git diff-index by default includes staged and non-staged changes.
The autogen-spicy-docs script copies over spicy-pygments.py from
the Spicy tree into doc/ext.
The job would attempt to commit scripts and script-reference even
though nothing actually is staged when a spurious diff existed in
ext/spicy-pygments.py. Guard from this by using --cached.
The Python docker-compose executable in the docker_builder image is not
functional anymore (TypeError: kwargs_from_env() got an unexpected
keyword argument 'ssl_version').
docker-compose v1 has been sun-setted for quite some time. Do the
minimal to switch to docker compose v2.
We already do have a test that check static builds, but none that tests
the exact same build flags we use for binary builds on OBS. This commit
fixes this oversight.
Related to GH-3647
* origin/topic/vern/script-opt-maint.Mar24:
AST location fixes for -O gen-C++
script optimization support for clearing tables/vectors using "delete"
fix for ZAM printing of instructions that modify slots other than slot 1
fixed bug where appending an empty vector to a vector-of-vector's had no effect
@vpax reported surprising behavior when working with "void values".
While these are not exposed to script land, plumb the places he
pointed out are causing confusing behavior.
Closes#3640.
Since 81a9745fb3, the assert condition is
evaluated twice. This leads to unexpected behavior when cond has a side
effect like publishing a message or creating a log stream or filter.
Found while using the following in ad-hoc testing code and wondering
why two messages were published.
assert publish(Cluster::worker_topic, hello, "abc")
When creating RE_Matcher instances at runtime and verifying the pattern
compiles via Compile(), the syntax_error flag wasn't respected and
Compile() would return true even for some invalid regular expressions.
For example, compiling /a{1,b}/, Compile() would return true even though
it produced a reporter error while parsing complaining about b not
being valid.
This patch improves the error handling, so that calling Compile() returns
false whenever zeek::detail::synerr() was called while a pattern was
parsed. The use-case is creation of patterns at runtime based on
JavaScript strings. These might be entered or received at runtime via
an API. This change allows to be a bit more robust to detect invalid
input and raising exceptions to notify the user.
This also move syntax_error and csize out of global scope.
If RE_Matcher was to be used as an actual API, we likely should squelch
the reporter errors and mark it as not thread safe, but this is a small
step forward.
0a89ca6 doc: Expand zeek.as() description and add an example
7e814d7 Types: Implement basic pattern support
43df9d2 Update docs to provide example of shared Node.js openSSL configuration
3bf2ea5 lsan suppressions: Add some for 21.11
640affa zeek.global_vars: Remove leftover internal field usage
3ee53c7 zeek.global_vars: Implement setter
8144061 zeek.as: Support more Zeek types
b453483 zeek.as: Fix crash for non-atomic types
Zeekygen implements its own make-style update logic to prevent
re-creation of files that have not changed. To fulfill this, we
currently encode the current time into spicyz generated .cc files.
This degrades ccache efficiency for built-in analyzers and also
for all .evt files compiled during testing. Switch SpicyModuleInfo
to return current time instead. This results in the re-generation
of documentation files unconditionally when running Zeekygen, but
that seems more acceptable IMO.
Generally wonder if Zeekygen should produce output unconditionally
and if we need to clobber prevention, compare with the content of
the existing file.
Closes#3619
* origin/topic/awelzel/move-iso-9660-sig-to-policy:
signatures/iso-9660: Add \x01 suffix to CD001
test-all-policy: Do not load iso-9660.zeek
signatures: Move ISO 9660 signature to policy
