Jean Privat [Fri, 25 Apr 2014 02:25:47 +0000 (22:25 -0400)]
nitls: parse options before instantiating modelbuilder
Because things, eg paths, may depend on options,
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 22 Apr 2014 20:30:59 +0000 (16:30 -0400)]
tests: update nitls
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 25 Apr 2014 01:36:56 +0000 (21:36 -0400)]
nitls: use `-P` for `--project`. `-p` is already used for `--path`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 22 Apr 2014 20:04:24 +0000 (16:04 -0400)]
nitls: option `-M` aliases `-d`, `-p` and `-s`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 22 Apr 2014 20:03:45 +0000 (16:03 -0400)]
nitls: can use `--depends` with others options.
~~~
$ nitls ../examples/hello_world.nit ../examples/calculator.nit -d
calculator (../examples/calculator.nit)
gtk (../lib/gtk.nit)
gtk3_4 (../lib/gtk3_4)
hello_world (../examples/hello_world.nit)
standard (../lib/standard)
~~~
~~~
$ nitls ../examples/hello_world.nit ../examples/calculator.nit -d -t -p
../examples/calculator.nit
`--../examples/calculator.nit
../lib/gtk.nit
`--../lib/gtk.nit
../lib/gtk3_4
|--../lib/gtk3_4/gdk_enums.nit
|--../lib/gtk3_4/gtk3_4.nit
|--../lib/gtk3_4/gtk_assistant.nit
|--../lib/gtk3_4/gtk_core.nit
|--../lib/gtk3_4/gtk_dialogs.nit
|--../lib/gtk3_4/gtk_enums.nit
`--../lib/gtk3_4/gtk_widgets_ext.nit
../examples/hello_world.nit
`--../examples/hello_world.nit
../lib/standard
|--../lib/standard/environ.nit
|--../lib/standard/exec.nit
|--../lib/standard/file.nit
|--../lib/standard/gc.nit
|--../lib/standard/kernel.nit
|--../lib/standard/math.nit
|--../lib/standard/posix.nit
|--../lib/standard/standard.nit
|--../lib/standard/stream.nit
|--../lib/standard/string.nit
|--../lib/standard/string_search.nit
|--../lib/standard/time.nit
`--../lib/standard/collection
|--../lib/standard/collection/abstract_collection.nit
|--../lib/standard/collection/array.nit
|--../lib/standard/collection/collection.nit
|--../lib/standard/collection/hash_collection.nit
|--../lib/standard/collection/list.nit
|--../lib/standard/collection/range.nit
`--../lib/standard/collection/sorter.nit
~~~
Jean Privat [Tue, 22 Apr 2014 19:53:02 +0000 (15:53 -0400)]
nitls: add option --path to force the output of paths only
Otherwise, a couple `name (filepath)` is displayed for each entry
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 22 Apr 2014 13:35:53 +0000 (09:35 -0400)]
Merge: Inkscape tools
Intro 2 tools based on Inkscape.
__svg_to_png_and_nit__ reads a SVG file and outputs 2 files:
* A single png file
* A Nit source file and class, with attributes pointing to all objects with IDs beginning with 0.
__svg_to_icons__ Generates PNG files from a SVG file. It can output the format expected by Android projects.
Pull-Request: #406
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Alexis Laferrière [Fri, 18 Apr 2014 18:16:26 +0000 (14:16 -0400)]
contrib/inkscape_tools: update to use a nested module
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 18 Apr 2014 18:18:14 +0000 (14:18 -0400)]
Merge branch 'master' into inkscape-tool
To get import of nested modules.
Jean Privat [Fri, 18 Apr 2014 17:36:53 +0000 (13:36 -0400)]
Merge: add NativeArray::length
Add and implement native methods `NativeArray::length` and `NativeArray::to_a`
The length could be available but was not for historial reasons.
`to_a` is trivial once `length` is defined.
A future series will add `NativeArray::new` and depreciate the ugly `calloc_array`.
Note: because of a bug in nitg, the compiler in `c_src` seems to return 0 on `length` instead of aborting.
Pull-Request: #405
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Fri, 18 Apr 2014 16:48:57 +0000 (12:48 -0400)]
tests: add test_native_array.nit
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 17 Apr 2014 02:20:10 +0000 (22:20 -0400)]
niti: implements `NativeArray::length`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 17 Apr 2014 02:19:35 +0000 (22:19 -0400)]
nitg: implements `NativeArray::length`
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Thu, 17 Apr 2014 17:04:49 +0000 (13:04 -0400)]
tests: add inkscape_tools to testfull.sh
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 17 Apr 2014 12:45:24 +0000 (08:45 -0400)]
contrib/inkscape_tools: add test for icons
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 17 Apr 2014 12:45:07 +0000 (08:45 -0400)]
contrib/inkscape_tools: intro the svg_to_icons tool
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sun, 6 Apr 2014 16:30:04 +0000 (12:30 -0400)]
lib: adds support lib for svg_to_png_and_nit tool
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sun, 6 Apr 2014 03:31:26 +0000 (23:31 -0400)]
contrib: intro the svg_to_png_and_nit project
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 17 Apr 2014 02:17:06 +0000 (22:17 -0400)]
lib: add `NativeArray::length`
Engines should be able to store the allocated size.
Unlike `NativeString` that is a char*, `NativeArray` is a special Nit
opaque structure.
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Sun, 6 Apr 2014 15:58:37 +0000 (11:58 -0400)]
lib: opts add OptionFloat
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Tue, 15 Apr 2014 23:01:10 +0000 (19:01 -0400)]
Merge: Shoot for android
Enable control with the pointer
Compile to android
Scale view port
Pull-Request: #399
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Tue, 15 Apr 2014 20:02:44 +0000 (16:02 -0400)]
Merge: Modernize syntax
Lot of cleanup and modernization in the syntax of Nit
* depreciate old FFI syntax: `extern` without `class` and non-dotted extern notation (quad or space)
* reduce the usage of old style getters and setters (they will be disabled soon)
* remove the shortcut `=` for one-liner function
* hexadecimal literals `var x = 0x7F`
Conflicts: lib/jvm.nit
Pull-Request: #401
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Tue, 15 Apr 2014 17:20:26 +0000 (13:20 -0400)]
tests: remove `base_return_assing`
The syntax is removed
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:52:01 +0000 (09:52 -0400)]
src: remove old-style attribute from src/ (except parser)
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:51:24 +0000 (09:51 -0400)]
parser: regenerate parser files
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:50:49 +0000 (09:50 -0400)]
parser: add AHexIntExpr
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:49:52 +0000 (09:49 -0400)]
parser: split `AIntExpr` and `ADecIntExpr`
next commit will introduce AHexIntExpr
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:30:38 +0000 (09:30 -0400)]
parser: force new style FFI in syntax.
`extern class` + dotted extern call notations.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:29:28 +0000 (09:29 -0400)]
parser: disable unused `=` shortcut for functions
So simplify the grammar.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:28:41 +0000 (09:28 -0400)]
parser: allow top-level methods between class definitions
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:26:56 +0000 (09:26 -0400)]
parser: parentheses in `.as` are optional
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:26:01 +0000 (09:26 -0400)]
parser: add 'public' keyword
to reserve it :)
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 13:22:19 +0000 (09:22 -0400)]
tests: remove usage of old-style getters and setters
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 12:59:55 +0000 (08:59 -0400)]
lib: remove usage of old-style getters and setters
old-style attributes are still available but getters and setters are manual.
This encourage the use of new-style attributes.
Support for old style attributes will be removed eventually.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 12:57:22 +0000 (08:57 -0400)]
syntax: remove last usages of `=` in one liner function
This optional syntax is largely unused and no so KISS.
I will drop it in a future grammar.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 12:52:43 +0000 (08:52 -0400)]
ffi: update code with last syntax
use `extern class` and dotted notation for extern call and casts.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 15 Apr 2014 12:51:26 +0000 (08:51 -0400)]
tests: update sav/shoot_logic.res
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 20:13:20 +0000 (16:13 -0400)]
tests: add sav/shoot_android.res
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 20:10:55 +0000 (16:10 -0400)]
shoot: use a scale factor to compute the game dimentions
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 20:09:28 +0000 (16:09 -0400)]
shoot: add scene to sprites and use it to gather global width and height
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 20:07:19 +0000 (16:07 -0400)]
shoot: specialize scene to add witdth and height
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 20:16:16 +0000 (16:16 -0400)]
Merge: FFI with Java
Supports:
* Implementing Nit methods in Java
* Associating a Nit extern class to a Java type
* Callback Nit from Java code
To do next:
* Write the manual
* Support callbacks to super and casts (which also implies have static types in Java)
* Support Java code on Android
* Add more complexe examples, such as using an external .jar
Pull-Request: #396
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Alexis Laferrière [Mon, 14 Apr 2014 15:09:11 +0000 (11:09 -0400)]
lib/java: improve documentation
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 12 Apr 2014 18:32:50 +0000 (14:32 -0400)]
lib/java: intro JavaObject::new/release_global_ref
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 12 Apr 2014 16:09:02 +0000 (12:09 -0400)]
lib/jvm: rename JObject class to JavaObject
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 10 Apr 2014 16:47:35 +0000 (12:47 -0400)]
lib/jvm: change a C function to a Nit extern method
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 10 Apr 2014 14:45:26 +0000 (10:45 -0400)]
tests: adds a test for java strings
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 10 Apr 2014 14:46:30 +0000 (10:46 -0400)]
lib/java: add services to support Java strings
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Mon, 14 Apr 2014 14:43:05 +0000 (10:43 -0400)]
shoot: control player with `PointerEvent`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 14:42:46 +0000 (10:42 -0400)]
shoot: add Player::going_target
A simple target used to control player with pointers.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 14:26:48 +0000 (10:26 -0400)]
shoot: make android
But the game is unplayable yet without keyboard, see next patches
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 14 Apr 2014 12:57:55 +0000 (08:57 -0400)]
Merge: Correct detection of property name conflicts
Fixes #395
Pull-Request: #398
Jean Privat [Fri, 11 Apr 2014 13:07:50 +0000 (09:07 -0400)]
tests: add error_prop_name*
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 11 Apr 2014 01:56:35 +0000 (21:56 -0400)]
tests: update base_virtual_type_redef.res
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 11 Apr 2014 01:40:56 +0000 (21:40 -0400)]
modelbuilder: correctly detect property name conflicts
The tricky part is the handling of existing constructors.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 10 Apr 2014 20:41:57 +0000 (16:41 -0400)]
Merge: Qualified import
Enable `import foo::bar`
Pull-Request: #391
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 13 Mar 2014 04:10:05 +0000 (00:10 -0400)]
tests: intro tests for the FFI with Java
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 14 Mar 2014 22:00:53 +0000 (18:00 -0400)]
nitg: support compilation of Java files to .jar
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Wed, 12 Mar 2014 21:47:30 +0000 (17:47 -0400)]
nitg: intro of the FFI with java
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 10 Apr 2014 15:13:50 +0000 (11:13 -0400)]
Merge: Clean generated parser
This series reduce the size of the generated parser.
* First, factorize generated clones (the first commit)
* Second, reduce inlining in the grammar (the other commits)
This was a tidy process because each slight modification in the grammar yields to numerous conflicts often
difficult to understand. But the results worth the pain:
Time of `nitg nitg.nit`
before: 0m10.784s
after: 0m7.956s
gain: 26%
Number of subclasses of `ReduceAction`
before: 780
after: 376
gain: 52%
Pull-Request: #388
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 10 Apr 2014 12:37:35 +0000 (08:37 -0400)]
nitg-g: use nitstack by default
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 10 Apr 2014 12:36:33 +0000 (08:36 -0400)]
Merge: Gperf fix
Since gperf was slow, better replace it with a slow to call solution at
runtime but fast to generate, it's not like we'll be generating
stacktraces often in the life cycle of a program.
Surprisingly enough, even on nitg, the stacktrace does not take long
to generate, which is a good thing.
For comparison :
~~~
$ time nitg nitg.nit --stacktrace libunwind
user 0m19.720s
$ time nitg nitg.nit --stacktrace nitstack
user 0m19.820s
~~~
Pull-Request: #387
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 10 Apr 2014 12:32:30 +0000 (08:32 -0400)]
Merge: Fix nitdbg_websocket_server
Regression tests did not catch that `nitdbg_websocket_server` was broken.
Fix that
Pull-Request: #390
Jean Privat [Wed, 9 Apr 2014 21:56:34 +0000 (17:56 -0400)]
nitunit: updade usage of get_mmodule_by_name
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:36:42 +0000 (16:36 -0400)]
test: test qualified importations
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:35:54 +0000 (16:35 -0400)]
modelbuilder: use qualified name in imports to import things
import foo::bar
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:35:10 +0000 (16:35 -0400)]
modelbuilder: split `search_mmodule_by_name` from `get_mmodule_by_name`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:34:17 +0000 (16:34 -0400)]
modelbuilder: search_module_in_paths return ModulePath
Instead of a simple String
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:10:57 +0000 (16:10 -0400)]
tests: add sav/nitdbg_websocket_server.res
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 20:10:30 +0000 (16:10 -0400)]
nitdbg_websocket_server: fix write
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 19:01:15 +0000 (15:01 -0400)]
modelbuilder: get_mmodule_by_name get a mgroup as context
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 18:08:07 +0000 (14:08 -0400)]
Merge: lib/socket: Bugfix
Bug introduced in pull request #377, in the read function of FFSocket, a free is called on a static variable, which provokes a warning and probably bugs at runtime.
Pull-Request: #389
Lucas Bajolet [Wed, 9 Apr 2014 17:50:11 +0000 (13:50 -0400)]
nitg: Removed references to gperf in documentation and code
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Lucas Bajolet [Wed, 9 Apr 2014 15:50:32 +0000 (11:50 -0400)]
nitg: Overridden gperf since it was slow to generate.
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Wed, 9 Apr 2014 15:50:22 +0000 (11:50 -0400)]
Merge: Streamable and Pattern update
Upgrade on Streamable and Pattern, they will now work with any kind of Text instead of just String.
Pull-Request: #373
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Lucas Bajolet [Tue, 8 Apr 2014 19:33:10 +0000 (15:33 -0400)]
lib/socket: Bugfix, tried to free a static variable, this provoked a warning when compiling.
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Wed, 9 Apr 2014 14:47:54 +0000 (10:47 -0400)]
parser: regenerate with changes in the grammar and file generation
Before:
~~~~
$ time nitg nitg.nit
user 0m10.784s
$ wc parser/parser.nit
29702 93153 922785
$ grep 'super ReduceAction' parser/parser.nit -c
780
~~~~
After:
~~~~
$ time nitg nitg.nit
user 0m7.956s
$ wc parser/parser.nit
10660 32841 319408
$ grep 'super ReduceAction' parser/parser.nit -c
376
~~~~
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 15:03:19 +0000 (11:03 -0400)]
parser: new nullable production `typing_o`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 15:02:48 +0000 (11:02 -0400)]
parser: new production `qclassid` to factorize `qualified? classid`
See previous commit for details.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 15:01:42 +0000 (11:01 -0400)]
parser: new production `qid` to factorize `qualified? id`
`qid` is still transformed into an simple `id` (thus the qualified is lost).
Latter the qualification should be preserved in the AST.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:59:42 +0000 (10:59 -0400)]
parser: accepts `kwredef` in front of importation clauses
OK, this is crazy.
The point is to avoid the inlining of the production `redef`.
Because `redef` is used a lot, the gain worth the crazyness.
TODO: catch the usage of `redef` and crash after parsing.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:56:29 +0000 (10:56 -0400)]
parser: remove useless production `writable2`
just accepts visibility in all versions
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:54:44 +0000 (10:54 -0400)]
parser: new nullable production `qualified_o`
Somme `qualified?` remain (yet) because of conflicts, see next commits.
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:52:06 +0000 (10:52 -0400)]
parser: new nullable production `extern_code_block_o`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:51:48 +0000 (10:51 -0400)]
parser: new nullable production `annotations_o`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:39:23 +0000 (10:39 -0400)]
parser: production `annotation_withend` is nullable
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:38:04 +0000 (10:38 -0400)]
parser: production `params` is nullable
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:37:07 +0000 (10:37 -0400)]
parser: production `extern_calls` is nullable
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:33:05 +0000 (10:33 -0400)]
parser: new nullable production `string_o` to factorize `string?`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:32:15 +0000 (10:32 -0400)]
parser: new nullable production `kwend_o` to factorize `kwend?`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:31:14 +0000 (10:31 -0400)]
parser: production `formaldefs` is nullable
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 14:30:52 +0000 (10:30 -0400)]
parser: production `implicit_main_class` is nullable
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 9 Apr 2014 15:05:02 +0000 (11:05 -0400)]
parser: clean and factorize services in parser_prod
Mainly, generalize cloned services in the class ANodes
Also rationalize the generated nit code.
The diff seems big because some reindentation
Signed-off-by: Jean Privat <jean@pryen.org>
Lucas Bajolet [Mon, 7 Apr 2014 16:09:30 +0000 (12:09 -0400)]
stdlib/streams: Streamable now working on any Text type instead of just String.
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Lucas Bajolet [Tue, 8 Apr 2014 21:21:09 +0000 (17:21 -0400)]
stdlib/pattern: Rendered Pattern compatible with all Text representations instead of just String.
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Tue, 8 Apr 2014 18:37:22 +0000 (14:37 -0400)]
Merge: nitg: use libunwind instead of gperf by default for stacktraces
f54e52e activate gperf by default, but gperf is in fact quite slow.
It was unnoticed since time regressions are
tracked by valgrind (and gperf was unmonitored because run in a subprocess).
However, the effect was noticeable:
~~~
$ time nitg nitg.nit --stacktrace gperf
user 1m13.377s
$ time nitg nitg.nit --stacktrace libunwind
user 0m14.189s
~~~
Pull-Request: #386
Jean Privat [Tue, 8 Apr 2014 18:12:58 +0000 (14:12 -0400)]
nitg: use libunwind instead of gperf by default for stacktraces
f54e52e activate gperf by default, but gperf is in fact quite slow.
It was unnoticed since time regressions are
tracked by valgrind (and gperf was unmonitored because run in a subprocess).
However, the effect was noticeable:
~~~
$ time nitg nitg.nit --stacktrace gperf
user 1m13.377s
$ time nitg nitg.nit --stacktrace libunwind
user 0m14.189s
~~~
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Tue, 8 Apr 2014 15:52:08 +0000 (11:52 -0400)]
tests: update base_gen_reassign for nitg-e
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 4 Apr 2014 19:22:56 +0000 (15:22 -0400)]
Merge: Numeric, a super-class to Int and Float
Will be used by my implementation of the Bentley-Ottmann algorithm,
allowing to be used on either Int or Float. Could also be used to
revamp the API of mnit::Display and to implement QuadTree.
Pull-Request: #344
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Reviewed-by: Julien Pagès <julien.m.pages@gmail.com
Jean Privat [Mon, 7 Apr 2014 20:42:24 +0000 (16:42 -0400)]
Merge: nitg: activate nit-stacktrace by default
The C stack-trace only is unreadable for most users.
The dependence on gperf is fine since it is packaged if Debian.
the option `--stacktrace` can be used to override the default behavior.
Pull-Request: #354
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
nitlanguage / nit.git / log