If there was a global default on a machine that was something other than
'master', these things would crash with:
Traceback (most recent call last):
File "/home/hans/code/fdroid/server/fdroid", line 22, in <module>
fdroidserver.__main__.main()
File "/home/hans/code/fdroid/server/fdroidserver/__main__.py", line 230, in main
raise e
File "/home/hans/code/fdroid/server/fdroidserver/__main__.py", line 211, in main
mod.main()
File "/home/hans/code/fdroid/server/fdroidserver/deploy.py", line 833, in main
push_binary_transparency(BINARY_TRANSPARENCY_DIR,
File "/home/hans/code/fdroid/server/fdroidserver/deploy.py", line 705, in push_binary_transparency
local.pull('master')
File "/usr/lib/python3/dist-packages/git/remote.py", line 1045, in pull
res = self._get_fetch_info_from_stderr(proc, progress, kill_after_timeout=kill_after_timeout)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3/dist-packages/git/remote.py", line 848, in _get_fetch_info_from_stderr
proc.wait(stderr=stderr_text)
File "/usr/lib/python3/dist-packages/git/cmd.py", line 604, in wait
raise GitCommandError(remove_password_if_present(self.args), status, errstr)
git.exc.GitCommandError: Cmd('git') failed due to: exit code(1)
cmdline: git pull -v -- local master
stderr: 'fatal: couldn't find remote ref master'
The builder should check the `AllowedAPKSigningKeys` at build time, so
that the CI can check if somebody gives a wrong value that doesn't match
a compared RB binary. In the event it fails, it gives useful
information, and in the event it succeeds, it makes it clear that this
build has verification back to the developer's original key.
Also, add tests for this to the test suite.
I tried to get this to indent the .yaml files properly so yamllint defaults
work with tests/metadata/dump/*.yaml, but it didn't take for some reason:
yaml.indent(mapping=4, sequence=4, offset=2)
Before this, there were separate post-parse paths for app-fields versus
build-flags. This makes all TYPE_STRING values always go through the same
post-parse code path.
My guess is that this is some kind of vestige of the old code structure,
back when there was .txt and .yml formats. This makes it a normal Python
function: input as arg, return value is the result.
This lets mirrors: in config.yml be the same list-of-dicts format as it is
in index-v2. This also includes a data format conversion to maintain the
right format for the old, unchanging index v0 and v1 formats.
#928#1107
It turns out that the maven: field was originally declared as a TYPE_STRING,
given that it was not given a different type in metadata.py's flagtypes.
The code was confused because it was given a default value of `False` rather
than `None` as the rest of the TYPE_STRING fields have.
This construct in build.py means maven: should always be a string:
if '@' in build.maven:
maven_dir = os.path.join(root_dir, build.maven.split('@', 1)[1])
else:
maven_dir = root_dir
The paths in the config must be strings because they are used in things
like env vars where they must be strings. Plus lots of other places in the
code assumes they are strings. This is the first step to defining the
border of where paths can be pathlib.Path() and where they must be strings.
This job just started failing, but wasn't before:
https://gitlab.com/eighthave/fdroidserver/-/jobs/4060582594
But I look at it, and it looks right that it fails. So how on earth
was it succeeding before? Basically the `os.getenv('ANDROID_HOME')`
returns `None` when `ANDROID_HOME` is not set. It is not set in both the
jobs, so how did it not stacktrace before?
make_v0() modified the apps data structure to copy an anti feature for a
specific version to all versions resulting in index-v1 and -v2 to
contain wrong anti feature annotations. This patch fixes this and adds a
test that the data structure is no longer modified.
The bug shadowed bugs in the AF implementation of -v1 and -v2 resulting
in not coping the version specific data. This is corrected as well.
This is also tested now.
For -v2 the AF dict is now sorted to make the result reproducible.
Finally The NoSourceSince AF was added as a per version and overall AF
in -v1 and is now only applied as an overall AF and the test is updated
accordingly.
This script is some of the only documentation of certain processes that
run in the buildserver setup. It is not really maintained, but is still
here as a kind of reference.
Signing files is a slow operation, especially with jarsigner. This speeds
up the full test run from 10-12 seconds to 2-3 seconds, which makes it
possible to run the tests interactively again. And it stops signing the
file entirely for tests that do not even touch that file.
In the long run, it would probably make sense to have each test case sign
the file as it needs it, but that's a much bigger change.
Things like apksigner and @obfusk's tools handle this now.
jarsigner is used in the test, since that's the most common use of
`common.find_sdk_tools_cmd()`.
closes#1100
* for f in locale/*/LC_MESSAGES/fdroidserver.po; do msgattrib --set-obsolete --no-wrap --ignore-file=locale/fdroidserver.pot -o $f $f; done
* sed -i 's, \.\./fdroidserver/stats\.py,,' locale/*/LC_MESSAGES/fdroidserver.po
Ran this to generate/refresh the test index files:
`cd tests; ../fdroid update --nosign --pretty`
* converts config.py description to a single line, since the values are no
longer stripped, so this give the same output.
closes#835
This adds the option to configure which set of signatures `fdroid
scanner` should use, by configuring it in `config.yml`. It allows
fetching signatures in our custom json format. It also adds 3 additional
sources: 'suss', 'exodus', 'etip'
makebuildserver.config.py is no more, builder/Vagrantfile is now where the
CPU and memory is configured for the buildserver VM. In fact, that was
always the actual place, the makebuildserver.config.py thing was just
confused.
This should have been part of !1222
openjdk-11 11.0.17 in Debian unstable fails to verify weak signatures:
jarsigner -verbose -strict -verify tests/signindex/guardianproject.jar
131 Fri Dec 02 20:10:00 CET 2016 META-INF/MANIFEST.MF
252 Fri Dec 02 20:10:04 CET 2016 META-INF/1.SF
2299 Fri Dec 02 20:10:04 CET 2016 META-INF/1.RSA
0 Fri Dec 02 20:09:58 CET 2016 META-INF/
m ? 48743 Fri Dec 02 20:09:58 CET 2016 index.xml
s = signature was verified
m = entry is listed in manifest
k = at least one certificate was found in keystore
? = unsigned entry
- Signed by "EMAILADDRESS=root@guardianproject.info, CN=guardianproject.info, O=Guardian Project, OU=FDroid Repo, L=New York, ST=New York, C=US"
Digest algorithm: SHA1 (disabled)
Signature algorithm: SHA1withRSA (disabled), 4096-bit key
WARNING: The jar will be treated as unsigned, because it is signed with a weak algorithm that is now disabled by the security property:
jdk.jar.disabledAlgorithms=MD2, MD5, RSA keySize < 1024, DSA keySize < 1024, SHA1 denyAfter 2019-01-01, include jdk.disabled.namedCurves
getsig.java was replaced by a Python implementation in 6e2d0a9e (2014)
and the test was only there to compare the results for the transition.
Dropping this as it no longer works starting with 11.0.17+8.
Make sudo, init prebuild, build and Prepare fields lists and only
concatenate them with '; ' before execution. This allows arbitrary
commands inside the fileds (even && and ';') as we don't need to split
the commands again for rewritemeta.
- Update usage documentation.
- Use argparse.
- Simplify config handling (similar to readmeta.py).
- Drop code for fdroidserver before 0.7.0.
- Reformat with black.
13016c5d63 in !602 used a set to prevent
duplicate entries, but that worked poorly because it required lots of
data wrapping. Instead, just normalize to JSON, then equality is easy.
The buildserver VM has not been upgraded yet to bullseye, so it is still on
Debian/stretch. The buildserver VM does not need to run `fdroid update`,
`fdroid signindex`, etc. so this new apksigner requirement should not
affect app builds even though they are stuck on Debian/stretch.
The current signing method uses apksigner to sign the JAR so that it
will automatically select algorithms that are compatible with Android
SDK 23, which added the most recent algorithms:
https://developer.android.com/reference/java/security/Signature
This signing method uses then inherits the default signing algothim
settings, since Java and Android both maintain those. That helps
avoid a repeat of being stuck on an old signing algorithm. That means
specifically that this call to apksigner does not specify any of the
algorithms.
The old indexes must be signed by SHA1withRSA otherwise they will no
longer be compatible with old Androids.
apksigner 30.0.0+ is available in Debian/bullseye, Debian/buster-backports,
Ubuntu 21.10, and Ubuntu 20.04 from the fdroid PPA. Here's a quick way to
test:
for f in `ls -1 /opt/android-sdk/build-tools/*/apksigner | sort ` /usr/bin/apksigner; do printf "$f : "; $f sign --v4-signing-enabled false; done
closes#1005
If a project uses fdroidserver as a library, then just calls
common.get_apk_id(), it will now work. Before, that project would have had
to include something like `common.config = {}` to avoid a stacktrace.
This needs a rerun of `fdroid update --clean`.
In case a build is disabled delete_disabled_builds takes care of
deleting it from the repo. But this only works if the apk follows the
normal name pattern. Otherwise it will stay in the folder and be picked
up by process_apks and added to the index.
Closes: #1002
With ~index-v2, the model is changing to offer the plain JSON file for easy
consumption. Then gpgsign will also provide a detached PGP signature for
systems that would rather verify based on PGP signatures than JAR signatures.
!1080closes#969
In case the version information is inside a submodule we need to
checkout the submodule at the version of the tag we test.
Found with org.courville.nova.
Closes: #622
Traceback (most recent call last):
File "/builds/fdroid/fdroidserver/./tests/gradle-release-checksums.py", line 130, in <module>
mr = project.mergerequests.create({
File "/usr/lib/python3/dist-packages/gitlab/exceptions.py", line 281, in wrapped_f
raise error(e.error_message, e.response_code, e.response_body) from e
gitlab.exceptions.GitlabCreateError: 409: ['Another open merge request already exists for this source branch: !1064']
In case the app repository has a broken submodule, checkupdates failed
and did not search for any version updates. Ignoring the error let's us
at least find new version in the main repo (which is probably the right
place anyhow) and thus an improvement.
allowlist and blocklist are much clearer terms with no cultural baggage.
This changes all "whitelist" references to "allowlist", and all "blacklist"
references to "blocklist".
config.yml requires ASCII or UTF-8 encoding because this code does not
auto-detect the file's encoding. That is left up to the YAML library.
YAML allows ASCII, UTF-8, UTF-16, and UTF-32 encodings. Since it is a
good idea to manage config.yml (WITHOUT PASSWORDS!) in git, it makes
sense to use a globally standard encoding.
Android Studio recommends "you use UTF-8 encoding whenever possible",
so this code assumes the files use UTF-8. UTF-8 is also the default
encoding on GNU/Linux and macOS.
https://sites.google.com/a/android.com/tools/knownissues/encoding
Windows will probably default to UTF16, since that's the native
encoding for files. So forcing things to use UTF-8 should help
compatibility.
Windows seems to require this, otherwise this happens:
Traceback (most recent call last):
File "tests/update.TestCase", line 737, in test_translate_per_build_anti_features
apps = fdroidserver.metadata.read_metadata(xref=True)
File "C:\Users\travis\build\fdroidtravis\fdroidserver\fdroidserver\metadata.py", line 813, in read_metadata
app = parse_metadata(metadatapath, appid in check_vcs, refresh)
File "C:\Users\travis\build\fdroidtravis\fdroidserver\fdroidserver\metadata.py", line 1023, in parse_metadata
parse_yaml_metadata(mf, app)
File "C:\Users\travis\build\fdroidtravis\fdroidserver\fdroidserver\metadata.py", line 1073, in parse_yaml_metadata
yamldata = yaml.safe_load(mf)
File "C:\python37\lib\site-packages\yaml\__init__.py", line 162, in safe_load
return load(stream, SafeLoader)
File "C:\python37\lib\site-packages\yaml\__init__.py", line 112, in load
loader = Loader(stream)
File "C:\python37\lib\site-packages\yaml\loader.py", line 34, in __init__
Reader.__init__(self, stream)
File "C:\python37\lib\site-packages\yaml\reader.py", line 85, in __init__
self.determine_encoding()
File "C:\python37\lib\site-packages\yaml\reader.py", line 124, in determine_encoding
self.update_raw()
File "C:\python37\lib\site-packages\yaml\reader.py", line 178, in update_raw
data = self.stream.read(size)
File "C:\python37\lib\encodings\cp1252.py", line 23, in decode
return codecs.charmap_decode(input,self.errors,decoding_table)[0]
UnicodeDecodeError: 'charmap' codec can't decode byte 0x81 in position 37: character maps to <undefined>
gotorevisionx tries to clean up the git repo before checking out a new
revision. In b848b99ba this was changed to reset and clean any submodule
as well. In case upstream has a broken submodule configuration this
could fail and we can't checkout the new revision. As we are doing a
reset and clean after checking out the new revision anyhow, this change
ignores submodule errors before the checkout and only makes sure that
the main repo is reset and clean.
This broke checkupdates for apps where old versions had broken
submodules. It checkout out the old version and got stuck, not able to
checkout any other version.
Use the tag as the version code if no regex was specified. This allows:
UpdateCheckData: '|||'
meaning the tag should be used for version code and name.
com.github.jameshnsears.quoteunquote defines flavours 'fdroid' and
'fdroidS'. The old code used flavour in line, which matches both and the
wrong one was selected.
Closes: #912
