

(Thanks again for sharing this project! I hope to test it on parsing ORA files in the near future...)

Hi ,

I've tested under VB6 so far and here your class works well - but the actual goal is, to include the class in an Excel marco: here the archives are created, but they are not encrypted.
I am using Excel 2010 32bit under Windows 10 - but the marco should work for 32 and 64 bit Excel version and under all Windows OS from XP on (lower encryption strength for XP and Vista is ok, because these operating systems are also not really secure anymore).
Is it possible to make encryption work under VBA too?

Thanks for you work,
Peter

EDIT: It is working - my bad! In the Excel macro I forgot to add the new ZIP_CRYPTO = 1 arguement...

Hi wqweto,

Trying to run your app :) Successfully ziped the 3 example files from C:\TEMP (replaces D:\TEMP as well as on the VM I have only C:\ ) to aaa.zip (removed level in the zip filename), with Command_3 sub.
So the aaa.zip has inside:

mem/enwik8.txt
mem/report.pdf
mem/report2.pdf
mem/thunk.bin

Now trying to extract a file (for example mem/report2.pdf) to C:\TEMP\Unzip but without creating the "mem" subfolder.
The goal is to get C:\TEMP\Unzip\report.pdf .

Tried Command_4 sub, where I changed bResult for :
.bResult = .Extract ("C:\TEMP\Unzip\", "mem/report2.pdf"

I get nothing.

Commented the lines :

Set m_oExtractInMemory = m_oZip
Set m_oExtractInMemory = Nothing

I get file extracted but it is here C:\TEMP\Unzip\mem\report.pdf

Is there a way to ride of the "mem" subfolder ?

Thanks :)
Couin

https://github.com/ebiggers/libdeflate/blob/master/lib/deflate_compress.c#L980-L1085

This seem to use thunking to enable some of the functionality. I am assuming this is only going to work on 32-bit and a different thunking code (if even possible?) would be needed to run on 64-bit? If so , how was that derived?

I read your answer from stackoverflow How to make VB6 console application which prints output to the command prompt .
It works.
ConsolePrint works on ASCII. But not work correctly on CJK.
eg.
ConsolePrint("确定" & vbCrLf)

How to output CJK correctly ?

I have tried extracting the test.zip file using the file path, and it succeeded in extracting all the files with all their contents,

With New cZipArchive
    .OpenArchive Text1.Text 'test.zip
    .Extract App.Path & "\test", Password:="test123"
End With

but when I put the test.zip file into the resource file, and tried to extract it, only one file was extracted, how do I solve this?

Dim bytResourceData() As Byte
    bytResourceData = LoadResData(404, "CUSTOM")
  
With New cZipArchive
    .OpenArchive bytResourceData
    .Extract App.Path & "\test", Password:="test123"
End With

Thanks @wqweto

Thank you for the good code.

I tested it with the attached sample program.

The compressed file is slightly larger than the Windows default compression.

Is there a reason?

Pros

• better compression ratio
• better decompression speed

Cons

• additional dependency
• official release is a cdecl only .dll

Thank you for the code.

I tested it with the attached sample program with a different .addfromfolder-command in command1 to my testdata (folder with subfolders, 3.5gb of about 3k documents).

When compressing this folder I get an error -1 "overflow" and the compressing stops.
Before that error 131 in pvVfsSeek is raised because of a negative lPosition value.

Do you have an idea to solve this issue to compress to archives larger than 2 gb?

Hi @wqweto. I'm looking at adding OpenRaster file format support to PhotoDemon. The spec is very straightforward, basically constructed around existing OpenDocument rules. cZipArchive makes my job much easier!

I haven't studied everything in great detail, but I notice a few odd requirements in the spec that (perhaps?) are not covered by ZipArchive at present. These are independent from things like Zip64 support, which are a totally different beast - and not what I'd dare ask for, since PhotoDemon couldn't handle 4gb+ files cleanly anyway.

• The required mimetype file must be uncompressed (e.g. STORED in zip parlance). I wonder if ZipArchive could expose a flag for adding individual entries as STORED vs DEFLATED.
• UTF-8 filenames are required by the spec. I'm not sure how much work it is to get/set the UTF-8 flag for zip filenames (or if this is tied to Zip64 format), but the spec mentions that the flag is important e.g.:

The file names within OpenRaster "zip files" are case-sensitive and must be UTF-8 encoded. If you decide to write non-ascii file names, be aware that the Zip format has an UTF-8 flag for each file name. Make sure this flag is set, otherwise the content might get decoded as cp437.

• As an unrelated convenience, it would be very helpful to retrieve a list of filenames as perhaps a string array, to validate ORA file format correctness before attempting actual extraction of individual files (which could be extremely large). I'm using my own (ugly) code to do this at present, and could send a PR if you don't mind some potentially hideous code... :)

Apologies if any of these items are already covered and I just didn't review the code carefully enough! TIA

I created a very small project that uses this class but the project icon has the small shield of the UAC and, on startup, UAC windows appears.
Is there any way to solve the problem?
Or should I try to solve the UAC problem in general?

In 2421dda, low-level Inflate() wrappers were made accessible to outside callers - a great idea for VB6 devs looking to drop a zlibwapi dependency.

Does it make sense to do something similar for low-level Deflate() wrappers? Your great contribution to VB6 compression tester already provides a possible implementation, as I recall.

Hi :)

I'm near from achieve backup and restore feature, looks running well with Jingle Palette's sample files.
And I tried with my jingles set (folder size is 788 MB but not all files are used), and on restoring , wow, a lot of jingles missing.
Started to investigate, files are listed, but archive is corrupted, I can not read or extract any file in Windows explorer. Zip file size in 262 144 kB (in Size column of Explorer)
Tried with removing files to archive, zip is no long corrupted since its under 262 144 kB.

Tried with some files with the test/basic project to rid of an eventually wrong code in my project.

Here is the co, based on existing Command3_click :

Private Sub Command3_Click()
    Dim dblTimer        As Double
    Dim bResult         As Boolean
    Dim sLastError      As String
    Dim oOutput         As cMemoryStream
    Dim lLevel          As Long

    dblTimer = Timer
    Set m_oZip = New cZipArchive
    Set oOutput = New cMemoryStream
    m_bCancel = False
    lLevel = 8
    With m_oZip
        .AddFile pvInitMemStream("C:\TEMP\Ep.2 2017-06-04 (Export FB mono).mp3"), "Ep.2 2017-06-04 (Export FB mono).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.5 2017-06-25 (via E90).mp3"), "Ep.5 2017-06-25 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.7 2017-07-09 (via E90).mp3"), "Ep.7 2017-07-09 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.8 2017-07-16 (via E90).mp3"), "Ep.8 2017-07-16 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.10 2017-07-30 (via E90).mp3"), "Ep.10 2017-07-30 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.11 2017-08-13 (via E90).mp3"), "Ep.11 2017-08-13 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.12 2017-08-20 (via E90).mp3"), "Ep.12 2017-08-20 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.13 2017-08-27 (via E90).mp3"), "Ep.13 2017-08-27 (via E90).mp3"
        .AddFile pvInitMemStream("C:\TEMP\Ep.14 2017-09-03 (via E90).mp3"), "Ep.14 2017-09-03 (via E90).mp3"
        bResult = .CompressArchive(oOutput, Level:=lLevel)
        sLastError = .LastError
    End With
    Set m_oZip = Nothing
    WriteBinaryFile "C:\TEMP\Podcasts.zip", oOutput.Contents
    labProgress.Caption = IIf(bResult, "Done. ", sLastError & ". ") & Format(Timer - dblTimer, "0.000") & " elapsed"
End Sub

The result is that only the Ep.2 2017-06-04 (Export FB mono).mp3 is in the zip file.
About progress, I see that the other files are displayed for a short time than first , and percentage not goes to 100% (or near) for each file, but only around 30% max.

Some idea of what I do wrong ? (Of course, files are present and not bad).

Thanks :)
Couin

This is an esoteric request, so feel free to disregard and I will simply muck up my own cZipArchive copy to implement. ;)

Because large array allocations are especially expensive in VB, I would prefer to simply reuse a single large destination buffer when extracting a series of sequential files. This is especially useful in my OpenRaster use-case, as I just need to extract PNG bytes via cZipArchive, then immediately decode it into a usable image buffer (and discard the original PNG stream). It would save me a ton of allocation time to take the approach of, "find largest extracted size, prep single buffer at that size, re-use for all PNG extractions" vs the current approach of allocating new memory for each temporary Extract() output buffer.

Obviously, cZipArchive supporting a naked long as extraction target would require some amount of trust that the caller has properly read .Size property and allocated accordingly. Not sure if this is desirable or not in a class designed for general use.

An alternate implementation could be a new optional parameter in Extract(), something to the effect of "only resize OutputTarget array if it's too small for this extraction", but this adds complexity on cZipArchive's end and the same problems persist - e.g. the caller still needs to confirm .Size to understand the output. That's why I thought this implementation (passing a naked pointer) may be preferable.

(Actually there are other use-cases too, e.g. the caller could pass a pointer to something like a memory-mapped file view, which otherwise would require extra steps to "wrap" that bare pointer in a VB array.)

I have create zip file from folder like code below, the zip file is successfully created, but with no password

Dim Zip As New cZipArchive
With Zip
    .AddFromFolder App.Path & "\test\*.*", Recursive:=True, TargetFolder:="test", IncludeEmptyFolders:=True, Password:="test123"
    .CompressArchive App.Path & "\test.zip"
End With

Please help @wqweto

Regards