Fixed CI & added Python 3.7

* Fixed Travis CI build

* Updated zstd test binary blob

* Added: Python 3.7

* Fixed: change type of "information" for RAR

* Added: test the guesser alone to get better error message

* Increase major and update dependency

Author: cecton

.travis.yml

@@ -4,12 +4,19 @@ python:
   - "3.4"
   - "3.5"
   - "3.6"
+  - "3.7"
+
+os:
+  - linux
+  # - osx         # NOTE: Python is not available yet (not tested)
+  # - windows     # NOTE: Python is not available yet
 
 cache: pip
 
 before_install:
   - sudo apt-get install p7zip rar
-  - sudo sh -c 'echo "deb http://archive.ubuntu.com/ubuntu/ artful main restricted universe" > /etc/apt/sources.list'
+  # NOTE: attempt to get a more updated version of file and libmagic >= 5.30
+  - sudo sh -c 'echo "deb http://archive.ubuntu.com/ubuntu/ bionic main restricted universe" > /etc/apt/sources.list'
   - sudo apt-get update
   - sudo apt-get install zstd
   - sudo apt-get install file libmagic1

README.md

@@ -6,6 +6,15 @@ destream
 
 _destream: decompress a stream_
 
+Compatibility
+-------------
+
+ *  Python: 2.7, 3.4, 3.5, 3.6, 3.7
+ *  OS:
+     -  Linux
+     -  OSX is not tested
+     -  Windows is not tested and requires file and libmagic for Windows
+
 Installation
 ------------
 
@@ -52,15 +61,15 @@ Lib Usage
     assert isinstance(archive, destream.Archive)
 
     # ==> we can read the content but not seek
-    print archive.read()
+    print(archive.read())
     ```
 3.  Open an archive that holds only one file,
     ```python
     archive = destream.open("some_file.tar.xz")
 
     # ==> we can read the archive like it is a stream
     if archive.single():
-        print archive.read()
+        print(archive.read())
     else:
         archive.extractall('/tmp/some/path/')
     ```

destream/decompressors/rar.py

@@ -12,28 +12,28 @@
 __all__ = ['Unrar']
 
 
-def parse_hunk(hunk):
-    info = {}
-    for m in re.finditer(
-            r'^[ \t\f]*(.+?)[ \t\f]*:[ \t\f]*(.*?)[ \t\f]*$',
-            hunk, flags=re.M):
-        key = re.sub(r'\W', '_', m.group(1).lower())
-        info[key] = m.group(2)
-    if info.get('service', '') == 'EOF':
-        raise StopIteration()
-    return info
+def iter_on_hunks(hunks):
+    for hunk in hunks:
+        info = {}
+        for m in re.finditer(
+                r'^[ \t\f]*(.+?)[ \t\f]*:[ \t\f]*(.*?)[ \t\f]*$',
+                hunk, flags=re.M):
+            key = re.sub(r'\W', '_', m.group(1).lower())
+            info[key] = m.group(2)
+        if info.get('service', '') == 'EOF':
+            break
+        yield info
 
 
 class Header(object):
-    def __init__(self, hunk):
-        self.__dict__.update(parse_hunk(hunk))
+    def __init__(self, info):
+        self.__dict__.update(info)
         assert 'RAR' in self.details, \
             "Maybe not a RAR file:%s\n" % self.details
 
 
 class Member(object):
-    def __init__(self, hunk):
-        info = parse_hunk(hunk)
+    def __init__(self, info):
         info['filename'] = info.pop('name')
         info['size'] = int(info.get('size', 0))
         info['packed_size'] = int(info.get('packed_size', 0))
@@ -76,8 +76,8 @@ def __init__(self, name, fileobj):
         self.fileobj = ArchiveTemp(fileobj)
         output = check_output(self._command +
                               ['vta', self.fileobj.name]).decode()
-        hunks = iter(output.split("\n\n"))
-        self.information = next(hunks).strip()
+        hunks = iter_on_hunks(output.split("\n\n"))
+        self.information = next(hunks)
         self.header = Header(next(hunks))
         self._members = [m for m in (Member(h) for h in hunks)]
         self._stream = (len(self._members) == 1)

setup.py

@@ -2,7 +2,7 @@
 
 setup(
     name = "destream",
-    version = "4.2.0",
+    version = "5.0.0",
     author = "Cecile Tonglet",
     author_email = "cecile.tonglet@gmail.com",
     description = ("A simple module to decompress streams compressed multiple "
@@ -12,7 +12,7 @@
     url = "https://github.com/cecton/destream",
     packages = find_packages(),
     scripts = ['scripts/destream'],
-    install_requires = ['python-magic==0.4.12'],
+    install_requires = ['python-magic>=0.4.12'],
     classifiers = [
         "Development Status :: 5 - Production/Stable",
         "Topic :: System :: Archiving :: Compression",

tests/test_30_decompressors.py

@@ -6,6 +6,7 @@
 from io import BytesIO
 import tarfile
 import zipfile
+import magic
 try:
     import unittest2 as unittest
 except ImportError:
@@ -25,6 +26,10 @@ def _check_decompressor(self, decompressor, compressed_fileobj,
             self.skipTest("decompressor not available")
         if expected_name is None:
             expected_name = decompressed_fileobj.name
+        archive = destream.ArchiveFile(compressed_fileobj, None, closefd=False)
+        mime = magic.from_buffer(archive.peek(1024), mime=True)
+        decompressor._guess(mime, str(archive.realname), compressed_fileobj)
+        compressed_fileobj.seek(0)
         with destream.open(
                 fileobj=compressed_fileobj, closefd=False) as archive:
             # check that the decompressor has been used
@@ -226,7 +231,7 @@ def test_20_external_pipe_zstd(self):
         uncompressed = BytesIO(b"Hello World\n")
         uncompressed.name = 'test_file'
         raw = BytesIO(
-            b'%\xb5/\xfd\x08@\x00\x0cHello World\n\xc0\x00\x00')
+            b'(\xb5/\xfd$\x0ca\x00\x00Hello World\n\x93C\x0f\x1a')
         raw.name = "test_file.zst"
         self._check_decompressor(
             destream.decompressors.Unzstd,