chiark / gitweb /
secnet.git
4 years agoREADME: Introduce spec for "base91s" encoding
Ian Jackson [Sun, 1 Dec 2019 16:33:21 +0000 (16:33 +0000)]
README: Introduce spec for "base91s" encoding

We have no implementations of this yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoprivcache: Remove some spurious \n in messages
Ian Jackson [Sat, 23 Nov 2019 23:41:19 +0000 (23:41 +0000)]
privcache: Remove some spurious \n in messages

I really oughtn't to have done this, but there is much code in secnet
which sets a bad example.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoprivcache: Better debug
Ian Jackson [Sat, 23 Nov 2019 19:13:13 +0000 (19:13 +0000)]
privcache: Better debug

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoprivcache: Use new pathprefix_template facility
Ian Jackson [Tue, 8 Oct 2019 16:22:49 +0000 (17:22 +0100)]
privcache: Use new pathprefix_template facility

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoprivcache: New closure for signature key handling
Ian Jackson [Sun, 29 Sep 2019 13:21:19 +0000 (14:21 +0100)]
privcache: New closure for signature key handling

This will be used by site.c, when it needs to load a key to make a
signature as demanded by the peer.

This commit introduces 5 spurious \n's in messages.
It also introduces a few ` = ' style errors.
These will be removed later.

No callers yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoutil: pathprefix_template: New facility
Ian Jackson [Tue, 8 Oct 2019 14:45:50 +0000 (15:45 +0100)]
util: pathprefix_template: New facility

We will use this for constructing public and private key filenames to
load.

No callers yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigscheme: Interface for signature schemes
Ian Jackson [Sun, 29 Sep 2019 13:34:00 +0000 (14:34 +0100)]
sigscheme: Interface for signature schemes

There are no implementations yet.  This is split out like this for
review of the api.

We use struct buffer_if to feed key data into the scheme, in both
cases.  This will be convenient for implementations, The loadpub
caller will have the data from base91s, probably.  The loadpriv caller
will have it from a file, but we read the file in the common code.

The API we are introducing now expects each private key buffer to be
fed to the sig schemes one by one until they it finds one that likes
it.  This is in fact not necessary; it was needed in an earlier design
which does not otherwise survive in the published git history.  This
rather suboptimal API will be changed later.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigpubkey/sigprivkey: Provide a dispose() method
Ian Jackson [Sun, 29 Sep 2019 13:21:31 +0000 (14:21 +0100)]
sigpubkey/sigprivkey: Provide a dispose() method

This is needed allow the site code to load and unload keys
dynamically.

This is just the interface, with no implemntation and no callers,
split out like this for early review.

The two xxx's will be fixed rather later in
  "rsa1: Provide dispose functions"
(but, obviously, before we add any callers).

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigpubkey/sigprivkey: Provide a hash_if
Ian Jackson [Sun, 29 Sep 2019 12:58:02 +0000 (13:58 +0100)]
sigpubkey/sigprivkey: Provide a hash_if

Sometimes, the caller is going to want to hash things for themselves.
(Notably, site.c will want to hash the advertisements of public key
set updates.)

The hash to be used should be the one used by the signature scheme, as
that has the same security properties.

In rsa.c this moves the hash_if from common to ops, and it is now
convenient to abolish rsa_hash which is a pointless veneer over
hash_hash.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigkey handling: Introduce serialt type
Ian Jackson [Tue, 8 Oct 2019 14:51:41 +0000 (15:51 +0100)]
sigkey handling: Introduce serialt type

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigkey handling: define MAX_SIG_KEYS
Ian Jackson [Sun, 29 Sep 2019 19:28:46 +0000 (20:28 +0100)]
sigkey handling: define MAX_SIG_KEYS

This is going to control the number of different public keys we will
use from a keyset, and also the number we will look at when listed as
acceptable by a relying peer.

No users yet in this commit, though.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosigkey handling: Introduce sigkeyid type
Ian Jackson [Sun, 29 Sep 2019 19:27:30 +0000 (20:27 +0100)]
sigkey handling: Introduce sigkeyid type

We provide helper macros for use with printf.

The reference to pubkeys.fl.pl comment is to code which doesn't exist
yet but comes later.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agopubkey handling: Document key sets, id, etc. plan
Ian Jackson [Thu, 28 Nov 2019 13:57:44 +0000 (13:57 +0000)]
pubkey handling: Document key sets, id, etc. plan

None of this is implemented yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agomtest/t-prefix: Drop redundant setting of seddery
Ian Jackson [Wed, 4 Dec 2019 21:55:15 +0000 (21:55 +0000)]
mtest/t-prefix: Drop redundant setting of seddery

This is in mtest/common.tcl already.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agorsa: emsa_pkcs1: Fix a message
Ian Jackson [Sat, 23 Nov 2019 19:52:43 +0000 (19:52 +0000)]
rsa: emsa_pkcs1: Fix a message

This function is called during verification too, so the message
shouldn't talk about signing.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoconffile: find_cl_if: fix fail_if_invalid==False
Ian Jackson [Sun, 6 Oct 2019 22:34:33 +0000 (23:34 +0100)]
conffile: find_cl_if: fix fail_if_invalid==False

If the key was not found, this would segfault, so passing False didn't
work properly.  Currently nothing does, so this bug is latent.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agomake-secnet-sites: Tainted: Fix a lot of bad return values
Ian Jackson [Fri, 8 Nov 2019 21:18:05 +0000 (21:18 +0000)]
make-secnet-sites: Tainted: Fix a lot of bad return values

This code had remnants of a previously intended calling convention,
where object return chaining would be used.

Unfortunately in the currently used calling style, we expect to get a
boolean back everywhere, where true meas `ok'.  Returning `self' is
always treated as `ok' because it's trueish.

Luckily this doesn't cause actual security bugs because we always
return from all of the top-level entrypoints via ._rtn[val] which
checks the ._ok setting, which does properly track problems.  So we
fail an assertion rather than printing a nice message.  This is not
pretty but it is not a vulnerability.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Drop a redundant test
Ian Jackson [Tue, 3 Dec 2019 22:10:08 +0000 (22:10 +0000)]
stest: Drop a redundant test

We were going to use this for key rollover tests etc.  But in fact
we (much later in this series) will introduce new tests for the new
key loading arrangements.  Actual key rollover will come *much* later
and there is little point having this placeholder in the meantime.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoautogen.sh: Write a comment about need for autoheader
Ian Jackson [Fri, 6 Dec 2019 13:42:50 +0000 (13:42 +0000)]
autogen.sh: Write a comment about need for autoheader

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoconfigure.ac: Do not check for lack of standard headers
Ian Jackson [Fri, 6 Dec 2019 13:27:18 +0000 (13:27 +0000)]
configure.ac: Do not check for lack of standard headers

This makes configure rather faster.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoconfigure.ac: Drop AC_STDC_HEADERS
Ian Jackson [Fri, 6 Dec 2019 13:37:23 +0000 (13:37 +0000)]
configure.ac: Drop AC_STDC_HEADERS

This is obsolete, as the docs say:

 AC_STDC_HEADERS
     Replaced by 'AC_HEADER_STDC'

 AC_HEADER_STDC
     This macro is obsolescent, as current systems have conforming
     header files.  New programs need not use this macro.

Unfortunately dropping it is not properly effective because it's
implied by some other stuff.  We are going to deal with that in a
moment.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoconfigure.ac: Drop checks for systems lacking stdint.h
Ian Jackson [Fri, 6 Dec 2019 13:03:46 +0000 (13:03 +0000)]
configure.ac: Drop checks for systems lacking stdint.h

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Check that received packet is as expected
Ian Jackson [Sun, 24 Nov 2019 00:59:39 +0000 (00:59 +0000)]
stest: Check that received packet is as expected

It turns out that if secnet decides the link to the peer is too
broken, it loops outbound packets back to its host.  Normally this is
not obvious, because the host naturally drops such packets.

But in stest we weren't actually looking at the incoming packets at
all, so we would treat this as success!

So, instead, match the start of the packet, including the source and
destination addresses and the icmp type.  (This crude match will fail
if the reply contains any IP options, but the ping comes from the peer
secnet so it won't have.)

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Decode the slip packets that come via fake netlink
Ian Jackson [Sun, 24 Nov 2019 00:59:31 +0000 (00:59 +0000)]
stest: Decode the slip packets that come via fake netlink

This will allow us to do something more sophisticated with the packet
contents.

The only functional change right now is to ignore empty `packets'.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: If one of our secnets dies, call the test a failure
Ian Jackson [Mon, 25 Nov 2019 00:01:34 +0000 (00:01 +0000)]
stest: If one of our secnets dies, call the test a failure

This involves use of TclX's `signal' facility.  In my tests it was
easy to make Tcl deadlock by doing too much work in the signal
handler.  In particular reaping children is a bad idea.  Also signals
are not blocked during the signal handler so it would have to be
reentrant.

Instead, use `after idle'.  That is quite soon enough for the reap to
run, and in my tests with TclX 8.4 it all works properly.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Use `finish' for success exits
Ian Jackson [Sun, 24 Nov 2019 22:20:51 +0000 (22:20 +0000)]
stest: Use `finish' for success exits

vwait is a bit funny.  This is better, and it also kills our child
secnets.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Kill our child secnets when we call finish
Ian Jackson [Mon, 25 Nov 2019 00:01:01 +0000 (00:01 +0000)]
stest: Kill our child secnets when we call finish

This stops them hanging around and, often, printing more stuff after
the test has actually failed.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Record the pids of of the secnets we spawn
Ian Jackson [Mon, 25 Nov 2019 00:00:30 +0000 (00:00 +0000)]
stest: Record the pids of of the secnets we spawn

Nothing uses this yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Replace the call to `exit 1' with a new proc `finish'
Ian Jackson [Sun, 24 Nov 2019 13:59:29 +0000 (13:59 +0000)]
stest: Replace the call to `exit 1' with a new proc `finish'

This is going to be used for success exits too, in a moment.
This will let us do more work when we are exiting.

The only change for now is some extra stderr output.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Support not running secnet ourselves
Ian Jackson [Sat, 23 Nov 2019 00:38:01 +0000 (00:38 +0000)]
stest: Support not running secnet ourselves

Now the user can set SECNET_STEST_DIVERT_inside=1 (say) and instead of
actually running secnet, we print the rune we would run and ask the
user when they have started it.

A freshly started secnet (with this config, at least) just sits there
waiting for its first packet to deal with.  So if the user does this
for the 2nd secnet, that's fine - the first one will happily wait
until the test actually gets going.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Don't prefix ./ with another ./, in socktmp
Ian Jackson [Sat, 23 Nov 2019 00:37:35 +0000 (00:37 +0000)]
stest: Don't prefix ./ with another ./, in socktmp

Makes things slightly prettier.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Print the special env vars we pass to secnet
Ian Jackson [Sat, 23 Nov 2019 00:19:11 +0000 (00:19 +0000)]
stest: Print the special env vars we pass to secnet

We don't dump the whole environment, just selected bits.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Print the secnet runes we run
Ian Jackson [Sat, 23 Nov 2019 00:03:58 +0000 (00:03 +0000)]
stest: Print the secnet runes we run

We can't use prexec here because we aren't using exec at all.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests, mtest: Provide and use `prexec' for helping with debugging
Ian Jackson [Fri, 22 Nov 2019 23:59:32 +0000 (23:59 +0000)]
tests, mtest: Provide and use `prexec' for helping with debugging

This prints the make-secnet-sites runes we are about to run.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agodir-locals: Provide python-indent-offset too
Ian Jackson [Fri, 29 Nov 2019 23:37:01 +0000 (23:37 +0000)]
dir-locals: Provide python-indent-offset too

python-indent is obsolete since Emacs 24.3.  It has been made into an
alias for python-indent-offset, but the alias is not marked safe as a
file variable.  Specifying python-indent-offset should make this work
on all relevant versions of Emacs (at least in the default
configuration).

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Add debug logs
Ian Jackson [Sat, 23 Nov 2019 19:13:46 +0000 (19:13 +0000)]
stest: Add debug logs

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Use new `prefix' option.
Ian Jackson [Sat, 23 Nov 2019 17:57:56 +0000 (17:57 +0000)]
stest: Use new `prefix' option.

This lets us distinguish output from our two secnets.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agostest: Use stderr, not tty, for logging
Ian Jackson [Sat, 23 Nov 2019 17:50:46 +0000 (17:50 +0000)]
stest: Use stderr, not tty, for logging

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agologfile: New `prefix' option.
Ian Jackson [Sat, 23 Nov 2019 17:57:46 +0000 (17:57 +0000)]
logfile: New `prefix' option.

This allows the config file to add a fixed string to log messages.
This will be useful in our test suite, which mixes up output from
two instances of secnet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agologfile: Do not log differently with --nodetach
Ian Jackson [Sat, 23 Nov 2019 17:52:31 +0000 (17:52 +0000)]
logfile: Do not log differently with --nodetach

This is actually quite anoying.  There is no good reason for leaving
off the timestamps in tests, or manual debugging, etc.

If at some point we want to support running under some kind of
log-capturing daemon supervisor which timestamps everything, we can
easily make this conditional again.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agologfile: Remove redundant check for st->f
Ian Jackson [Sat, 23 Nov 2019 17:51:04 +0000 (17:51 +0000)]
logfile: Remove redundant check for st->f

st->f is now always non-null.  If no filename is specified, it is
stderr.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agologfile: Log to stderr by default (`filename' key now optional)
Ian Jackson [Sat, 23 Nov 2019 17:50:28 +0000 (17:50 +0000)]
logfile: Log to stderr by default (`filename' key now optional)

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosecnet: Make stderr line buffered
Ian Jackson [Sun, 24 Nov 2019 01:00:20 +0000 (01:00 +0000)]
secnet: Make stderr line buffered

This helps not interleave output in tests.  We write output in lines,
so this is fine.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoMerge subdirmk 0.3
Ian Jackson [Mon, 30 Dec 2019 13:14:45 +0000 (13:14 +0000)]
Merge subdirmk 0.3

4 years agotests: Provide make-release script
Ian Jackson [Mon, 30 Dec 2019 12:59:57 +0000 (12:59 +0000)]
tests: Provide make-release script

Not really `tests' but I don't want to add a whole new directory just
for this.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests: Provide advance-tested script
Ian Jackson [Mon, 30 Dec 2019 12:50:30 +0000 (12:50 +0000)]
tests: Provide advance-tested script

Useful for our own testing.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosubdirmk: Merge new version and fix everything
Ian Jackson [Mon, 30 Dec 2019 12:06:47 +0000 (12:06 +0000)]
subdirmk: Merge new version and fix everything

git subtree pull, but many incompatible changes - so also fixed up the
secnet code to match.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoMerge subdirmk
Ian Jackson [Mon, 30 Dec 2019 12:06:12 +0000 (12:06 +0000)]
Merge subdirmk

4 years agoMerge subdirmk
Ian Jackson [Sun, 29 Dec 2019 14:23:00 +0000 (14:23 +0000)]
Merge subdirmk

4 years agoLegal: Update tests/filter/ inputs and outputs
Ian Jackson [Mon, 30 Dec 2019 11:58:11 +0000 (11:58 +0000)]
Legal: Update tests/filter/ inputs and outputs

We judiciously use # vs. &# for the legal comment so that each
expected output file contains one notice.

The following files, which are wholly constructed by running
generate and reviewing the diffs, do not have a legal notice:
  tests/filter/main.mk.expected
  tests/filter/stderr.expected
  tests/filter/sub/Dir.mk.expected
This is OK I think.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoLegal: Add copyright, licence and warranty notice to many files
Ian Jackson [Mon, 30 Dec 2019 11:17:17 +0000 (11:17 +0000)]
Legal: Add copyright, licence and warranty notice to many files

The .sd.mk and .expected files in tests/filter/ are more complicated.
We'll do those in a moment.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoLegal: Add NO WARRANTY everywhere
Ian Jackson [Mon, 30 Dec 2019 11:09:36 +0000 (11:09 +0000)]
Legal: Add NO WARRANTY everywhere

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Avoid $err_file undefined warning during startup
Ian Jackson [Mon, 30 Dec 2019 11:02:59 +0000 (11:02 +0000)]
generate: Avoid $err_file undefined warning during startup

If, for example, one of the input files implied by the command line
arguments is missing, err is called with $err_file not yet set.

Handle this case.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Mention suppressions when warning re VAR vs &VAR
Ian Jackson [Sun, 29 Dec 2019 15:56:44 +0000 (15:56 +0000)]
generate: Mention suppressions when warning re VAR vs &VAR

When we are actually printing the warning, report all the locations,
including the suppressed ones.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoSyntax: Support &:local+global !...
Ian Jackson [Sun, 29 Dec 2019 15:50:28 +0000 (15:50 +0000)]
Syntax: Support &:local+global !...

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests/filter: Test &:local+global & vs non-&
Ian Jackson [Sun, 29 Dec 2019 15:46:54 +0000 (15:46 +0000)]
tests/filter: Test &:local+global & vs non-&

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoSyntax: Rescope effect of &:local+global
Ian Jackson [Sun, 29 Dec 2019 15:18:39 +0000 (15:18 +0000)]
Syntax: Rescope effect of &:local+global

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agosubdirmk: Fix a Subdir left over in README
Ian Jackson [Sun, 29 Dec 2019 14:32:49 +0000 (14:32 +0000)]
subdirmk: Fix a Subdir left over in README

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoSyntax: Expand &$( and &$NN to use ${ } rather than $( )
Ian Jackson [Sun, 29 Dec 2019 14:17:00 +0000 (14:17 +0000)]
Syntax: Expand &$( and &$NN to use ${ } rather than $( )

And now document the use of $&+ for recipes.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agodoctests: Allow parenthetical comments as an expansion RHS
Ian Jackson [Sun, 29 Dec 2019 14:18:49 +0000 (14:18 +0000)]
doctests: Allow parenthetical comments as an expansion RHS

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Swap order of &$( and &$NN
Ian Jackson [Sun, 29 Dec 2019 14:15:15 +0000 (14:15 +0000)]
README: Swap order of &$( and &$NN

We are going to add a note to &$( which applies to &$NN too, and this
makes it nicer.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoSyntax: Use ${top_srcdir} rather than $(top_srcdir)
Ian Jackson [Sun, 29 Dec 2019 14:13:52 +0000 (14:13 +0000)]
Syntax: Use ${top_srcdir} rather than $(top_srcdir)

Again, this makes it possible to use in $-doubled shell runes as well
as in make syntax.  (Assuming you have made top_srcdir be a shell
variable as well as a make variable.)

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoSyntax: expand &$... to ${....} rather than $(....)
Ian Jackson [Sun, 29 Dec 2019 14:03:28 +0000 (14:03 +0000)]
Syntax: expand &$... to ${....} rather than $(....)

This makes it possible to use in $-doubled shell runes as well as in
make syntax.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Introduce new `broken-var-ref' warning
Ian Jackson [Sun, 29 Dec 2019 14:01:22 +0000 (14:01 +0000)]
Warnings: Introduce new `broken-var-ref' warning

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests/filter: Print all the diffs, not just the first
Ian Jackson [Sun, 29 Dec 2019 14:00:58 +0000 (14:00 +0000)]
tests/filter: Print all the diffs, not just the first

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Refer to example/
Ian Jackson [Sat, 28 Dec 2019 19:07:17 +0000 (19:07 +0000)]
README: Refer to example/

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Move the "how to use" up
Ian Jackson [Sat, 28 Dec 2019 19:06:17 +0000 (19:06 +0000)]
README: Move the "how to use" up

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Add another layer of structure
Ian Jackson [Sat, 28 Dec 2019 19:04:26 +0000 (19:04 +0000)]
README: Add another layer of structure

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Move the Warnings section up into what is going to be spec
Ian Jackson [Sat, 28 Dec 2019 19:04:10 +0000 (19:04 +0000)]
README: Move the Warnings section up into what is going to be spec

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: deinent the macro explaation
Ian Jackson [Sat, 28 Dec 2019 18:55:45 +0000 (18:55 +0000)]
README: deinent the macro explaation

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Further miscellaneous fixes and clarifications
Ian Jackson [Sat, 28 Dec 2019 18:55:25 +0000 (18:55 +0000)]
README: Further miscellaneous fixes and clarifications

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Document that local+global is not 100% accurate
Ian Jackson [Sat, 28 Dec 2019 18:46:58 +0000 (18:46 +0000)]
README: Document that local+global is not 100% accurate

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: More syntax further up
Ian Jackson [Sat, 28 Dec 2019 18:43:18 +0000 (18:43 +0000)]
README: More syntax further up

This is primary.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarning reporting: Report each warning only once
Ian Jackson [Sat, 28 Dec 2019 18:36:31 +0000 (18:36 +0000)]
Warning reporting: Report each warning only once

In particular, Prefix and Suffix, or conventional &:include's can
generate a lot of repeated warnings.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarning reporting: Provide &:local+global directive
Ian Jackson [Sat, 28 Dec 2019 18:26:09 +0000 (18:26 +0000)]
Warning reporting: Provide &:local+global directive

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarning reporting: Add a test of the warning suppression system
Ian Jackson [Sat, 28 Dec 2019 13:55:59 +0000 (13:55 +0000)]
Warning reporting: Add a test of the warning suppression system

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarning reporting: Warning suppression system
Ian Jackson [Sat, 28 Dec 2019 13:55:12 +0000 (13:55 +0000)]
Warning reporting: Warning suppression system

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotest/filter; Cause some warnings, to check they appear
Ian Jackson [Sat, 28 Dec 2019 12:58:03 +0000 (12:58 +0000)]
test/filter; Cause some warnings, to check they appear

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Sort occurrences properly in local+global warnings
Ian Jackson [Sat, 28 Dec 2019 12:52:01 +0000 (12:52 +0000)]
Warnings: Sort occurrences properly in local+global warnings

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Warn for confusing single-char $ expansions
Ian Jackson [Sat, 28 Dec 2019 14:05:19 +0000 (14:05 +0000)]
Warnings: Warn for confusing single-char $ expansions

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Track variable references in &-expansions
Ian Jackson [Sat, 28 Dec 2019 12:31:40 +0000 (12:31 +0000)]
Warnings: Track variable references in &-expansions

For &-escapes which match the whole variable name, we can conveniently
track variable expansions as part of the processing.  (We don't change
any of the matching regexps.)

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Track basic $-references
Ian Jackson [Sat, 28 Dec 2019 12:31:49 +0000 (12:31 +0000)]
Warnings: Track basic $-references

We must always stop at $'s in the input now.

Having stopped at $, we normally just output it and carry on.  Ie, we
only stop so we can do some inspection: as before, we do this variable
tracking as inspection before processing, rather entangled with
processing.

We can deal reasonably properly with ${ } and $( ).

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Track settings of variables
Ian Jackson [Sat, 28 Dec 2019 12:27:40 +0000 (12:27 +0000)]
Warnings: Track settings of variables

Where the variable tracking does not involve &-escapes, or only
involves &-escapes which work like a prefix so the processor does not
have the whole variable name, we do this separately from processing.

This is simpler because we can look ahead more.  It also avoids
disturbing the processing logic (which needs to be precisely accurate,
unlike this variable tracking).

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: Infrastructure for tracking and warning about variables
Ian Jackson [Sat, 28 Dec 2019 12:59:11 +0000 (12:59 +0000)]
Warnings: Infrastructure for tracking and warning about variables

We are going to track when we see FOO or &FOO.  If we see both, we
issue a warning, as that might mean the programmer forgot a &.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoWarnings: generate: Basic `wrn' function
Ian Jackson [Sat, 28 Dec 2019 12:58:46 +0000 (12:58 +0000)]
Warnings: generate: Basic `wrn' function

This is sufficient to add warnning calls to the code.  We'll soup it
up later.  No callers yet.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests/filter/check: Capture generate's stderr
Ian Jackson [Sat, 28 Dec 2019 12:55:17 +0000 (12:55 +0000)]
tests/filter/check: Capture generate's stderr

We are going to cause `generate' to emit warnings, which we want to
check are working.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotest/filter: Test &${ ... } rather better
Ian Jackson [Sat, 28 Dec 2019 23:07:01 +0000 (23:07 +0000)]
test/filter: Test &${ ... } rather better

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Fix &${ } end condition
Ian Jackson [Sat, 28 Dec 2019 23:01:49 +0000 (23:01 +0000)]
generate: Fix &${ } end condition

It is quite wrong to use $ddbl, which might be adjusted by &$- &$+.
We must maintain a separate counter.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Fix { } in &${ to actually output the { }
Ian Jackson [Sat, 28 Dec 2019 12:17:14 +0000 (12:17 +0000)]
generate: Fix { } in &${ to actually output the { }

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agotests/filter: Rename `doctest' -> `doctests' everywhere
Ian Jackson [Sat, 28 Dec 2019 11:34:56 +0000 (11:34 +0000)]
tests/filter: Rename `doctest' -> `doctests' everywhere

This is more uniform.  A lot of churn, though.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Move &:changequote further down
Ian Jackson [Sat, 28 Dec 2019 18:19:17 +0000 (18:19 +0000)]
README: Move &:changequote further down

This is a bit of a minority interest.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Explain a wrinkle in &:include filename semantics
Ian Jackson [Sat, 28 Dec 2019 00:24:15 +0000 (00:24 +0000)]
README: Explain a wrinkle in &:include filename semantics

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: use oraw in a few more places
Ian Jackson [Thu, 26 Dec 2019 02:10:03 +0000 (02:10 +0000)]
generate: use oraw in a few more places

This gets rid of some open-coded `print O'.

No functional change.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Miscellaneous fixes and clarifications
Ian Jackson [Thu, 26 Dec 2019 01:49:49 +0000 (01:49 +0000)]
README: Miscellaneous fixes and clarifications

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Document restriction on $-doubling and lack of -quadrupling
Ian Jackson [Thu, 26 Dec 2019 02:03:07 +0000 (02:03 +0000)]
README: Document restriction on $-doubling and lack of -quadrupling

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoREADME: Move `Tables of file reference syntaxes' down
Ian Jackson [Thu, 26 Dec 2019 01:48:38 +0000 (01:48 +0000)]
README: Move `Tables of file reference syntaxes' down

This is more of an appendix or addendum.  It should come after the
complete description of the substitution syntax.

Pure motion.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoBig incompatible change: Rename `Subdir' to `Dir'
Ian Jackson [Thu, 26 Dec 2019 01:40:09 +0000 (01:40 +0000)]
Big incompatible change: Rename `Subdir' to `Dir'

We want completion to be easy, and we have Suffix now.  `Dir' is as
good as `Subdir' I think (and a bit shorter).

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agoError handling: Fix messages resulting from bad directive arguments
Ian Jackson [Thu, 26 Dec 2019 01:32:43 +0000 (01:32 +0000)]
Error handling: Fix messages resulting from bad directive arguments

"Unknown directive" is not necessarily accurate; another possibility
is that the directive is known but didn't match the parsing regexp
because of problems with the arguments.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Nested scope: Change `Eval' to `eval'
Ian Jackson [Thu, 26 Dec 2019 01:33:01 +0000 (01:33 +0000)]
generate: Nested scope: Change `Eval' to `eval'

This is the nesting kind name.  But it also appears in error messages
where it ought not to be capitalised.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
4 years agogenerate: Nested scope: Change `Macro' to `macro'
Ian Jackson [Thu, 26 Dec 2019 01:32:36 +0000 (01:32 +0000)]
generate: Nested scope: Change `Macro' to `macro'

This is the nesting kind name.  But it also appears in error messages
where it ought not to be capitalised.

We'll change `Eval' in a moment.

Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>