Alexandre Terrasa [Wed, 20 Apr 2016 06:37:09 +0000 (02:37 -0400)]
contrib: fix call on nullable receiver warnings
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Wed, 20 Apr 2016 06:36:31 +0000 (02:36 -0400)]
contrib: fix useless signature in redef warning
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Wed, 20 Apr 2016 22:40:33 +0000 (18:40 -0400)]
contrib/inkscape_tools: fix weird parenthesis
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Jean Privat [Wed, 20 Apr 2016 17:22:02 +0000 (13:22 -0400)]
Merge: nitcorn: fix path resolution for vararg routes
Routes with parameters like `/blog/articles/:articleId` where not matched correctly.
It worked until today and I'm not really sure how? Anyway, this PR resolve the route with request parameters before performing the match.
Also done some cleaning in warnings and doc.
Pull-Request: #2023
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Wed, 20 Apr 2016 17:22:01 +0000 (13:22 -0400)]
Merge: lib/core: fix `Text::is_int`
Fix #2019, `is_int` was always returning true. It was caused by the use of the wrong local variable.
Ping @R4PaSs and @Freddrickk.
Pull-Request: #2022
Reviewed-by: Frédéric Vachon <fredvac@gmail.com>
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Reviewed-by: Alexandre Terrasa <alexandre@moz-code.org>
Jean Privat [Wed, 20 Apr 2016 17:21:58 +0000 (13:21 -0400)]
Merge: iOS and text: fix UTF-8 support on iOS and improve doc on FFI related text services
This PR fix the support for UTF-8 on iOS by simply updating the encoding and use the correct number of bytes. This does not completely resolve #1945 as the error messages are still present and I suspect that the UTF-8 cleaning does not work, however I'll update the issue (or open a different one) to reflect that.
Intro the service `to_s_with_copy_and_length` which complements `to_s_with_copy` and `to_s_with_length` . The name is a mouthful, but using optional parameters was not desirable since this service will often be used from a foreign language.
Also improves the doc of the existing services to get a Nit `String` from a `NativeString`. The new doc is more detailed and oriented towards users of the FFI by putting emphasis on where the raw data resides and wether it is cleaned or not.
Also note the removal of return type declarations in method refinements which was not consistent with the introduction.
Pull-Request: #2021
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexandre Terrasa <alexandre@moz-code.org>
Jean Privat [Wed, 20 Apr 2016 17:21:55 +0000 (13:21 -0400)]
Merge: Interpreter FFI: fix running the interpreter FFI in the interpreter
Let's start with a bit of context, the interpreter supports only the light FFI, a subset of the full FFI supported by the compiler. The interpreter uses an heuristic to guess which extern methods are supported. The problem of #2018, was in this heuristic. The heuristic states that if a parameter of an extern method is a standard Nit class and not an universal or an extern class, then that method probably uses the full FFI. However, sometimes this is false, like in the interpreter where instances of Nit standard classes are stored as pointer in the C code, still without using the full FFI.
To fix #2018, this PR intro a new annotation `light_ffi` to be used on extern methods wrongly detected as full FFI. It bypasses the heuristic and force the interpreter to compile and run the method. (or it tries to) This annotation should not be used otherwise, the heuristic is pretty good at detecting when a method is supported or not.
Pull-Request: #2020
Reviewed-by: Romain Chanoir <romain.chanoir@viacesi.fr>
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Wed, 20 Apr 2016 17:20:33 +0000 (13:20 -0400)]
lib/core/bytes: add `Int::to_bytes` and `Bytes::to_i`
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Mon, 18 Apr 2016 17:48:14 +0000 (13:48 -0400)]
tests: add test for the interpreter interpreting the interpreter with its FFI
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Wed, 20 Apr 2016 11:34:29 +0000 (07:34 -0400)]
lib/opts: report invalid integer passed to an OptionInt
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexandre Terrasa [Wed, 20 Apr 2016 06:35:06 +0000 (02:35 -0400)]
contrib/inkscape_tools: fix noinit in redef warning
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Fri, 18 Dec 2015 05:18:12 +0000 (00:18 -0500)]
uml: fix not escaped names
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Wed, 20 Apr 2016 04:54:35 +0000 (00:54 -0400)]
vararg_routes: improve module documentation
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Wed, 20 Apr 2016 04:54:04 +0000 (00:54 -0400)]
nitcorn: remove warnings from reactor.nit
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexandre Terrasa [Wed, 20 Apr 2016 04:53:47 +0000 (00:53 -0400)]
nitcorn: fix path resolution for vararg routes
Signed-off-by: Alexandre Terrasa <alexandre@moz-code.org>
Alexis Laferrière [Tue, 19 Apr 2016 17:44:06 +0000 (13:44 -0400)]
lib/core: fix `Text::is_int`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 15 Apr 2016 21:33:50 +0000 (17:33 -0400)]
lib/cocoa: reading text from NSData should copy it to Nit memory
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 15 Apr 2016 21:06:24 +0000 (17:06 -0400)]
lib/cocoa: fix to_nsstring to use UTF-8 encoding
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 18 Apr 2016 19:01:32 +0000 (15:01 -0400)]
lib/core: remove duplicates and improve doc of `NativeString` related services
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 15 Apr 2016 22:01:35 +0000 (18:01 -0400)]
lib/core: intro to_s_with_copy_and_length
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 18 Apr 2016 17:23:36 +0000 (13:23 -0400)]
interpreter: use light_ffi to force support for the `instance` extern methods
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 18 Apr 2016 17:23:01 +0000 (13:23 -0400)]
interpreter: the annotation light_ffi identifies a method as light FFI only
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Mon, 18 Apr 2016 12:51:54 +0000 (08:51 -0400)]
Merge: Split catalog into a lib and a program
As requested by @Morriar, the nitcatalog tool is now splited into a lib and a program.
Pull-Request: #2017
Reviewed-by: Alexandre Terrasa <alexandre@moz-code.org>
Jean Privat [Mon, 18 Apr 2016 12:51:53 +0000 (08:51 -0400)]
Merge: iOS: support apps with multiple windows, and a better ListLayout
Support multiple windows on iOS by using a UINavigationController as the root controller which shows the navigation bar at the top of the screen. The navigation bar shows the "Back" button automatically when switching windows. It has room for a title, which can be set manually. This is not ideal as pure portable apps don't define a title leaving the bar mostly empty. This could be fixed either by always asking for a title or by using the app names a default title.
Rewrite the implementation of the ListLayout using a UIStackView in a UIScrollView. This implementation is easier to adapt to different content, and it much more simple than the previous one which used a UITableView. The table view implementation has been kept as a different iOS specific control, but it may be removed in the future.
Pull-Request: #2016
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 18 Apr 2016 12:51:51 +0000 (08:51 -0400)]
Merge: app.nit services to set the size and alignment of labels
Intro services to customize the size and alignment of text views. Only labels are supported on all platforms but Android also supports customizing text buttons. The behaviour on each platform varies a bit according to the available features. The goal is always to have the behaviour that is more natural on the platform instead of having the same exact look on each platform. Sticking to the recommended values should produce more consistent results, but using other values can create interesting effects.
For example, on Android and Linux, the text size is locked to a small/medium/large, but on iOS the size can be set to any variations relative to the default font size. On all platforms, 0.5 should result in a small font, 1.0 the default/medium size and 1.5 the large size. On iOS, other values will result in different text sizes.
I've used the same logic as `Text::justify` to set the alignment with a single float parameter.
Pull-Request: #2015
Reviewed-by: Alexandre Terrasa <alexandre@moz-code.org>
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Mon, 18 Apr 2016 00:53:14 +0000 (20:53 -0400)]
nitcatalog: add a new class Person
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Sun, 17 Apr 2016 01:49:14 +0000 (21:49 -0400)]
nitc: split nitcatalog into a lib and a program
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Sat, 2 Apr 2016 21:54:44 +0000 (17:54 -0400)]
lib/ios: implement `Label::size= & align=`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 14 Mar 2016 20:53:06 +0000 (16:53 -0400)]
lib/ios: support multiple windows using a UINavigationController
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 15 Apr 2016 15:23:57 +0000 (11:23 -0400)]
lib/ios: ListLayout use a UIStackView inside a UIScrollView
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 14 Apr 2016 18:44:33 +0000 (14:44 -0400)]
lib/ios: use simple init constructor in Objective-C when possible
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 8 Apr 2016 18:57:51 +0000 (14:57 -0400)]
lib/ios: do not crash when removing a non-View control
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 14 Mar 2016 20:52:58 +0000 (16:52 -0400)]
nitc iOS: use ???? as bundle signature (now we can)
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 2 Apr 2016 17:27:08 +0000 (13:27 -0400)]
lib/linux: implement `Label::size= & align=`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 2 Apr 2016 22:06:38 +0000 (18:06 -0400)]
lib/android: implement `TextView::size= & align=` and replace `text_size`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Wed, 25 Nov 2015 21:29:58 +0000 (16:29 -0500)]
lib/app: intro `TextView::size= & align=`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 15 Apr 2016 14:26:40 +0000 (10:26 -0400)]
lib/gtk: intro `GtkLabel::justify` and the `set_markup` service
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 2 Apr 2016 17:26:43 +0000 (13:26 -0400)]
lib/gtk: fix uncallable / unimplemented alignment and padding services
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 14 Apr 2016 01:03:06 +0000 (21:03 -0400)]
Merge: Introducing the do ... catch ... end structure
This PR is a first step in trying to handle exceptions in Nit, replacing the behaviour of `abort` if it happens within a do ... catch ... end.
In the compiler, setjmp() et longjmp() are used to jump directly from the `abort` to the nearest `catch` bloc.
Pull-Request: #2011
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 14 Apr 2016 01:03:03 +0000 (21:03 -0400)]
Merge: Benitlux & sqlite3: fix closing SQLite statements and malformed superstring
Fix statements that were left open and did not remove the lock on the database. It now relies on iterators to close the statement after the end of the loop. This requires to actually complete the loop and not return before it is complete. A shortcut is to use `statement.iterator.to_a` when the result is expected to be short, like with a single row.
Note that I did not use `with` because it has the same problem with a `return` skipping the call to `finish`. Neither did I use `Finalizable` because it is invoked by the GC which may be much later.
Also fix one malformed superstring and improve the style of another one.
Pull-Request: #2009
Reviewed-by: Jean Privat <jean@pryen.org>
BlackMinou [Sat, 9 Apr 2016 01:24:10 +0000 (21:24 -0400)]
tests: introducing some tests for the catch
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:23:09 +0000 (20:23 -0400)]
Compiler: using longjumps for the catch blocs
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:22:21 +0000 (20:22 -0400)]
Interpreter: using catch to modify abort behaviour
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:20:43 +0000 (20:20 -0400)]
Scope & Typing: visits the catch part of a do ... catch ... end
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:19:43 +0000 (20:19 -0400)]
Flow: managing the flow context for a do ... catch ... end
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:18:27 +0000 (20:18 -0400)]
Parser: Modifications due to the introduction of catch
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
BlackMinou [Sat, 9 Apr 2016 00:17:47 +0000 (20:17 -0400)]
Grammar: Adding an optionnal catch bloc to a do end
Signed-off-by: BlackMinou <romain.chanoir@viacesi.fr>
Lucas Bajolet [Mon, 11 Apr 2016 17:22:09 +0000 (13:22 -0400)]
Added services to NativeString and Byte
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Alexis Laferrière [Mon, 11 Apr 2016 16:10:25 +0000 (12:10 -0400)]
lib/sqlite3: improve doc and customization of the iterator closing statements
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Sat, 9 Apr 2016 13:58:02 +0000 (09:58 -0400)]
Merge: Unify escaping
For some historical reasons, the `return` was managed slightly differently from `break` and `continue`.
This PR unify the whole thing thus simplify how the escaping mechanism is handled by the interpreted. By the way this also fixes #2008 (that is caused by an unhandled `return` specific case)
This will also help the implementation of correct escaping and `finally` in the compiler.
Pull-Request: #2010
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Sat, 9 Apr 2016 13:57:59 +0000 (09:57 -0400)]
Merge: Tnitter app: fix list layout on GNU/Linux with GTK+
Update GNU/Linux implementation of `ListLayout` to follow the specification of `GtkListBox` which should only have children of type `GtkListBoxRow`. This was an issue in Tnitter where removing children from the `GtkListBox` was broken.
Pull-Request: #2007
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Sat, 9 Apr 2016 13:57:57 +0000 (09:57 -0400)]
Merge: More byte digests
Improve and some back-and-forth transformations between Byte and Text.
~~~nit
var b = "abcd".to_bytes
assert b.chexdigest == "\\x61\\x62\\x63\\x64"
assert b.chexdigest.unescape_to_bytes == b
assert b.binarydigest == "
01100001011000100110001101100100"
assert b.binarydigest.binarydigest_to_bytes == b
assert "1".binarydigest_to_bytes.hexdigest == "01"
assert "1 0 1".binarydigest_to_bytes.hexdigest == "05"
assert "7".hexdigest_to_bytes.hexdigest == "07"
assert "a B cd \\xe".hexdigest_to_bytes.hexdigest == "0ABCDE"
~~~
Pull-Request: #2005
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Sat, 9 Apr 2016 13:56:49 +0000 (09:56 -0400)]
core/bytes: make Text::hexdigest_to_bytes more permissive
It now accepts a odd number of digit and garbage (eg. whitespaces) in
the strings
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 8 Apr 2016 19:17:46 +0000 (15:17 -0400)]
tests: add base_with_return.res
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Fri, 8 Apr 2016 16:57:39 +0000 (12:57 -0400)]
tests: update line number in test_new_native.nit
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Fri, 8 Apr 2016 15:49:23 +0000 (11:49 -0400)]
nitc/niti: unify returnmark and escapemark
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 8 Apr 2016 15:46:47 +0000 (11:46 -0400)]
nitc/scope: attach AReturnExpr to an escape mark of the whole propdef
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Fri, 8 Apr 2016 15:45:15 +0000 (11:45 -0400)]
nitc/parser: Make `AReturnExpr` an AEscapeExpr
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Wed, 6 Apr 2016 15:40:54 +0000 (11:40 -0400)]
contrib/benitlux: fix superstring in benitlux_social
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 8 Apr 2016 15:08:41 +0000 (11:08 -0400)]
contrib/benitlux: let the iterators close the sqlite statements
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 8 Apr 2016 15:20:17 +0000 (11:20 -0400)]
lib/core: fix `Iterator::to_a` not calling `finish`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 8 Apr 2016 12:50:31 +0000 (08:50 -0400)]
lib/sqlite3: protect close statements
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 8 Apr 2016 11:47:31 +0000 (07:47 -0400)]
lib/sqlite3: fix closing a statement after iterating
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 7 Apr 2016 17:45:22 +0000 (13:45 -0400)]
contrib/tnitter: use the new `clear` for a cleaner behavior
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 7 Apr 2016 17:43:34 +0000 (13:43 -0400)]
lib/linux: fix ListBox that should contain only ListBoxRows
Reported-by: Abdelkarim Belkhir <abdelkarim.belkhir.exalger@viacesi.fr>
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 7 Apr 2016 17:43:01 +0000 (13:43 -0400)]
lib/gtk: GtkListBoxRow is a GtkBin
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Thu, 7 Apr 2016 16:37:36 +0000 (12:37 -0400)]
lib/app: intro `CompositeControl::clear`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 7 Apr 2016 13:31:07 +0000 (09:31 -0400)]
core/bytes: add `Bytes::binarydigest` and `Text::binarydigest_to_bytes`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 7 Apr 2016 13:28:46 +0000 (09:28 -0400)]
core/bytes: add `Byte::chexdigest`
Signed-off-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 7 Apr 2016 12:50:34 +0000 (08:50 -0400)]
Merge: Augmented literal strings
As a follow-up to #1991, here's the support mentioned in #1734 for bytestrings, regex and raw strings
NOTE: Depends on #1991 for integration, you may review the last 3 commits while #1991 is reviewed and eventually merged
TODO: Support degraded mode (Bytes and Byte only) for byte SuperStrings instead of refusing at compile-time
Pull-Request: #1992
Reviewed-by: Jean Privat <jean@pryen.org>
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 7 Apr 2016 12:50:29 +0000 (08:50 -0400)]
Merge: Friendz: fix missing music file
This PR moves the music file in the Chainz of Friendz project to the new location of Android specific resources (changed in #1952). This missing file crashed the app at launch on Android.
Also use a Nit constructor/factory to wrap the static Java function `MediaPlayer.create` which is a convenience method acting as a constructor and applying some of the boilerplate code.
Pull-Request: #2004
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 7 Apr 2016 12:50:24 +0000 (08:50 -0400)]
Merge: interpreter: fix rand by leaving rand methods to the FFI
This PR fix `rand` with the interpreter so setting a seed with `srand_from` produces predictable results in the following calls to `rand`, and the same results as the same compiled program. The problem was caused by the interception of `rand` which was executed in the interpreter itself, while `srand_from` was executed on the side of the interpreted program. So setting the seed did not affect in anyway the values returned by `rand`.
Pull-Request: #2003
Reviewed-by: Jean Privat <jean@pryen.org>
Jean Privat [Thu, 7 Apr 2016 12:50:21 +0000 (08:50 -0400)]
Merge: Benitlux: intro server side of social network
This PR extends the Benitlux web server with a RESTful interface for a social network. This interface will be used by the mobile client, which will be introduced in a future PR.
The new module `benitlux_social` implements all the core logic and SQL request to implement the database. It uses existing services of `benitlux_db` which only keeps track of the beers and their availability.
Most of this code is already deployed on xymus.net and used by the commuNity using the alpha version of the mobile app. Also note that this is the first real-world use of nitrestful, which handles very well all the `restful` methods of `benitlux_controller`.
Pull-Request: #2002
Reviewed-by: Jean Privat <jean@pryen.org>
Lucas Bajolet [Tue, 22 Mar 2016 16:03:18 +0000 (12:03 -0400)]
tests: Added augmented strings test
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Lucas Bajolet [Tue, 22 Mar 2016 16:01:57 +0000 (12:01 -0400)]
interpreter: Added support for prefixed/suffixed strings
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Lucas Bajolet [Tue, 22 Mar 2016 15:42:30 +0000 (11:42 -0400)]
compiler: Added prefixed and suffixed `String` support
Signed-off-by: Lucas Bajolet <r4pass@hotmail.com>
Alexis Laferrière [Tue, 5 Apr 2016 15:24:55 +0000 (11:24 -0400)]
xymus.net: update to latest benitlux server features
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Tue, 5 Apr 2016 15:24:35 +0000 (11:24 -0400)]
contrib/benitlux: rename DB to BenitluxDB waiting for working qualified names
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 4 Apr 2016 20:26:42 +0000 (16:26 -0400)]
contrib/benitlux: share a single connexion to the DB
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 4 Apr 2016 21:08:52 +0000 (17:08 -0400)]
lib/sqlite: to_sql_string escape `\`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 20:44:58 +0000 (16:44 -0400)]
lib/nitcorn: use generate_token in sessions
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 15:58:33 +0000 (11:58 -0400)]
contrib/benitlux: add push notification support
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 15:58:13 +0000 (11:58 -0400)]
contrib/benitlux: drop the non-nitrestful service `since/`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 15:54:16 +0000 (11:54 -0400)]
contrib/benitlux: use nitrestful to expose social network services to clients
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 16:02:53 +0000 (12:02 -0400)]
contrib/benitlux: intro core logic and DB services for a social network
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sat, 22 Aug 2015 21:35:34 +0000 (17:35 -0400)]
contrib/benitlux: intro shared model objects for a social network
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Tue, 5 Apr 2016 04:05:19 +0000 (00:05 -0400)]
core/bytes: document hexdigest
Signed-off-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Mon, 4 Apr 2016 17:24:51 +0000 (13:24 -0400)]
lib/android: Java's `MediaPlayer.create` is a pseudo constructor
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 4 Apr 2016 17:22:05 +0000 (13:22 -0400)]
contrib/friendz: fix missing sound
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Mon, 4 Apr 2016 15:07:30 +0000 (11:07 -0400)]
interpreter: do not intercept rand methods
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 15:50:56 +0000 (11:50 -0400)]
contrib/benitlux: use print_error to print unexpected errors
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 18:42:08 +0000 (14:42 -0400)]
contrib/benitlux: update pre-existing client of beer_events_today
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 15:50:16 +0000 (11:50 -0400)]
contrib/benitlux: fix beer_events_today returning null on error
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Fri, 1 Apr 2016 20:31:40 +0000 (16:31 -0400)]
lib/nitcorn: intro a method to generate a random token
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Sun, 25 Oct 2015 17:31:18 +0000 (13:31 -0400)]
lib/sqlite3: use less callbacks
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Thu, 31 Mar 2016 13:59:00 +0000 (09:59 -0400)]
Merge: Rewrite benchlang
This is a major rewrite of the existing but broken `bench_languages` script.
I squashed most of the whole work since a detailed bunch of commits does not make so much sense, just consider it as a new work.
what's new?
* update to lastest eiffel studio thing
* added pypy engine for python
* bench method and attribute access (not only type tests)
* add real polymorphism
* add null bench to test the bottom line
* improved html result
Pull-Request: #1995
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Jean Privat [Thu, 31 Mar 2016 13:58:57 +0000 (09:58 -0400)]
Merge: Improve `Text::justify` and use it for a pretty `PerfMap::to_s`
This PR introduces padding services ( `pad_left` and `pad_right`) to `Text` and uses it for a cleaner print in `performance_analysis`.
Seeing how the left-pad package was critical to the node.js ecosystem motivated this PR. These methods should be useful in the long term!
Pull-Request: #2000
Reviewed-by: Lucas Bajolet <r4pass@hotmail.com>
Reviewed-by: Jean Privat <jean@pryen.org>
Alexis Laferrière [Sun, 27 Mar 2016 00:56:30 +0000 (20:56 -0400)]
lib/performance_analysis: cleaner output for to_s
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Alexis Laferrière [Tue, 29 Mar 2016 14:54:55 +0000 (10:54 -0400)]
lib/core: add optional char parameter to `Text::justify`
Signed-off-by: Alexis Laferrière <alexis.laf@xymus.net>
Jean Privat [Fri, 25 Mar 2016 16:44:09 +0000 (12:44 -0400)]
Merge: niti: fix escaping in `with`
Possible escape flag was let as is in the finish code of a with statement, that caused issues when executing complex code. The correct semantic is to temporarily clear any escape flag while executing the finish.
If an uncatched espace mark is generated while executing the finish, it takes precedence over the possible old one.
Pull-Request: #1997