cmss-compressing

The missing compressing and uncompressing lib for node.

Currently supported:

• tar
• gzip
• tgz
• zip

Install

1npm install cmss-compressing

Usage

Compress a single file

Use gzip as an example, tar, tgz and zip is same as gzip.

promise style

1const compressing = require('cmss-compressing');
2
3// compress a file
4compressing.gzip.compressFile('file/path/to/compress', 'path/to/destination.gz')
5.then(compressDone)
6.catch(handleError);
7
8// compress a file buffer
9compressing.gzip.compressFile(buffer, 'path/to/destination.gz')
10.then(compressDone)
11.catch(handleError);
12
13// compress a stream
14compressing.gzip.compressFile(stream, 'path/to/destination.gz')
15.then(compressDone)
16.catch(handleError);

stream style

1const compressing = require('cmss-compressing');
2
3new compressing.gzip.FileStream({ source: 'file/path/to/compress' })
4  .on('error', handleError)
5  .pipe(fs.createWriteStream('path/to/destination.gz'))
6  .on('error', handleError);
7
8// It's a transform stream, so you can pipe to it
9fs.createReadStream('file/path/to/compress')
10  .on('error', handleError)
11  .pipe(new compressing.gzip.FileStream())
12  .on('error', handleError)
13  .pipe(fs.createWriteStream('path/to/destination.gz'))
14  .on('error', handleError);
15
16// You should take care of stream errors in caution, use pump to handle error in one place
17const pump = require('pump';)
18const sourceStream = fs.createReadStream('file/path/to/compress')
19const gzipStream = new compressing.gzip.FileStream();
20const destStream = fs.createWriteStream('path/to/destination.gz');
21pump(sourceStream, gzipStream, destStream, handleError);

Compress a dir

Use tar as an example, tgz and zip is same as gzip.

Gzip only support compressing a single file. if you want to compress a dir with gzip, then you may need tgz instead.

promise style

1const compressing = require('cmss-compressing');
2compressing.tar.compressDir('dir/path/to/compress', 'path/to/destination.tar')
3.then(compressDone)
4.catch(handleError);

stream style

1const compressing = require('cmss-compressing');
2
3const tarStream = new compressing.tar.Stream();
4tarStream.addEntry('dir/path/to/compress');
5
6tarStream
7  .on('error', handleError)
8  .pipe(fs.createWriteStream('path/to/destination.tar'))
9  .on('error', handleError);
10
11// You should take care of stream errors in caution, use pump to handle error in one place
12const tarStream = new compressing.tar.Stream();
13tarStream.addEntry('dir/path/to/compress');
14const destStream = fs.createWriteStream('path/to/destination.tar');
15pump(tarStream, destStream, handleError);

Stream is very powerful, you can compress multiple entries in it;

1const tarStream = new compressing.tar.Stream();
2// dir
3tarStream.addEntry('dir/path/to/compress');
4
5// file
6tarStream.addEntry('file/path/to/compress');
7
8// buffer
9tarStream.addEntry(buffer);
10
11// stream
12tarStream.addEntry(stream);
13
14const destStream = fs.createWriteStream('path/to/destination.tar');
15pipe(tarStream, destStream, handleError);

Uncompress a file

promise style

1const compressing = require('cmss-compressing');
2
3// uncompress a file
4compressing.tgz.uncompress('file/path/to/uncompress.tgz', 'path/to/destination/dir')
5.then(uncompressDone)
6.catch(handleError);
7
8// uncompress a file buffer
9compressing.tgz.uncompress(buffer, 'path/to/destination/dir')
10.then(uncompressDone)
11.catch(handleError);
12
13// uncompress a stream
14compressing.tgz.uncompress(stream, 'path/to/destination/dir')
15.then(uncompressDone)
16.catch(handleError);

Note: tar, tgz and zip have the same uncompressing API as above: destination should be a path of a directory, while that of gzip is slightly different: destination must be a file or filestream.

And working with urllib is super easy. Let's download a tgz file and uncompress to a directory:

1const urllib = require('urllib');
2const targetDir = require('os').tmpdir();
3const compressing = require('cmss-compressing');
4
5urllib.request('http://registry.npmjs.org/pedding/-/pedding-1.1.0.tgz', {
6  streaming: true,
7  followRedirect: true,
8})
9.then(result => compressing.tgz.uncompress(result.res, targetDir))
10.then(() => console.log('uncompress done'))
11.catch(console.error);

stream style

1const compressing = require('cmss-compressing');
2const mkdirp = require('mkdirp');
3
4function onEntry(header, stream, next) => {
5  stream.on('end', next);
6
7  // header.type => file | directory
8  // header.name => path name
9
10  if (header.type === 'file') {
11    stream.pipe(fs.createWriteStream(path.join(destDir, header.name)));
12  } else { // directory
13    mkdirp(path.join(destDir, header.name), err => {
14      if (err) return handleError(err);
15      stream.resume();
16    });
17  }
18}
19
20new compressing.tgz.UncompressStream({ source: 'file/path/to/uncompress.tgz' })
21  .on('error', handleError)
22  .on('finish', handleFinish) // uncompressing is done
23  .on('entry', onEntry);
24
25// It's a writable stream, so you can pipe to it
26fs.createReadStream('file/path/to/uncompress')
27  .on('error', handleError)
28  .pipe(new compressing.tgz.UncompressStream())
29  .on('error', handleError)
30  .on('finish', handleFinish) // uncompressing is done
31  .on('entry', onEntry);

Note: tar, tgz and zip have the same uncompressing streaming API as above: it's a writable stream, and entries will be emitted while uncompressing one after one another, while that of gzip is slightly different: gzip.UncompressStream is a transform stream, so no entry event will be emitted and you can just pipe to another stream

This constrants is brought by Gzip algorithm itself, it only support compressing one file and uncompress one file.

1new compressing.gzip.UncompressStream({ source: 'file/path/to/uncompress.gz' })
2  .on('error', handleError)
3  .pipe(fs.createWriteStream('path/to/dest/file'))
4  .on('error', handleError);

API

compressFile

Use this API to compress a single file. This is a convenient method, which wraps FileStream API below, but you can handle error in one place.

• gzip.compressFile(source, dest, opts)
• tar.compressFile(source, dest, opts)
• tgz.compressFile(source, dest, opts)
• zip.compressFile(source, dest, opts)

Params

• source {String|Buffer|Stream} - source to be compressed, could be a file path, buffer, or a readable stream
• dest {String|Stream} - compressing destination, could be a file path(eg. /path/to/xx.tgz), or a writable stream.
• opts {Object} - usually you don't need it

Returns a promise object.

compressDir

Use this API to compress a dir. This is a convenient method, which wraps Stream API below, but you can handle error in one place.

Note: gzip do not have a compressDir method, you may need tgz instead.

• tar.compressDir(source, dest, opts)
• tgz.compressDir(source, dest, opts)
• zip.compressDir(source, dest, opts)

Params

• source {String|Buffer|Stream} - source to be compressed
• dest {String|Stream} - compressing destination, could be a file path(eg. /path/to/xx.tgz), or a writable stream.
• opts {Object} - usually you don't need it

uncompress

Use this API to uncompress a file. This is a convenient method, which wraps UncompressStream API below, but you can handle error in one place. RECOMMANDED.

• tar.uncompress(source, dest, opts)
• tgz.uncompress(source, dest, opts)
• zip.uncompress(source, dest, opts)
• gzip.uncompress(source, dest, opts)

Params

• source {String|Buffer|Stream} - source to be uncompressed
• dest {String|Stream} - uncompressing destination. When uncompressing tar, tgz and zip, it should be a directory path (eg. /path/to/xx). When uncompressing gzip, it should be a file path or a writable stream.
• opts {Object} - usually you don't need it

  • opts.zipFileNameEncoding {String} - Only work on zip format, default is 'utf8'. Major non-UTF8 encodings by languages:

    • Korean: cp949, euc-kr
    • Japanese: sjis (shift_jis), cp932, euc-jp
    • Chinese: gbk, gb18030, gb2312, cp936, hkscs, big5, cp950

FileStream

The transform stream to compress a single file.

Note: If you are not very familiar with streams, just use compressFile() API, error can be handled in one place.

• new gzip.FileStream(opts)
• new tar.FileStream(opts)
• new tgz.FileStream(opts)
• new zip.FileStream(opts)

Common params:

• opts.source {String|Buffer|Stream} - source to be compressed, could be a file path, buffer, or a readable stream.

Gzip params:

• opts.zlib - {Object} gzip.FileStream uses zlib to compress, pass this param to control the behavior of zlib.

Tar params:

• opts.relativePath {String} - Adds a file from source into the compressed result file as opts.relativePath. Uncompression programs would extract the file from the compressed file as relativePath. If opts.source is a file path, opts.relativePath is optional, otherwise it's required.
• opts.size {Number} - Tar compression requires the size of file in advance. When opts.source is a stream, the size of it cannot be calculated unless load all content of the stream into memory(the default behavior, but loading all data into memory could be a very bad idea). Pass opts.size to avoid loading all data into memory, or a warning will be shown.
• opts.suppressSizeWarning {Boolean} - Pass true to suppress the size warning mentioned.

Tgz params:

tgz.FileStream is a combination of tar.FileStream and gzip.FileStream, so the params are the combination of params of tar and gzip.

Zip params:

• opts.relativePath {String} - Adds a file from source into the compressed result file as opts.relativePath. Uncompression programs would extract the file from the compressed file as relativePath. If opts.source is a file path, opts.relativePath is optional, otherwise it's required.
• opts.yazl {Object} - zip.FileStream compression uses yazl, pass this param to control the behavior of yazl.

Stream

The readable stream to compress anything as you need.

Note: If you are not very familiar with streams, just use compressFile() and compressDir() API, error can be handled in one place.

Gzip only support compressing a single file. So gzip.Stream is not available.

Constructor

• new tar.Stream()
• new tgz.Stream()
• new zip.Stream()

No options in all constructors.

Instance methods

• addEntry(entry, opts)

Params

• entry {String|Buffer|Stream} - entry to compress, cound be a file path, a dir path, a buffer, or a stream.
• opts.relativePath {String} - uncompression programs would extract the file from the compressed file as opts.relativePath. If entry is a file path or a dir path, opts.relativePath is optional, otherwise it's required.
• opts.ignoreBase {Boolean} - when entry is a dir path, and opts.ignoreBase is set to true, the compression will contain files relative to the path passed, and not with the path included.

UncompressStream

The writable stream to uncompress anything as you need.

Note: If you are not very familiar with streams, just use uncompress() API, error can be handled in one place.

Gzip only support compressing and uncompressing one single file. So gzip.UncompressStream is a transform stream which is different from others.

Constructor

• new gzip.UncompressStream(opts)
• new tar.UncompressStream(opts)
• new tgz.UncompressStream(opts)
• new zip.UncompressStream(opts)

Common params:

• opts.source {String|Buffer|Stream} - source to be uncompressed, could be a file path, buffer, or a readable stream.

CAUTION for zip.UncompressStream

Due to the design of the .zip file format, it's impossible to interpret a .zip file without loading all data into memory.

Although the API is streaming style(try to keep it handy), it still loads all data into memory.

https://github.com/thejoshwolfe/yauzl#no-streaming-unzip-api