| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
This patch reverts the semantically significant parts of #2890 due to the
issues discussed on #3360 (security concerns when used with autosign,
inconsistency between REST & XMLRPC semantics) but leaves the semantically
neutral changes (code cleanup, added tests) in place.
This patch is intended for 0.25.x, but may also be applied as a step in the
resolution of #3450 (refactored #2890, add "remove_certs" flag) in Rolwf.
|
|
|
|
|
|
|
|
|
|
| |
This will rarely be used, but it enables even more architectural
flexibility, such as precompiling catalogs and storing them in memcached
or equivalent. With this setup, a single host can probably serve all
catalogs and you would then just have as many compiling hosts as
needed.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
This basically allows a sysadmin to control when a client
will compile a new catalog - with this option enabled,
the client will use the cached catalog as long as it has
one, only recompiling when run with the option disabled.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The following manifest:
case $var {
/match/: {
if $var =~ /matchagain/ {
}
}
}
is failing because the "=~" operators when matching sets an ephemeral
variable in the scope. But the case regex also did it, and since they
both belong to the same scope, and Puppet variables are immutables, the
scope raises an error.
This patch fixes this issue by adding to the current scope a stack
of ephemeral symbol tables. Each new match operator or case/selector
with regex adds a new scope. When we get out of the case/if/selector
structure the scope is reset to the ephemeral level we were when
entering it.
This way the following manifest produces the correct output:
case $var {
/match(rematch)/: {
notice("1. \$0 = $0, \$1 = $1")
if $var =~ /matchagain/ {
notice("2. \$0 = $0, \$1 = $1")
}
notice("3. \$0 = $0, \$1 = $1")
}
}
notice("4. \$0 = $0")
And the output is:
1. $0 = match, $1 = rematch
2. $0 = matchagain, $1 = rematch
3. $0 = match, $1 = rematch
4. $0 =
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
| |
It would fail if a directory unexpectedly existed.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
There are still a few unported tests, but it's at least
better now.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
|
|
| |
If we were removing a field, which really only 'cron'
does, then we got an exception.
This is fixed, and now tested.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
It was only apparently working with constant keys,
not, say, AST strings.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
Most of it was just complicated, and the rest
useless.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
This was failing because it was expecting a call to
Puppet::Type#evaluate, which is no longer used.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The point of this test was to confirm that we could take
a given file resource and have it work multiple times in
memory, but that's not actually possible given our current
code. We copy the values over from the source metadata,
and suddenly have no method for differentiation between
values set by the user and those set by the remote source.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
|
|
|
| |
This is probably a slight refactor, but only because
it fixed a bug (content not being copied over correctly
from metadata), which required that slight refactor.
Mostly this just makes the code a bit cleaner.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
Also making them less likely to try to modify
the local filesystem in any way.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
signature of.
|
| |
|
|
|
|
|
| |
spec/unit/provider/mount/parsed.rb would show a failure when
spec/unit/type/mount.rb had been run prior.
|
| |
|
|
|
|
|
|
| |
Also making the code a bit cleaner.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
| |
This involved making some tests better, but mostly
just involved fixing calls to use new APIs and such.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It is a setting that was added years ago as a backward
compatibility option and even if it still works, which
is questionable, it has no purpose any longer.
It just complicated the code and didn't do much, so it's gone
now.
Also simplified the interface of Leaf#evaluate_match, since it
was now using none of the passed-in options.
Finally, removed/migrated the last of the Selector/CaseStatement
test/unit tests.
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@puppetlabs.com>
|
|
|
|
| |
"sensitive"
|
| |
|
|
|
|
| |
failures
|
|
|
|
|
|
|
|
|
| |
At the same time I removed all of the code in checksum
that managed tracking changes to the checksum over time.
I'll add it back in as I fix the fact that changes aren't
being tracked like the should at the moment.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
Replaced use of a stub resource with a real resource.
This is in preparation for turning the 'checksum' property
into a parameter.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Also changed the internals - we're no longer using
Resource instances with the ruby block, instead
we're using a simple new class. We had to do this
because Resource has too many methods - e.g.,
'file' returned the file name rather than
created a new resource type.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
This is a simplistic DSL - you can create
resource types (defined resources), classes,
and nodes, and they can call functions and
create resources. Nothing else, at this point.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
| |
We just use the scope's environment.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch moves the syntactic aspects of string interpolation up
into the lexer/parser phase, preparatory to moving the semantic
portions down to the as yet unnamed futures resolution phase.
This is an enabling move, designed to allow:
* Futures resolution in and between interpolated strings
* Interpolation of hash elements into strings
* Removal of certain order-dependent paths
* Further modularization of the lexer/parser
The key change is switching from viewing strings with interpolation
as single lexical entities (which await later special case processing)
to viewing them as formulas for constructing strings, with the internal
structure of the string exposed by the parser.
Thus a string like:
"Hello $name, are you enjoying ${language_feature}?"
internally becomes something like:
concat("Hello ",$name,", are you enjoying ",$language_feature,"?")
where "concat" is an internal string concatenation function.
A few test cases to show the user observable effects of this change:
notice("string with ${'a nested single quoted string'} inside it.")
$v2 = 3+4
notice("string with ${['an array ',3,'+',4,'=',$v2]} in it.")
notice("string with ${(3+5)/4} nested math ops in it.")
...and so forth.
The key changes in the internals are:
* Unification of SQTEXT and DQTEXT into a new token type STRING (since
nothing past the lexer cares about the distinction.
* Creation of several new token types to represent the components of
an interpolated string:
DQPRE The initial portion of an interpolated string
DQMID The portion of a string betwixt two interpolations
DQPOST The final portion of an interpolated string
DQCONT The as-yet-unlexed portion after an interpolation
Thus, in the example above (phantom curly braces added for clarity),
DQPRE "Hello ${
DQMID }, are you enjoying ${
DQPOST }?"
DQCONT is a bookkeeping token and is never generated.
* Creation of a DOLLAR_VAR token to strip the "$" off of variables
with explicit dollar signs, so that the VARIABLEs produced from
things like "Test ${x}" (where the "$" has already been consumed)
do not fail for want of a "$"
* Reworking the grammar rules in the obvious way
* Introduction of a "concatenation" AST node type (which will be going
away in a subsequent refactor).
Note finally that this is a component of a set of interrelated refactors,
and some of the changes around the edges of the above will only makes
sense in context of the other parts.
|
|
|
|
|
|
|
|
|
|
|
|
| |
I had only done this partway, because it seemed easier,
but not surprisingly, it ended up being more complex.
In addition to those renames, this commit includes fixes
to whatever tests I needed to fix to confirm that things
were again working. I think most of these broken
tests have been broken for a while.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
When a class is evaluated, its parent class
needs to be evaluated first. This forces that
evaluation. We somehow lost it when we converted
the resource types out of AST.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This code is impressively difficult, because
sometimes resource types act like resources (classes
and nodes are singletons) and sometimes like resource
types (defined and builtin resources).
So, to get nodes to show as Node[foo] and classes as
Class[Foo::Bar], but defined resources to show up as
Foo::Bar[baz], we have to do some silliness.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
| |
I was using 'params' and 'parameters', so
I fixed that and extracted the differences in
how they handle parameters into a stubbable method.
This allowed me to almost entirely remove the subclass's
'initialize' method.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
| |
This is used for AST resources (and fixed the last
of the tests I broke in spec/).
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
| |
This involves a bit of refactoring in the rest
of the code to make it all work, but most of the
changes are fixing or removing old tests.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit is hopefully less messy than it
first appears, but it's certainly cross-cutting.
The reason for all of this is that we previously only
looked up builtin resource types from outside the parser,
but now that the defined resource types are available globally
via environments, we can push that lookup code to Resource.
Once we do that, however, we have to have environment and
namespace information in every resource.
Here I remove the Resource::Reference classes (except
the AST class), and use Resource instances instead. I
did this because the shared code between the two classes
got incredibly complicated, such that they should have had
a hierarchical relationship disallowed by their constants.
This complexity convinced me just to get rid of References
entirely.
I also make Puppet::Parser::Resource a subclass
of Puppet::Resource.
There are still broken tests in test/, but this was a big
enough commit I wanted to get it in.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
This can happen because we're almost always converting
to environment instances from strings.
Shouldn't happen often, but it's easier to be more failure-tolerant.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
| |
We previously only supported a single namespace when searching for
resource types et al, but the whole system actually relies on
an array of namespaces and search paths, so this adds
that functionality all the way down, as it were.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This will allow us to remove all of the parameter
validation from the other Resource classes.
This is possible because resource types defined
in the language are visible outside of the parser,
via the environment.
This will enable lots of code removal and simplication.
Signed-off-by: Luke Kanies <luke@reductivelabs.com>
|