Detect installed windows programs

requirements.txt

@@ -10,6 +10,7 @@ chardet==4.0.0
 click==8.0.1
 colorama==0.4.4
 commoncode==21.6.11
+construct==2.10.67
 cryptography==3.4.7
 debian-inspector==21.5.25
 dparse==0.5.1
@@ -24,17 +25,21 @@ gemfileparser==0.8.0
 html5lib==1.1
 idna==2.10
 importlib-metadata==4.4.0
+inflection==0.5.1
 intbitset==2.4.1
 isodate==0.6.0
 jaraco.functools==3.3.0
 javaproperties==0.8.0
 Jinja2==3.0.1
+jsonlines==2.0.0
 jsonstreams==0.6.0
 license-expression==21.6.14
+Logbook==1.5.3
 lxml==4.6.3
 MarkupSafe==2.0.1
 more-itertools==8.8.0
 normality==2.2.2
+packagedcode-msitools==0.101.210706
 packageurl-python==0.9.4
 packaging==20.9
 patch==1.16
@@ -52,8 +57,10 @@ Pygments==2.9.0
 pygmars==0.7.0b2
 pymaven-patch==0.3.0
 pyparsing==2.4.7
+pytz==2021.1
 PyYAML==5.4.1
 rdflib==5.0.0
+regipy==1.9.3
 requests==2.25.1
 rpm-inspector-rpm==4.16.1.3.210404
 saneyaml==0.5.2
@@ -63,8 +70,10 @@ six==1.16.0
 sortedcontainers==2.4.0
 soupsieve==2.2.1
 spdx-tools==0.6.1
+tabulate==0.8.9
 text-unidecode==1.3
 toml==0.10.2
+tqdm==4.61.2
 typecode==21.6.1
 typecode-libmagic==5.39.210531
 typing==3.6.6

setup-mini.cfg

@@ -112,6 +112,8 @@ full =
 # linux-only package handling
 packages =
     rpm_inspector_rpm >= 4.16.1.3; platform_system == 'Linux'
+    regipy >= 1.9.3; platform_system == 'Linux'
+    packagedcode_msitools >= 0.101.210706; platform_system == 'Linux'
 
 dev =
     pytest

setup.cfg

@@ -112,6 +112,8 @@ full =
 # linux-only package handling
 packages =
     rpm_inspector_rpm >= 4.16.1.3; platform_system == 'Linux'
+    regipy >= 1.9.3; platform_system == 'Linux'
+    packagedcode_msitools >= 0.101.210706; platform_system == 'Linux'
 
 dev =
     pytest

src/packagedcode/__init__.py

@@ -22,6 +22,7 @@
 from packagedcode import haxe
 from packagedcode import maven
 from packagedcode import models
+from packagedcode import msi
 from packagedcode import npm
 from packagedcode import nuget
 from packagedcode import opam
@@ -32,6 +33,8 @@
 from packagedcode import rpm
 from packagedcode import rubygems
 from packagedcode import win_pe
+from packagedcode import windows
+
 
 # Note: the order matters: from the most to the least specific
 # Package classes MUST be added to this list to be active
@@ -67,7 +70,6 @@
     pypi.PythonPackage,
     golang.GolangPackage,
     models.CabPackage,
-    models.MsiInstallerPackage,
     models.InstallShieldPackage,
     models.NSISInstallerPackage,
     nuget.NugetPackage,
@@ -83,6 +85,8 @@
     win_pe.WindowsExecutable,
     readme.ReadmePackage,
     build.MetadataBzl,
+    msi.MsiInstallerPackage,
+    windows.MicrosoftUpdateManifestPackage,
     pubspec.PubspecPackage,
 ]
 

src/packagedcode/models.py

@@ -810,14 +810,6 @@ class CabPackage(Package):
     default_type = 'cab'
 
 
-@attr.s()
-class MsiInstallerPackage(Package):
-    filetypes = ('msi installer',)
-    mimetypes = ('application/x-msi',)
-    extensions = ('.msi',)
-    default_type = 'msi'
-
-
 @attr.s()
 class InstallShieldPackage(Package):
     filetypes = ('installshield',)

src/packagedcode/msi.py

@@ -0,0 +1,180 @@
+#
+# Copyright (c) nexB Inc. and others. All rights reserved.
+# ScanCode is a trademark of nexB Inc.
+# SPDX-License-Identifier: Apache-2.0
+# See http://www.apache.org/licenses/LICENSE-2.0 for the license text.
+# See https://github.com/nexB/scancode-toolkit for support or download.
+# See https://aboutcode.org for more information about nexB OSS projects.
+#
+
+import os
+import re
+import warnings
+from shutil import which
+
+import attr
+
+from commoncode.command import execute
+from commoncode.command import find_in_path
+from commoncode.system import on_linux
+from commoncode.version import VERSION_PATTERNS_REGEX
+from packagedcode import models
+
+
+MSIINFO_BIN_LOCATION = 'packagedcode_msitools.msiinfo'
+
+
+def get_msiinfo_bin_location(_cache=[]):
+    """
+    Return the binary location for msiinfo from either:
+    - a plugin-provided path,
+    - the system PATH.
+    Raise an Exception if no msiinfo command can be found.
+    """
+    if _cache:
+        return _cache[0]
+
+    from plugincode.location_provider import get_location
+
+    # try a plugin-provided path first
+    cmd_loc = get_location(MSIINFO_BIN_LOCATION)
+
+    # try the PATH
+    if not cmd_loc:
+        cmd = 'msiinfo'
+        cmd_loc = find_in_path(cmd)
+
+        if not cmd_loc:
+            cmd_loc = which(cmd)
+
+        if cmd_loc:
+            warnings.warn(
+                'Using "msiinfo" command found in the PATH. '
+                'Install instead a plugincode-msitools plugin for best support.'
+            )
+
+    if not cmd_loc or not os.path.isfile(cmd_loc):
+        raise Exception(
+            'CRITICAL: msiinfo not provided. '
+            'Unable to continue: you need to install the plugin packagedcode-msitools'
+        )
+    _cache.append(cmd_loc)
+    return cmd_loc
+
+
+class MsiinfoException(Exception):
+    pass
+
+
+def parse_msiinfo_suminfo_output(output_string):
+    """
+    Return a dictionary containing information from the output of `msiinfo suminfo`
+    """
+    # Split lines by newline and place lines into a list
+    output_list = output_string.splitlines()
+    results = {}
+    # Partition lines by the leftmost ":", use the string to the left of ":" as
+    # the key and use the string to the right of ":" as the value
+    for output in output_list:
+        key, _, value = output.partition(':')
+        if key:
+            results[key] = value.strip()
+    return results
+
+
+def get_msi_info(location):
+    """
+    Run the command `msiinfo suminfo` on the file at `location` and return the
+    results in a dictionary
+
+    This function requires `msiinfo` to be installed on the system, either by
+    installing the `packagedcode-msiinfo` plugin or by installing `msitools`
+    through a package manager.
+    """
+    rc, stdout, stderr = execute(
+        cmd_loc=get_msiinfo_bin_location(),
+        args=[
+            'suminfo',
+            location,
+        ],
+    )
+    if stderr:
+        error_message = f'Error encountered when reading MSI information from {location}: '
+        error_message = error_message + stderr
+        raise MsiinfoException(error_message)
+    return parse_msiinfo_suminfo_output(stdout)
+
+
+def get_version_from_subject_line(subject_line):
+    """
+    Return a version number from `subject_line`
+
+    `subject_line` is the `Subject` field from the output of
+    `msiinfo suminfo <msi installer file>`. This string sometimes contains
+    the version number of the package contained in the MSI installer.
+    """
+    for pattern in VERSION_PATTERNS_REGEX():
+        version = re.search(pattern, subject_line)
+        if version:
+            v = version.group(0)
+            # prefix with v space
+            if not v.lower().startswith('v'):
+                v = f'v {v}'
+            return v
+
+
+def create_package_from_msiinfo_results(msiinfo_results):
+    """
+    Return an MsiInstallerPackage from the dictionary `msiinfo_results`
+    """
+    author_name = msiinfo_results.get('Author', '')
+    parties = []
+    if author_name:
+        parties.append(
+            models.Party(
+                type=None,
+                role='author',
+                name=author_name
+            )
+        )
+
+    # Currently, we use the contents `Subject` field from the msiinfo suminfo
+    # results as the package name because it contains the package name most of
+    # the time. Getting the version out of the `Subject` string is not
+    # straightforward because the format of the string is usually different
+    # between different MSIs
+    subject = msiinfo_results.get('Subject', '')
+    name = subject
+    version = get_version_from_subject_line(subject)
+    description = msiinfo_results.get('Comments', '')
+    keywords = msiinfo_results.get('Keywords', '')
+
+    return MsiInstallerPackage(
+        name=name,
+        version=version,
+        description=description,
+        parties=parties,
+        keywords=keywords,
+        extra_data=msiinfo_results
+    )
+
+
+def msi_parse(location):
+    """
+    Return an MsiInstallerPackage from `location`
+    """
+    info = get_msi_info(location)
+    return create_package_from_msiinfo_results(info)
+
+
+@attr.s()
+class MsiInstallerPackage(models.Package):
+    filetypes = ('msi installer',)
+    mimetypes = ('application/x-msi',)
+    extensions = ('.msi',)
+    default_type = 'msi'
+
+    @classmethod
+    def recognize(cls, location):
+        if on_linux:
+            yield msi_parse(location)

src/packagedcode/win_pe.py

@@ -276,10 +276,10 @@ def parse(location):
     infos = pe_info(location)
 
     version = get_first(
-        infos, 
-        'Full Version', 
-        'ProductVersion', 
-        'FileVersion', 
+        infos,
+        'Full Version',
+        'ProductVersion',
+        'FileVersion',
         'Assembly Version',
     )
     release_date = get_first(infos, 'BuildDate')
@@ -289,9 +289,9 @@ def parse(location):
         release_date = release_date.replace('/', '-')
 
     name = get_first(
-        infos, 
-        'ProductName', 
-        'OriginalFilename', 
+        infos,
+        'ProductName',
+        'OriginalFilename',
         'InternalName',
     )
     copyr = get_first(infos, 'LegalCopyright')