Nice!
Jeg har bare lige eet spørgsmål - hvad mener du med "binary grammars as
text in mangled structure", and hvorfor er det nødvendigt? Øger det ikke
risikoen for dekompilering?
-- Eckhard
On 05/20/2017 01:47 AM, Tino Didriksen wrote:
A new release of CG-3 has been tagged v1.0.0.12200, in preparation for
the NoDaLiDa 2017 Constraint Grammar Workshop (
https://visl.sdu.dk/nodalida2017.html ) which takes place on Monday.
Finally calling it version 1.0. Doesn't change anything formally, but
CG-3 is stable enough and widely enough used that it's about time for
the 1.0 tag.
New features:
- Cmdline flag --dump-ast to output an XML dump of the parser
representation. Useful for 3rd party transformation tools and highlighters
- Rule SPLITCOHORT that splits a cohort into multiple, while allowing
the creation of an internal dependency tree. See
https://visl.sdu.dk/cg3/chunked/rules.html#splitcohort
- Added a Bag of Tags feature. See
https://visl.sdu.dk/cg3/chunked/contexts.html#test-bag-of-tags
- Added rule flag REPEAT. See
https://visl.sdu.dk/cg3/chunked/rules.html#rule-options-repeat
- Added mathematical set difference set operator as \. See
https://visl.sdu.dk/cg3/chunked/sets.html#set-operator-difference
- Renamed the old set difference operator (-) to the except operator
- Cmdline flag --trace now optionally takes a range of rules to break
execution at
- ADDCOHORT now takes WITHCHILD to further narrow down where to insert
the cohort
- New tool cg-strictify to aid in making your existing grammars
STRICT-TAGS compliant. See
https://visl.sdu.dk/cg3/chunked/cmdreference.html#cg-strictify
Changes:
- Now requires C++11 to build, but will test for and use C++14 and
C++17 where available
- Regex captures are now done on a per-reading basis, which eliminates
weird gotchas about matches
- Linking from a position override template will now go from the
min/max edges of cohorts the template reached, rather than just the
final cohort it touched
- OR'ed templates will now backtrack instead of trying only first match
- Input stream initial SETVAR commands are now honored immediately, so
that variables can be used in DELIMITERS
- Context modifier A now works for almost all rule types
- KEEPORDER will now automatically be added for easily detectable
cases where it is needed
- SUBSTITUTE will add the replacement tags for each found contiguous
tag to be removed
- Better detection of endless loops
- Non-scanning contexts with modifers O and o will now throw an error
as the writer probably meant 0
- Binary grammars can now be dumped as text, but in the internal
mangled structure
- MOVE WITHCHILD changed to more logically interact with moving into
and out of trees
- cg-conv no longer creates readings for plain text input. Cmdline
flag --add-tags added to get these back where desired
- Numeric tags are now double-precision floating point
- cg-conv no longer multiplies weights in FST input by default
Fixed Bugs:
- Fixed removing a cohort that owned temporarily enclosed cohorts
- Fixed bug where e.g. <C:NN> was considered numerical with value 0
- Fixed a lot of bugs related to cohort moving, adding, and removing
- Fixed segfault when parsing tags longer than 256 bytes
- Fixed modifier pS acting as if it was p*
Main site is https://visl.sdu.dk/cg3.html
Google Group is https://groups.google.com/group/constraint-grammar
Source snapshots available at https://visl.sdu.dk/download/cg3/
Windows binary is at https://apertium.projectjj.com/win32/nightly/
OS X binary is at https://apertium.projectjj.com/osx/nightly/
RHEL/Fedora/CentOS/OpenSUSE packages are at
https://apertium.projectjj.com/rpm/howto.txt
Debian/Ubuntu packages are at https://apertium.projectjj.com/apt/howto.txt
-- Tino Didriksen
CG-3 Developer
--
You received this message because you are subscribed to the Google
Groups "Constraint Grammar" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected]
<mailto:[email protected]>.
To post to this group, send email to
[email protected]
<mailto:[email protected]>.
Visit this group at https://groups.google.com/group/constraint-grammar.
For more options, visit https://groups.google.com/d/optout.
--
Eckhard Bick,
cand.med., dr.phil.
University of Southern Denmark
e-mail: [email protected]
web: http://beta.visl.sdu.dk
--
You received this message because you are subscribed to the Google Groups
"Constraint Grammar" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/constraint-grammar.
For more options, visit https://groups.google.com/d/optout.