merge devel

doc/endb.rst

@@ -128,7 +128,7 @@ The ``watchpoint`` pragma is syntactically a statement. It can be used
 to mark a location as a watchpoint:
 
 .. code-block:: Nim
-  var a: array [0..20, int]
+  var a: array[0..20, int]
 
   {.watchpoint: a[3].}
   for i in 0 .. 20: a[i] = i

doc/koch.rst

@@ -35,32 +35,8 @@ options:
   By default a debug version is created, passing this option will
   force a release build, which is much faster and should be preferred
   unless you are debugging the compiler.
--d:useGnuReadline
-  Includes the `rdstdin module <rdstdin.html>`_ for `interactive
-  mode <nimc.html#nim-interactive-mode>`_ (aka ``nim i``).
-  This is not needed on Windows. On other platforms this may
-  incorporate the GNU readline library.
--d:nativeStacktrace
-  Use native stack traces (only for Mac OS X or Linux).
--d:avoidTimeMachine
-  Only for Mac OS X, activating this switch will force excluding
-  the generated ``nimcache`` directories from Time Machine backups.
-  By default ``nimcache`` directories will be included in backups,
-  and just for the Nim compiler itself it means backing up 20MB
-  of generated files each time you update the compiler. Using this
-  option will make the compiler invoke the `tmutil
-  <https://developer.apple.com/library/mac/documentation/Darwin/Reference/Manpages/man8/tmutil.8.html>`_
-  command on all ``nimcache`` directories, setting their backup
-  exclusion bit.
-
-  You can use the following command to locate all ``nimcache``
-  directories and check their backup exclusion bit::
-
-    $ find . -type d -name nimcache -exec tmutil isexcluded \{\} \;
---gc:refc|v2|markAndSweep|boehm|go|none
-  Selects which garbage collection strategy to use for the compiler
-  and generated code. See the `Nim's Garbage Collector <gc.html>`_
-  documentation for more information.
+-d:useLinenoise
+  Use the linenoise library for interactive mode (not needed on Windows).
 
 After compilation is finished you will hopefully end up with the nim
 compiler in the ``bin`` directory. You can add Nim's ``bin`` directory to
@@ -104,8 +80,3 @@ be rerun in serial fashion so that meaninful error output can be gathered for
 inspection. The ``--parallelBuild:n`` switch or configuration option can be
 used to force a specific number of parallel jobs or run everything serially
 from the start (``n == 1``).
-
-zip command
------------
-
-The `zip`:idx: command builds the installation ZIP package.

doc/manual.rst

@@ -2531,7 +2531,7 @@ The implicit initialization can be avoided for optimization reasons with the
 
 .. code-block:: nim
   var
-    a {.noInit.}: array [0..1023, char]
+    a {.noInit.}: array[0..1023, char]
 
 If a proc is annotated with the ``noinit`` pragma this refers to its implicit
 ``result`` variable:
@@ -3200,7 +3200,7 @@ notation. (Thus an operator can have more than two parameters):
     # Multiply and add
     result = a * b + c
 
-  assert `*+`(3, 4, 6) == `*`(a, `+`(b, c))
+  assert `*+`(3, 4, 6) == `+`(`*`(a, b), c)
 
 
 Export marker
@@ -6302,6 +6302,9 @@ modules don't need to import a module's dependencies:
   var x: MyObject
   echo $x
 
+When the exported symbol is another module, all of its definitions will
+be forwarded. You can use an ``except`` list to exclude some of the symbols.
+
 Note on paths
 -----------
 In module related statements, if any part of the module name /
@@ -6633,7 +6636,7 @@ Syntactically it has to be used as a statement inside the loop:
       enumA, enumB, enumC, enumD, enumE
 
   proc vm() =
-    var instructions: array [0..100, MyEnum]
+    var instructions: array[0..100, MyEnum]
     instructions[2] = enumC
     instructions[3] = enumD
     instructions[4] = enumA
@@ -7802,8 +7805,9 @@ Future directions:
 Threadvar pragma
 ----------------
 
-A global variable can be marked with the ``threadvar`` pragma; it is
-a `thread-local`:idx: variable then:
+A variable can be marked with the ``threadvar`` pragma, which makes it a
+`thread-local`:idx: variable; Additionally, this implies all the effects
+of the ``global`` pragma.
 
 .. code-block:: nim
   var checkpoints* {.threadvar.}: seq[string]

doc/nep1.rst

@@ -181,7 +181,7 @@ Conventions for multi-line statements and expressions
       LongTupleA = tuple[wordyTupleMemberOne: int, wordyTupleMemberTwo: string,
                          wordyTupleMemberThree: float]
 
-- Similarly, any procedure and procedure type declarations that are longer#
+- Similarly, any procedure and procedure type declarations that are longer
   than one line should do the same thing.
 
   .. code-block:: nim
@@ -198,4 +198,4 @@ Conventions for multi-line statements and expressions
 
   .. code-block:: nim
     startProcess(nimExecutable, currentDirectory, compilerArguments
-                 environment, processOptions)
+                 environment, processOptions)

doc/tut1.rst

@@ -944,12 +944,8 @@ String variables are **mutable**, so appending to a string
 is possible, and quite efficient. Strings in Nim are both zero-terminated and have a
 length field. A string's length can be retrieved with the builtin ``len``
 procedure; the length never counts the terminating zero. Accessing the
-terminating zero is not an error and often leads to simpler code:
-
-.. code-block:: nim
-  if s[i] == 'a' and s[i+1] == 'b':
-    # no need to check whether ``i < len(s)``!
-    ...
+terminating zero is an error, it only exists so that a Nim string can be converted
+to a ``cstring`` without doing a copy.
 
 The assignment operator for strings copies the string. You can use the ``&``
 operator to concatenate strings and ``add`` to append to a string.
@@ -960,7 +956,7 @@ enforced. For example, when reading strings from binary files, they are merely
 a sequence of bytes. The index operation ``s[i]`` means the i-th *char* of
 ``s``, not the i-th *unichar*.
 
-String variables are initialized with the empty strings ``""``.
+A string variable is initialized with the empty string ``""``.
 
 
 Integers