I'm interested in compressing many versions of a similar file. The files are PDFs with (often minor) differences between them.
My question is: Is the zip or gzip algorithm able to use the similarity between these files to improve compression? Or does it handle each file individually?
I've looked at http://www.infinitepartitions.com/art001.html from How does the GZip algorithm work?, which goes over the algorithms themselves, but doesn't answer whether implementation handles all files individually or not.
Follow-up questions: If not, are there file compression algorithms that would be able to leverage the similarity between files to aid in compression?
zip no. Files are compressed independently from each other. gzip by itself will only compress one file. What you want is tar and gzip, where tar will put the files adjacent to each other (with intervening headers), and gzip will then compress the whole thing a one stream, producing a .tar.gz file.
That will be able to take advantage of similarities that are ~32K distant from each other. If your files are much larger than 32K, then you should try xz instead of gzip, producing a .tar.xz file.
Related
I have a bunch of large HDF5 files (all around 1.7G), which share a lot of their content – I guess that more than 95% of the data of each file is found repeated in every other.
I would like to compress them in an archive.
My first attempt using GNU tar with the -z option (gzip) failed: the process was terminated when the archive reached 50G (probably a file size limitation imposed by the sysadmin). Apparently, gzip wasn't able to take advantage of the fact that the files are near-identical in this setting.
Compressing these particular files obviously doesn't require a very fancy compression algorithm, but a veeery patient one.
Is there a way to make gzip (or another tool) detect these large repeated blobs and avoid repeating them in the archive?
Sounds like what you need is a binary diff program. You can google for that, and then try using binary diff between two of them, and then compressing one of them and the resulting diff. You could get fancy and try diffing all combinations, picking the smallest ones to compress, and send only one original.
NOTE: I am using the term tar loosely here. I mean compress whether it be tar.gz, tar.bz2, zip, etc.
Is there a flag for rsync to negotiate the changed files between source/destination, tar the changed source files, send the single tar file to the destination machine and untar the changed files once arrived?
I have millions of files and remotely rsyncing across the internet to AWS seems very slow.
I know that rsync has a compression option (z), but it's my understanding that that compresses changed files on a per file basis. If there are many small files, the overhead of sending a 1KB as opposed to a 50KB file is still the bottleneck.
Also, simply tarring the whole directory is not efficient either as it will take an hour to archive
You can use the rsyncable option of gzip or pigz to compress the tar file to .gz format. (You will likely have to find a patch for gzip to add that. It's already part of pigz.)
The option partitions the resulting gzip file in a way that permits rsync to find only the modified portions for much more efficient transfers when only some of the files in the .tar.gz file have been changed.
I was looking for exact same thing as you and I landed on using borg.
tar cf - -C $DIR . | borg create $REPO::$NAME
tar will still read entire folder so you won't avoid a read penalty versus just rsyncing two dirs (since I believe rsync uses tricks to avoid reading each file for changes), but you will avoid the write penalty because borg will only write blocks it hasn't encountered before. Also borg auto compresses so no need for xz/gzip. Also, if borg is installed on both ends it won't send over superfluous data either since the two borgs can let each other know what they have versus don't.
If avoiding that read penalty is crucial for you, you could possibly use rsync to use its tricks to just tell you which files changed, create a difftar and send that to borg, but then getting borg to merge archives is whole second headache. You'd likely end up creating a filter that removes paths that were deleted from the original archive and then creating a new archive of just file additions/changes. And then you'd have to do that for each archive recursively. In the end it will create the original archive by extracting each version in sequence, but like I said a total headache.
In the Unix world, there is a famous format called "tar.gz".
But now, I want to develop a game and random accessing a file will be more efficient. If it is archived first, it will cause sequential access.
I know that there is an alternative format called zip or 7z, but what about other formats?
Not only gz.tar, I'd like to a minor compressing library and also get archiving features.
Should I use *.tar or other solutions are available?
PS: I'm using C++.
"Random" access is not good on a .tar.gz, since that is a .tar file that has been wrapped in a .gz compression, so to get to things in the .tar file, you'd first have to decompress the .tar file.
It would be possible to use a .tar file that contains individual files compressed with .gz. You can read the table of content of the .tar file and find/store where all the files are in the archive, and then extract as you need. However, you may find that using your own format is "better" (for example, if I remember correctly, the "header" for a tar-archive is a file at a time, you may want to build your header in one lump, before you store the files [which does mean at least enumerating all the relevant files first, then forming the compressed variant and "patching up" the header with the offsets in compressed form]
For a game, one critical factor would probably be the decompression speed, so you may want to look at different libraries and which one has the best decompression speed. I found this when searching for a comparison:
http://catchchallenger.first-world.info//wiki/Quick_Benchmark:_Gzip_vs_Bzip2_vs_LZMA_vs_XZ_vs_LZ4_vs_LZO
You may also care about memory usage, which also varies a bit depending on algorithm.
And I'm guessing your individual files will be much smaller than the entire tar-ball of Linux, so you may want to do your own benchmark, with your own data - after all, the speed of different compression formats does, to some degree, depend on the format of the data.
Normally, for computer games, what you need is a format where each file is compressed individually before being assembled into one file. This is the crucial difference between .tar.gz and .zip / .7z formats, that is, tar-gz is a "compressed archive" while zip / 7z are "archives of compressed files". In fact, both file formats use the same compression algorithm (by default), and the only reason that .tar.gz files are typically smaller is because they compress the entire archive instead of file-by-file, which increases the overall compression ratio.
AFAIK, most computer games use a zip format or a custom format that closely matches it, because it does per-file compression. For instance, Quake engines have always (.pak, .pk3, .pk4) relied on an off-the-shelf zip format with a few minor additions (like a built-in checksum, I think).
The .tar.gz format is created by first making an archive that puts all the (uncompressed) files into one .tar file. Then, that big archive file is compressed with the gzip method to create the final .tar.gz file. The point is that to get any one of the files from the archive you have the decompress the entire thing. This is very appropriate for backups or large transfers, but not appropriate at all for a game engine media archive.
That said, you could technically do the reverse of tar-gz, which is to compress each file individually with gzip, and then put them together in a .tar archive. But this is probably not worth the extra trouble, as it is pretty much exactly what zip files are (in "one easy step"). So, it will be a lot easier to use an off-the-shelf all-in-one format like zip that will allow you to extract individual files at a time. There are many off-the-shelf libraries for extracting and manipulating files in zip archives, just start with libzip (not to be confused with zlib (for gzip or .gz)).
In the Unix world, there is a famous format called "tar.gz".
Probably the biggest reason why "tar-ballz" are so popular and famously used in Unix-like systems is that they preserve file permissions (and other meta-data, I guess). I think that some implementations of zip and 7z might provide that feature as an extension to the format, but most don't have it. The convenient thing with tar archives is that whatever you put in there comes out exactly the same at the other end, with all permissions and whatever else preserved. And the "gzip" compression (from zlib) has just been historically an industry-standard compression algorithm, although, now, there are better ones, also supported by tar, such as .tar.lzma (or .tlz) or .tar.xz.
but what about other formats?
There aren't really that many other formats. Mostly, compressed archive formats often reuse the same few algorithms (DEFLATE, LZ77 / LZMA / LZMA2, BZIP, etc.), and often, formats like zip / 7z / rar are only really container formats that can employ any of those compression algorithms (and even mix and match depending on the individual file types). The point is that you won't really find much that is better than zip or 7z. And their competitors are more or less gone today (like rar?).
Should I use *.tar or other solutions are available?
No, use zip or 7z. Tar-balls are for backups. They are optimized for that purpose (e.g., dump a large folder full of files into a tar-ball, and recover it later, with everything preserved and with best full-archive compression). For your application, zip or 7z is more appropriate.
The compression algorithm used in zlib is essentially the same as that in gzip and zip. What are gzip and zip? How are they different and how are they same?
Short form:
.zip is an archive format using, usually, the Deflate compression method. The .gz gzip format is for single files, also using the Deflate compression method. Often gzip is used in combination with tar to make a compressed archive format, .tar.gz. The zlib library provides Deflate compression and decompression code for use by zip, gzip, png (which uses the zlib wrapper on deflate data), and many other applications.
Long form:
The ZIP format was developed by Phil Katz as an open format with an open specification, where his implementation, PKZIP, was shareware. It is an archive format that stores files and their directory structure, where each file is individually compressed. The file type is .zip. The files, as well as the directory structure, can optionally be encrypted.
The ZIP format supports several compression methods:
0 - The file is stored (no compression)
1 - The file is Shrunk
2 - The file is Reduced with compression factor 1
3 - The file is Reduced with compression factor 2
4 - The file is Reduced with compression factor 3
5 - The file is Reduced with compression factor 4
6 - The file is Imploded
7 - Reserved for Tokenizing compression algorithm
8 - The file is Deflated
9 - Enhanced Deflating using Deflate64(tm)
10 - PKWARE Data Compression Library Imploding (old IBM TERSE)
11 - Reserved by PKWARE
12 - File is compressed using BZIP2 algorithm
13 - Reserved by PKWARE
14 - LZMA
15 - Reserved by PKWARE
16 - IBM z/OS CMPSC Compression
17 - Reserved by PKWARE
18 - File is compressed using IBM TERSE (new)
19 - IBM LZ77 z Architecture
20 - deprecated (use method 93 for zstd)
93 - Zstandard (zstd) Compression
94 - MP3 Compression
95 - XZ Compression
96 - JPEG variant
97 - WavPack compressed data
98 - PPMd version I, Rev 1
99 - AE-x encryption marker (see APPENDIX E)
Methods 1 to 7 are historical and are not in use. Methods 9 through 98 are relatively recent additions and are in varying, small amounts of use. The only method in truly widespread use in the ZIP format is method 8, Deflate, and to some smaller extent method 0, which is no compression at all. Virtually every .zip file that you will come across in the wild will use exclusively methods 8 and 0, likely just method 8. (Method 8 also has a means to effectively store the data with no compression and relatively little expansion, and Method 0 cannot be streamed whereas Method 8 can be.)
The ISO/IEC 21320-1:2015 standard for file containers is a restricted zip format, such as used in Java archive files (.jar), Office Open XML files (Microsoft Office .docx, .xlsx, .pptx), Office Document Format files (.odt, .ods, .odp), and EPUB files (.epub). That standard limits the compression methods to 0 and 8, as well as other constraints such as no encryption or signatures.
Around 1990, the Info-ZIP group wrote portable, free, open-source implementations of zip and unzip utilities, supporting compression with the Deflate format, and decompression of that and the earlier formats. This greatly expanded the use of the .zip format.
In the early '90s, the gzip format was developed as a replacement for the Unix compress utility, derived from the Deflate code in the Info-ZIP utilities. Unix compress was designed to compress a single file or stream, appending a .Z to the file name. compress uses the LZW compression algorithm, which at the time was under patent and its free use was in dispute by the patent holders. Though some specific implementations of Deflate were patented by Phil Katz, the format was not, and so it was possible to write a Deflate implementation that did not infringe on any patents. That implementation has not been so challenged in the last 20+ years. The Unix gzip utility was intended as a drop-in replacement for compress, and in fact is able to decompress compress-compressed data (assuming that you were able to parse that sentence). gzip appends a .gz to the file name. gzip uses the Deflate compressed data format, which compresses quite a bit better than Unix compress, has very fast decompression, and adds a CRC-32 as an integrity check for the data. The header format also permits the storage of more information than the compress format allowed, such as the original file name and the file modification time.
Though compress only compresses a single file, it was common to use the tar utility to create an archive of files, their attributes, and their directory structure into a single .tar file, and then compress it with compress to make a .tar.Z file. In fact, the tar utility had and still has the option to do the compression at the same time, instead of having to pipe the output of tar to compress. This all carried forward to the gzip format, and tar has an option to compress directly to the .tar.gz format. The tar.gz format compresses better than the .zip approach, since the compression of a .tar can take advantage of redundancy across files, especially many small files. .tar.gz is the most common archive format in use on Unix due to its very high portability, but there are more effective compression methods in use as well, so you will often see .tar.bz2 and .tar.xz archives.
Unlike .tar, .zip has a central directory at the end, which provides a list of the contents. That and the separate compression provides random access to the individual entries in a .zip file. A .tar file would have to be decompressed and scanned from start to end in order to build a directory, which is how a .tar file is listed.
Shortly after the introduction of gzip, around the mid-1990s, the same patent dispute called into question the free use of the .gif image format, very widely used on bulletin boards and the World Wide Web (a new thing at the time). So a small group created the PNG losslessly compressed image format, with file type .png, to replace .gif. That format also uses the Deflate format for compression, which is applied after filters on the image data expose more of the redundancy. In order to promote widespread usage of the PNG format, two free code libraries were created. libpng and zlib. libpng handled all of the features of the PNG format, and zlib provided the compression and decompression code for use by libpng, as well as for other applications. zlib was adapted from the gzip code.
All of the mentioned patents have since expired.
The zlib library supports Deflate compression and decompression, and three kinds of wrapping around the deflate streams. Those are no wrapping at all ("raw" deflate), zlib wrapping, which is used in the PNG format data blocks, and gzip wrapping, to provide gzip routines for the programmer. The main difference between zlib and gzip wrapping is that the zlib wrapping is more compact, six bytes vs. a minimum of 18 bytes for gzip, and the integrity check, Adler-32, runs faster than the CRC-32 that gzip uses. Raw deflate is used by programs that read and write the .zip format, which is another format that wraps around deflate compressed data.
zlib is now in wide use for data transmission and storage. For example, most HTTP transactions by servers and browsers compress and decompress the data using zlib, specifically HTTP header Content-Encoding: deflate means deflate compression method wrapped inside the zlib data format.
Different implementations of deflate can result in different compressed output for the same input data, as evidenced by the existence of selectable compression levels that allow trading off compression effectiveness for CPU time. zlib and PKZIP are not the only implementations of deflate compression and decompression. Both the 7-Zip archiving utility and Google's zopfli library have the ability to use much more CPU time than zlib in order to squeeze out the last few bits possible when using the deflate format, reducing compressed sizes by a few percent as compared to zlib's highest compression level. The pigz utility, a parallel implementation of gzip, includes the option to use zlib (compression levels 1-9) or zopfli (compression level 11), and somewhat mitigates the time impact of using zopfli by splitting the compression of large files over multiple processors and cores.
ZIP is a file format used for storing an arbitrary number of files and folders together with lossless compression. It makes no strict assumptions about the compression methods used, but is most frequently used with DEFLATE.
Gzip is both a compression algorithm based on DEFLATE but less encumbered with potential patents et al, and a file format for storing a single compressed file. It supports compressing an arbitrary number of files and folders when combined with tar. The resulting file has an extension of .tgz or .tar.gz and is commonly called a tarball.
zlib is a library of functions encapsulating DEFLATE in its most common LZ77 incarnation.
The most important difference is that gzip is only capable to compress a single file while zip compresses multiple files one by one and archives them into one single file afterwards.
Thus, gzip comes along with tar most of the time (there are other possibilities, though). This comes along with some (dis)advantages.
If you have a big archive and you only need one single file out of it, you have to decompress the whole gzip file to get to that file. This is not required if you have a zip file.
On the other hand, if you compress 10 similiar or even identical files, the zip archive will be much bigger because each file is compressed individually, whereas in gzip in combination with tar a single file is compressed which is much more effective if the files are similiar (equal).
This question on archiving PDF's got me wondering -- if I wanted to compress (for archival purposes) lots of files which are essentially small changes made on top of a master template (a letterhead), it seems like huge compression gains can be had with inter-file compression.
Do any of the standard compression/archiving formats support this? AFAIK, all the popular formats focus on compressing each single file.
Several formats do inter-file compression.
The oldest example is .tar.gz; a .tar has no compression but concatenates all the files together, with headers before each file, and a .gz can compress only one file. Both are applied in sequence, and it's a traditional format in the Unix world. .tar.bz2 is the same, only with bzip2 instead of gzip.
More recent examples are formats with optional "solid" compression (for instance, RAR and 7-Zip), which can internally concatenate all the files before compressing, if enabled by a command-line flag or GUI option.
Take a look at google's open-vcdiff.
http://code.google.com/p/open-vcdiff/
It is designed for calculating small compressed deltas and implements RFC 3284.
http://www.ietf.org/rfc/rfc3284.txt
Microsoft has an API for doing something similar, sans any semblance of a standard.
In general the algorithms you are looking for are ones based on Bentley/McIlroy:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.11.8470
In particular these algorithms will be a win if the size of the template is larger than the window size (~32k) used by gzip or the block size (100-900k) used by bzip2.
They are used by Google internally inside of their BIGTABLE implementation to store compressed web pages for much the same reason you are seeking them.
Since LZW compression (which pretty much they all use) involves building a table of repeated characters as you go along, such as schema as you desire would limit you to having to decompress the entire archive at once.
If this is acceptable in your situation, it may be simpler to implement a method which just joins your files into one big file before compression.