Made setuptools mandatory, since now I am generating a wheel and build dependencies are free

5 files changed, 136 insertions, 97 deletions

diff --git a/Makefile b/Makefile
index b284284..6a48393 100644
--- a/Makefile
+++ b/Makefile
@@ -12,4 +12,4 @@ pdf: documentation.rst
 	rst2pdf documentation.rst -o documentation.pdf
 
 upload: documentation.pdf
-	python3 setup.py register sdist upload 
+	python3 setup.py register sdist bdist_wheel upload 
diff --git a/documentation.pdf b/documentation.pdf
index 06ec760..b1e85e2 100644
--- a/documentation.pdf
+++ b/documentation.pdf
@@ -133,16 +133,16 @@ endobj
 << /Border [ 0 0 0 ] /Contents () /Dest [ 71 0 R /XYZ 62.69291 607.8236 0 ] /Rect [ 521.4627 272.7736 532.5827 284.7736 ] /Subtype /Link /Type /Annot >>

 endobj

 45 0 obj

-<< /Border [ 0 0 0 ] /Contents () /Dest [ 72 0 R /XYZ 62.69291 147.0236 0 ] /Rect [ 62.69291 254.0236 174.3929 266.0236 ] /Subtype /Link /Type /Annot >>

+<< /Border [ 0 0 0 ] /Contents () /Dest [ 72 0 R /XYZ 62.69291 129.0236 0 ] /Rect [ 62.69291 254.0236 174.3929 266.0236 ] /Subtype /Link /Type /Annot >>

 endobj

 46 0 obj

-<< /Border [ 0 0 0 ] /Contents () /Dest [ 72 0 R /XYZ 62.69291 147.0236 0 ] /Rect [ 521.4627 254.7736 532.5827 266.7736 ] /Subtype /Link /Type /Annot >>

+<< /Border [ 0 0 0 ] /Contents () /Dest [ 72 0 R /XYZ 62.69291 129.0236 0 ] /Rect [ 521.4627 254.7736 532.5827 266.7736 ] /Subtype /Link /Type /Annot >>

 endobj

 47 0 obj

-<< /Border [ 0 0 0 ] /Contents () /Dest [ 75 0 R /XYZ 62.69291 330.6236 0 ] /Rect [ 62.69291 236.0236 106.0329 248.0236 ] /Subtype /Link /Type /Annot >>

+<< /Border [ 0 0 0 ] /Contents () /Dest [ 75 0 R /XYZ 62.69291 270.6236 0 ] /Rect [ 62.69291 236.0236 106.0329 248.0236 ] /Subtype /Link /Type /Annot >>

 endobj

 48 0 obj

-<< /Border [ 0 0 0 ] /Contents () /Dest [ 75 0 R /XYZ 62.69291 330.6236 0 ] /Rect [ 521.4627 236.7736 532.5827 248.7736 ] /Subtype /Link /Type /Annot >>

+<< /Border [ 0 0 0 ] /Contents () /Dest [ 75 0 R /XYZ 62.69291 270.6236 0 ] /Rect [ 521.4627 236.7736 532.5827 248.7736 ] /Subtype /Link /Type /Annot >>

 endobj

 49 0 obj

 << /Annots [ 5 0 R 6 0 R 7 0 R 8 0 R 9 0 R 10 0 R 11 0 R 12 0 R 13 0 R 14 0 R 

@@ -255,7 +255,7 @@ endobj
 << /Outlines 79 0 R /PageLabels 119 0 R /PageMode /UseNone /Pages 99 0 R /Type /Catalog >>

 endobj

 78 0 obj

-<< /Author (Michele Simionato) /CreationDate (D:20150722085001-01'00') /Creator (\(unspecified\)) /Keywords () /Producer (ReportLab PDF Library - www.reportlab.com) /Subject (\(unspecified\)) 

+<< /Author (Michele Simionato) /CreationDate (D:20150722093236-01'00') /Creator (\(unspecified\)) /Keywords () /Producer (ReportLab PDF Library - www.reportlab.com) /Subject (\(unspecified\)) 

   /Title (The decorator module) >>

 endobj

 79 0 obj

@@ -313,10 +313,10 @@ endobj
 << /Dest [ 71 0 R /XYZ 62.69291 607.8236 0 ] /Next 97 0 R /Parent 79 0 R /Prev 95 0 R /Title (Generic functions and virtual ancestors) >>

 endobj

 97 0 obj

-<< /Dest [ 72 0 R /XYZ 62.69291 147.0236 0 ] /Next 98 0 R /Parent 79 0 R /Prev 96 0 R /Title (Caveats and limitations) >>

+<< /Dest [ 72 0 R /XYZ 62.69291 129.0236 0 ] /Next 98 0 R /Parent 79 0 R /Prev 96 0 R /Title (Caveats and limitations) >>

 endobj

 98 0 obj

-<< /Dest [ 75 0 R /XYZ 62.69291 330.6236 0 ] /Parent 79 0 R /Prev 97 0 R /Title (LICENSE) >>

+<< /Dest [ 75 0 R /XYZ 62.69291 270.6236 0 ] /Parent 79 0 R /Prev 97 0 R /Title (LICENSE) >>

 endobj

 99 0 obj

 << /Count 19 /Kids [ 49 0 R 51 0 R 56 0 R 57 0 R 58 0 R 60 0 R 61 0 R 62 0 R 63 0 R 64 0 R 

@@ -5893,14 +5893,14 @@ Q
 endstream

 endobj

 113 0 obj

-<< /Length 16060 >>

+<< /Length 16068 >>

 stream

 1 0 0 1 0 0 cm  BT /F1 12 Tf 14.4 TL ET

 q

 1 0 0 1 62.69291 717.0236 cm

 q

 0 0 0 rg

-BT 1 0 0 1 0 38 Tm /F1 10 Tf 12 TL .352209 Tw (I could give a down-to-earth example of situations in which it is desiderable to dispatch on more than one) Tj T* 0 Tw 3.67998 Tw (argument \(for instance once I implemented a database-access library where the first dispatching) Tj T* 0 Tw .164987 Tw (argument was the the database driver and the second the database record\), but here I prefer to follow the) Tj T* 0 Tw (tradition and show the time-honored Rock-Paper-Scissor example:) Tj T* ET

+BT 1 0 0 1 0 38 Tm /F1 10 Tf 12 TL .352209 Tw (I could give a down-to-earth example of situations in which it is desiderable to dispatch on more than one) Tj T* 0 Tw 3.67998 Tw (argument \(for instance once I implemented a database-access library where the first dispatching) Tj T* 0 Tw .442765 Tw (argument was the the database driver and the second one was the database record\), but here I prefer to) Tj T* 0 Tw (follow the tradition and show the time-honored Rock-Paper-Scissor example:) Tj T* ET

 Q

 Q

 q

@@ -6817,14 +6817,14 @@ Q
 q

 1 0 0 1 62.69291 86.82362 cm

 q

-BT 1 0 0 1 0 38 Tm .646651 Tw 12 TL /F1 10 Tf 0 0 0 rg (The implementation of generic functions in the decorator module is still experimental. In this initial phase) Tj T* 0 Tw 2.659069 Tw (implicity was preferred over consistency with the way ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (works in the) Tj T* 0 Tw .571235 Tw (standard library. So there some subtle differences in special case. I will only show an example. Suppose) Tj T* 0 Tw (with are using a third party set-like class like the following:) Tj T* ET

+BT 1 0 0 1 0 38 Tm .646651 Tw 12 TL /F1 10 Tf 0 0 0 rg (The implementation of generic functions in the decorator module is still experimental. In this initial phase) Tj T* 0 Tw 2.659069 Tw (implicity was preferred over consistency with the way ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (works in the) Tj T* 0 Tw .258735 Tw (standard library. So there some subtle differences in special cases. I will only show an example. Suppose) Tj T* 0 Tw (you are using a third party set-like class like the following:) Tj T* ET

 Q

 Q

  

 endstream

 endobj

 115 0 obj

-<< /Length 14323 >>

+<< /Length 14773 >>

 stream

 1 0 0 1 0 0 cm  BT /F1 12 Tf 14.4 TL ET

 q

@@ -6873,7 +6873,9 @@ n 120 12 12 12 re f*
 n 48 0 36 12 re f*

 .960784 .960784 .862745 rg

 n 90 0 6 12 re f*

-BT 1 0 0 1 0 50 Tm 12 TL /F5 10 Tf 0 .501961 0 rg (class) Tj /F3 10 Tf 0 0 0 rg ( ) Tj /F5 10 Tf 0 0 1 rg (SomeSet) Tj /F3 10 Tf 0 0 0 rg (\() Tj 0 0 0 rg (collections) Tj .4 .4 .4 rg (.) Tj 0 0 0 rg (Sized) Tj 0 0 0 rg (\):) Tj 0 0 0 rg  T* (    ) Tj /F6 10 Tf .25098 .501961 .501961 rg (# methods that make SomeSet set-like) Tj /F3 10 Tf 0 0 0 rg  T* (    ) Tj /F6 10 Tf .25098 .501961 .501961 rg (# not shown ...) Tj /F3 10 Tf 0 0 0 rg  T* (    ) Tj /F5 10 Tf 0 .501961 0 rg (def) Tj /F3 10 Tf 0 0 0 rg ( ) Tj 0 0 1 rg (__len__) Tj 0 0 0 rg (\() Tj 0 .501961 0 rg (self) Tj 0 0 0 rg (\):) Tj 0 0 0 rg  T* (        ) Tj /F5 10 Tf 0 .501961 0 rg (return) Tj /F3 10 Tf 0 0 0 rg ( ) Tj .4 .4 .4 rg (0) Tj T* ET

+.960784 .960784 .862745 rg

+n 108 0 264 12 re f*

+BT 1 0 0 1 0 50 Tm 12 TL /F5 10 Tf 0 .501961 0 rg (class) Tj /F3 10 Tf 0 0 0 rg ( ) Tj /F5 10 Tf 0 0 1 rg (SomeSet) Tj /F3 10 Tf 0 0 0 rg (\() Tj 0 0 0 rg (collections) Tj .4 .4 .4 rg (.) Tj 0 0 0 rg (Sized) Tj 0 0 0 rg (\):) Tj 0 0 0 rg  T* (    ) Tj /F6 10 Tf .25098 .501961 .501961 rg (# methods that make SomeSet set-like) Tj /F3 10 Tf 0 0 0 rg  T* (    ) Tj /F6 10 Tf .25098 .501961 .501961 rg (# not shown ...) Tj /F3 10 Tf 0 0 0 rg  T* (    ) Tj /F5 10 Tf 0 .501961 0 rg (def) Tj /F3 10 Tf 0 0 0 rg ( ) Tj 0 0 1 rg (__len__) Tj 0 0 0 rg (\() Tj 0 .501961 0 rg (self) Tj 0 0 0 rg (\):) Tj 0 0 0 rg  T* (        ) Tj /F5 10 Tf 0 .501961 0 rg (return) Tj /F3 10 Tf 0 0 0 rg ( ) Tj .4 .4 .4 rg (0) Tj 0 0 0 rg (  ) Tj /F6 10 Tf .25098 .501961 .501961 rg (# in reality one would return more than zero) Tj T* ET

 Q

 Q

 Q

@@ -7044,13 +7046,19 @@ Q
 Q

 Q

 q

-1 0 0 1 62.69291 362.2236 cm

+1 0 0 1 62.69291 386.2236 cm

+q

+BT 1 0 0 1 0 38 Tm 2.257752 Tw 12 TL /F1 10 Tf 0 0 0 rg (Generic functions implemented via ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (use a more sophisticated lookup) Tj T* 0 Tw .632209 Tw (algorithm; in particular they are able to discern that a ) Tj /F3 10 Tf 0 0 0 rg (Set ) Tj /F1 10 Tf 0 0 0 rg (is a ) Tj /F3 10 Tf 0 0 0 rg (Sized ) Tj /F1 10 Tf 0 0 0 rg (object, so the implementation for) Tj T* 0 Tw 2.135777 Tw /F3 10 Tf 0 0 0 rg (Set ) Tj /F1 10 Tf 0 0 0 rg (is taken and the result is 1, not 0. Still, the implementation in the decorator module is easy to) Tj T* 0 Tw (undestand, once one declare that real ancestors take the precedence over virtual ancestors.) Tj T* ET

+Q

+Q

+q

+1 0 0 1 62.69291 344.2236 cm

 q

-BT 1 0 0 1 0 62 Tm 2.257752 Tw 12 TL /F1 10 Tf 0 0 0 rg (Generic functions implemented via ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (use a more sophisticated lookup) Tj T* 0 Tw .632209 Tw (algorithm; in particular they are able to discern that a ) Tj /F3 10 Tf 0 0 0 rg (Set ) Tj /F1 10 Tf 0 0 0 rg (is a ) Tj /F3 10 Tf 0 0 0 rg (Sized ) Tj /F1 10 Tf 0 0 0 rg (object, so the implementation for) Tj T* 0 Tw 10.16355 Tw /F3 10 Tf 0 0 0 rg (Set ) Tj /F1 10 Tf 0 0 0 rg (is taken and the result is 1, not 0. Also, the functions implemented via) Tj T* 0 Tw .927251 Tw /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (are smarter when there are conflicting implementations and are able to) Tj T* 0 Tw .844488 Tw (solve more potential conflicts. Just to have an idea of what I am talking about, here is a situation with a) Tj T* 0 Tw (conflict:) Tj T* ET

+BT 1 0 0 1 0 26 Tm 1.365976 Tw 12 TL /F1 10 Tf 0 0 0 rg (The functions implemented via ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (are smarter when there are conflicting) Tj T* 0 Tw .691654 Tw (implementations and are able to solve more potential conflicts. Just to have an idea of what I am talking) Tj T* 0 Tw (about, here is a situation with a conflict:) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 245.0236 cm

+1 0 0 1 62.69291 227.0236 cm

 q

 q

 1 0 0 1 0 0 cm

@@ -7192,25 +7200,25 @@ Q
 Q

 Q

 q

-1 0 0 1 62.69291 189.0236 cm

+1 0 0 1 62.69291 171.0236 cm

 q

-BT 1 0 0 1 0 38 Tm 9.280976 Tw 12 TL /F1 10 Tf 0 0 0 rg (Since ) Tj /F3 10 Tf 0 0 0 rg (WithLength ) Tj /F1 10 Tf 0 0 0 rg (is both a \(virtual\) subclass of ) Tj /F3 10 Tf 0 0 0 rg (collections.Iterable ) Tj /F1 10 Tf 0 0 0 rg (and of) Tj T* 0 Tw .08998 Tw /F3 10 Tf 0 0 0 rg (collections.Sized) Tj /F1 10 Tf 0 0 0 rg (, it is impossible to decide which implementation should be taken. Consistently with) Tj T* 0 Tw .488314 Tw (the ) Tj /F4 10 Tf (refuse the temptation to guess ) Tj /F1 10 Tf (philosophy, an error is raised. ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (works) Tj T* 0 Tw (exactly the same in this case.) Tj T* ET

+BT 1 0 0 1 0 38 Tm 9.280976 Tw 12 TL /F1 10 Tf 0 0 0 rg (Since ) Tj /F3 10 Tf 0 0 0 rg (WithLength ) Tj /F1 10 Tf 0 0 0 rg (is both a \(virtual\) subclass of ) Tj /F3 10 Tf 0 0 0 rg (collections.Iterable ) Tj /F1 10 Tf 0 0 0 rg (and of) Tj T* 0 Tw 4.906647 Tw /F3 10 Tf 0 0 0 rg (collections.Sized) Tj /F1 10 Tf 0 0 0 rg (, which are not related by subclassing, it is impossible to decide which) Tj T* 0 Tw .214269 Tw (implementation should be taken. Consistently with the ) Tj /F4 10 Tf (refuse the temptation to guess ) Tj /F1 10 Tf (philosophy, an error) Tj T* 0 Tw (is raised. ) Tj /F3 10 Tf 0 0 0 rg (functools.singledispatch ) Tj /F1 10 Tf 0 0 0 rg (would work exactly the same in this case.) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 159.0236 cm

+1 0 0 1 62.69291 141.0236 cm

 q

 BT 1 0 0 1 0 14 Tm .39561 Tw 12 TL /F1 10 Tf 0 0 0 rg (Finally let me notice that the decorator module implementation does not use any cache, whereas the one) Tj T* 0 Tw (in ) Tj /F3 10 Tf 0 0 0 rg (singledispatch ) Tj /F1 10 Tf 0 0 0 rg (has a cache.) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 126.0236 cm

+1 0 0 1 62.69291 108.0236 cm

 q

 BT 1 0 0 1 0 3.5 Tm 21 TL /F2 17.5 Tf 0 0 0 rg (Caveats and limitations) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 96.02362 cm

+1 0 0 1 62.69291 78.02362 cm

 q

 0 0 0 rg

 BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL .474987 Tw (The first thing you should be aware of, it the fact that decorators have a performance penalty. The worse) Tj T* 0 Tw (case is shown by the following example:) Tj T* ET

@@ -7220,7 +7228,7 @@ Q
 endstream

 endobj

 116 0 obj

-<< /Length 10091 >>

+<< /Length 10109 >>

 stream

 1 0 0 1 0 0 cm  BT /F1 12 Tf 14.4 TL ET

 q

@@ -7281,7 +7289,7 @@ q
 1 0 0 1 62.69291 396.6236 cm

 q

 0 0 0 rg

-BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL .867984 Tw (You should be aware that decorators will make your tracebacks longer and more difficult to understand.) Tj T* 0 Tw (Consider this example:) Tj T* ET

+BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL .091984 Tw (More importantly, you should be aware that decorators will make your tracebacks longer and more difficult) Tj T* 0 Tw (to understand. Consider this example:) Tj T* ET

 Q

 Q

 q

@@ -7498,29 +7506,29 @@ Q
 endstream

 endobj

 117 0 obj

-<< /Length 12826 >>

+<< /Length 13088 >>

 stream

 1 0 0 1 0 0 cm  BT /F1 12 Tf 14.4 TL ET

 q

-1 0 0 1 62.69291 717.0236 cm

+1 0 0 1 62.69291 669.0236 cm

 q

-BT 1 0 0 1 0 38 Tm .74186 Tw 12 TL /F1 10 Tf 0 0 0 rg (could happen in future versions of Python \(see PEP ) Tj 0 0 .501961 rg (362) Tj 0 0 0 rg (\) and then the decorator module would become) Tj T* 0 Tw 2.385318 Tw (obsolete. However, at present, even in Python 3.5 it is impossible to change the function signature) Tj T* 0 Tw 1.372485 Tw (directly, therefore the ) Tj /F3 10 Tf 0 0 0 rg (decorator ) Tj /F1 10 Tf 0 0 0 rg (module is still useful. Actually, this is the main reasons why I keep) Tj T* 0 Tw (maintaining the module and releasing new versions.) Tj T* ET

+BT 1 0 0 1 0 86 Tm .74186 Tw 12 TL /F1 10 Tf 0 0 0 rg (could happen in future versions of Python \(see PEP ) Tj 0 0 .501961 rg (362) Tj 0 0 0 rg (\) and then the decorator module would become) Tj T* 0 Tw 2.385318 Tw (obsolete. However, at present, even in Python 3.4 it is impossible to change the function signature) Tj T* 0 Tw 1.372485 Tw (directly, therefore the ) Tj /F3 10 Tf 0 0 0 rg (decorator ) Tj /F1 10 Tf 0 0 0 rg (module is still useful. Actually, this is the main reasons why I keep) Tj T* 0 Tw 2.169398 Tw (maintaining the module and releasing new versions. It should be noticed that in Python 3.5 a lot of) Tj T* 0 Tw 9.189147 Tw (improvements have been made: in that version you can decorated a function with) Tj T* 0 Tw .084147 Tw /F3 10 Tf 0 0 0 rg (func_tools.update_wrapper ) Tj /F1 10 Tf 0 0 0 rg (and ) Tj /F3 10 Tf 0 0 0 rg (pydoc ) Tj /F1 10 Tf 0 0 0 rg (will see the correct signature; still internally the function will) Tj T* 0 Tw 1.47748 Tw (have an incorrect signature, as you can see by using ) Tj /F3 10 Tf 0 0 0 rg (inspect.getfullargspec) Tj /F1 10 Tf 0 0 0 rg (: all documentation) Tj T* 0 Tw (tools using such function \(which has been correctly deprecated\) will see the wrong signature.) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 675.0236 cm

+1 0 0 1 62.69291 615.0236 cm

 q

-BT 1 0 0 1 0 26 Tm 1.043828 Tw 12 TL /F1 10 Tf 0 0 0 rg (In the present implementation, decorators generated by ) Tj /F3 10 Tf 0 0 0 rg (decorator ) Tj /F1 10 Tf 0 0 0 rg (can only be used on user-defined) Tj T* 0 Tw .152485 Tw (Python functions or methods, not on generic callable objects, nor on built-in functions, due to limitations of) Tj T* 0 Tw (the ) Tj /F3 10 Tf 0 0 0 rg (inspect ) Tj /F1 10 Tf 0 0 0 rg (module in the standard library, especially for Python 2.X.) Tj T* ET

+BT 1 0 0 1 0 38 Tm 1.043828 Tw 12 TL /F1 10 Tf 0 0 0 rg (In the present implementation, decorators generated by ) Tj /F3 10 Tf 0 0 0 rg (decorator ) Tj /F1 10 Tf 0 0 0 rg (can only be used on user-defined) Tj T* 0 Tw .152485 Tw (Python functions or methods, not on generic callable objects, nor on built-in functions, due to limitations of) Tj T* 0 Tw 2.15881 Tw (the ) Tj /F3 10 Tf 0 0 0 rg (inspect ) Tj /F1 10 Tf 0 0 0 rg (module in the standard library, especially for Python 2.X \(in Python 3.5 a lot of such) Tj T* 0 Tw (limitations have been removed\).) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 645.0236 cm

+1 0 0 1 62.69291 585.0236 cm

 q

 BT 1 0 0 1 0 14 Tm .785777 Tw 12 TL /F1 10 Tf 0 0 0 rg (There is a restriction on the names of the arguments: for instance, if try to call an argument ) Tj /F3 10 Tf 0 0 0 rg (_call_ ) Tj /F1 10 Tf 0 0 0 rg (or) Tj T* 0 Tw /F3 10 Tf 0 0 0 rg (_func_ ) Tj /F1 10 Tf 0 0 0 rg (you will get a ) Tj /F3 10 Tf 0 0 0 rg (NameError) Tj /F1 10 Tf 0 0 0 rg (:) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 527.8236 cm

+1 0 0 1 62.69291 467.8236 cm

 q

 q

 1 0 0 1 0 0 cm

@@ -7622,14 +7630,14 @@ Q
 Q

 Q

 q

-1 0 0 1 62.69291 495.8236 cm

+1 0 0 1 62.69291 435.8236 cm

 q

 0 0 0 rg

 BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL 1.720651 Tw (Finally, the implementation is such that the decorated function makes a \(shallow\) copy of the original) Tj T* 0 Tw (function dictionary:) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 342.6236 cm

+1 0 0 1 62.69291 282.6236 cm

 q

 q

 1 0 0 1 0 0 cm

@@ -7769,30 +7777,30 @@ Q
 Q

 Q

 q

-1 0 0 1 62.69291 309.6236 cm

+1 0 0 1 62.69291 249.6236 cm

 q

 BT 1 0 0 1 0 3.5 Tm 21 TL /F2 17.5 Tf 0 0 0 rg (LICENSE) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 291.6236 cm

+1 0 0 1 62.69291 231.6236 cm

 q

 0 0 0 rg

 BT 1 0 0 1 0 2 Tm /F1 10 Tf 12 TL (Copyright \(c\) 2005-2015, Michele Simionato All rights reserved.) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 261.6236 cm

+1 0 0 1 62.69291 201.6236 cm

 q

 0 0 0 rg

 BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL 1.328555 Tw (Redistribution and use in source and binary forms, with or without modification, are permitted provided) Tj T* 0 Tw (that the following conditions are met:) Tj T* ET

 Q

 Q

 q

-1 0 0 1 62.69291 255.6236 cm

+1 0 0 1 62.69291 195.6236 cm

 Q

 q

-1 0 0 1 62.69291 207.6236 cm

+1 0 0 1 62.69291 147.6236 cm

 0 0 0 rg

 BT /F1 10 Tf 12 TL ET

 BT 1 0 0 1 0 2 Tm  T* ET

@@ -7807,24 +7815,31 @@ q
 Q

 Q

 q

-1 0 0 1 62.69291 207.6236 cm

+1 0 0 1 62.69291 147.6236 cm

 Q

 q

 1 0 0 1 62.69291 81.62362 cm

 q

 0 0 0 rg

-BT 1 0 0 1 0 110 Tm /F1 10 Tf 12 TL .17998 Tw (THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND) Tj T* 0 Tw 2.911797 Tw (ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED) Tj T* 0 Tw 5.165529 Tw (WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE) Tj T* 0 Tw 1.395433 Tw (DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE) Tj T* 0 Tw 5.53122 Tw (FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL) Tj T* 0 Tw 2.705976 Tw (DAMAGES \(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR) Tj T* 0 Tw 3.868976 Tw (SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION\) HOWEVER) Tj T* 0 Tw 1.326647 Tw (CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR) Tj T* 0 Tw 1.525366 Tw (TORT \(INCLUDING NEGLIGENCE OR OTHERWISE\) ARISING IN ANY WAY OUT OF THE USE OF) Tj T* 0 Tw (THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.) Tj T* ET

+BT 1 0 0 1 0 50 Tm /F1 10 Tf 12 TL .17998 Tw (THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND) Tj T* 0 Tw 2.911797 Tw (ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED) Tj T* 0 Tw 5.165529 Tw (WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE) Tj T* 0 Tw 1.395433 Tw (DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE) Tj T* 0 Tw 5.53122 Tw (FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL) Tj T* 0 Tw ET

 Q

 Q

  

 endstream

 endobj

 118 0 obj

-<< /Length 361 >>

+<< /Length 939 >>

 stream

 1 0 0 1 0 0 cm  BT /F1 12 Tf 14.4 TL ET

 q

-1 0 0 1 62.69291 741.0236 cm

+1 0 0 1 62.69291 705.0236 cm

+q

+0 0 0 rg

+BT 1 0 0 1 0 50 Tm /F1 10 Tf 12 TL 2.705976 Tw (DAMAGES \(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR) Tj T* 0 Tw 3.868976 Tw (SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION\) HOWEVER) Tj T* 0 Tw 1.326647 Tw (CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR) Tj T* 0 Tw 1.525366 Tw (TORT \(INCLUDING NEGLIGENCE OR OTHERWISE\) ARISING IN ANY WAY OUT OF THE USE OF) Tj T* 0 Tw (THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.) Tj T* ET

+Q

+Q

+q

+1 0 0 1 62.69291 675.0236 cm

 q

 0 0 0 rg

 BT 1 0 0 1 0 14 Tm /F1 10 Tf 12 TL .407132 Tw (If you use this software and you are happy with it, consider sending me a note, just to gratify my ego. On) Tj T* 0 Tw (the other hand, if you use this software and you are unhappy with it, send me a patch!) Tj T* ET

@@ -8012,36 +8027,36 @@ xref
 0000182361 00000 n

 0000198523 00000 n

 0000212236 00000 n

-0000228355 00000 n

-0000241287 00000 n

-0000255669 00000 n

-0000265819 00000 n

-0000278704 00000 n

-0000279122 00000 n

-0000279369 00000 n

-0000279407 00000 n

-0000279445 00000 n

-0000279483 00000 n

-0000279521 00000 n

-0000279559 00000 n

-0000279597 00000 n

-0000279635 00000 n

-0000279673 00000 n

-0000279711 00000 n

-0000279750 00000 n

-0000279789 00000 n

-0000279828 00000 n

-0000279867 00000 n

-0000279906 00000 n

-0000279945 00000 n

-0000279984 00000 n

-0000280023 00000 n

-0000280062 00000 n

+0000228363 00000 n

+0000241295 00000 n

+0000256127 00000 n

+0000266295 00000 n

+0000279442 00000 n

+0000280438 00000 n

+0000280685 00000 n

+0000280723 00000 n

+0000280761 00000 n

+0000280799 00000 n

+0000280837 00000 n

+0000280875 00000 n

+0000280913 00000 n

+0000280951 00000 n

+0000280989 00000 n

+0000281027 00000 n

+0000281066 00000 n

+0000281105 00000 n

+0000281144 00000 n

+0000281183 00000 n

+0000281222 00000 n

+0000281261 00000 n

+0000281300 00000 n

+0000281339 00000 n

+0000281378 00000 n

 trailer

 << /ID 

  % ReportLab generated PDF document -- digest (http://www.reportlab.com)

- [(\001/{;`\255:\224G\357U\210\302\220\020\246) (\001/{;`\255:\224G\357U\210\302\220\020\246)]

+ [(cT\276\320$\217\341<\304d\353\241\327\302A\250) (cT\276\320$\217\341<\304d\353\241\327\302A\250)]

  /Info 78 0 R /Root 77 0 R /Size 139 >>

 startxref

-280101

+281417

 %%EOF

diff --git a/documentation.rst b/documentation.rst
index 79f3b06..2c4aa7d 100644
--- a/documentation.rst
+++ b/documentation.rst
@@ -894,8 +894,8 @@ Now the XMLWriter is able to serialize floats:
 I could give a down-to-earth example of situations in which it is desiderable
 to dispatch on more than one argument (for instance once I implemented
 a database-access library where the first dispatching argument was the
-the database driver and the second the database record), but here I prefer
-to follow the tradition and show the time-honored
+the database driver and the second one was the database record),
+but here I prefer to follow the tradition and show the time-honored
 Rock-Paper-Scissor example:
 
 .. code-block:: python
@@ -1038,8 +1038,8 @@ The implementation of generic functions in the decorator module is
 still experimental. In this initial phase implicity was preferred
 over consistency with the way ``functools.singledispatch`` works in
 the standard library. So there some subtle differences in special
-case. I will only show an example.
-Suppose with are using a third party set-like class like
+cases. I will only show an example.
+Suppose you are using a third party set-like class like
 the following:
 
 .. code-block:: python
@@ -1048,7 +1048,7 @@ the following:
      # methods that make SomeSet set-like
      # not shown ...
      def __len__(self):
-         return 0
+         return 0  # in reality one would return more than zero
 
 
 Here the author of ``SomeSet`` made a mistake by not inheriting
@@ -1084,7 +1084,11 @@ Generic functions implemented via ``functools.singledispatch`` use
 a more sophisticated lookup algorithm; in particular they are able
 to discern that a ``Set`` is a ``Sized`` object, so the
 implementation for ``Set`` is taken and the result is 1, not 0.
-Also, the functions implemented via ``functools.singledispatch``
+Still, the implementation in the decorator module is easy to
+undestand, once one declare that real ancestors take the precedence
+over virtual ancestors.
+
+The functions implemented via ``functools.singledispatch``
 are smarter when there are conflicting implementations and are
 able to solve more potential conflicts. Just to have an idea
 of what I am talking about, here is a situation with a conflict:
@@ -1101,10 +1105,11 @@ of what I am talking about, here is a situation with a conflict:
  RuntimeError: Ambiguous dispatch for WithLength instance: Sized or Iterable?
 
 Since ``WithLength`` is both a (virtual) subclass
-of ``collections.Iterable`` and of ``collections.Sized``, it is impossible
+of ``collections.Iterable`` and of ``collections.Sized``, which are
+not related by subclassing, it is impossible
 to decide which implementation should be taken. Consistently with
 the *refuse the temptation to guess* philosophy, an error is raised.
-``functools.singledispatch`` works exactly the same in this case.
+``functools.singledispatch`` would work exactly the same in this case.
 
 Finally let me notice that the decorator module implementation does
 not use any cache, whereas the one in ``singledispatch`` has a cache.
@@ -1147,8 +1152,9 @@ performance penalty could be completely negligible.  As always, the
 only way to know if there is
 a penalty in your specific use case is to measure it.
 
-You should be aware that decorators will make your tracebacks
-longer and more difficult to understand. Consider this example:
+More importantly, you should be aware that decorators will make your
+tracebacks longer and more difficult to understand. Consider this
+example:
 
 .. code-block:: python
 
@@ -1184,17 +1190,25 @@ would require to change the CPython implementation of functions and
 add an hook to make it possible to change their signature directly.
 That could happen in future versions of Python (see PEP 362_) and
 then the decorator module would become obsolete. However, at present,
-even in Python 3.5 it is impossible to change the function signature
+even in Python 3.4 it is impossible to change the function signature
 directly, therefore the ``decorator`` module is still useful.
 Actually, this is the main reasons why I keep maintaining
 the module and releasing new versions.
+It should be noticed that in Python 3.5 a lot of improvements
+have been made: in that version you can decorated a function
+with ``func_tools.update_wrapper`` and ``pydoc`` will see the correct
+signature; still internally the function will have an incorrect
+signature, as you can see by using ``inspect.getfullargspec``:
+all documentation tools using such function (which has been
+correctly deprecated) will see the wrong signature.
 
 .. _362: http://www.python.org/dev/peps/pep-0362
 
 In the present implementation, decorators generated by ``decorator``
 can only be used on user-defined Python functions or methods, not on generic
 callable objects, nor on built-in functions, due to limitations of the
-``inspect`` module in the standard library, especially for Python 2.X.
+``inspect`` module in the standard library, especially for Python 2.X
+(in Python 3.5 a lot of such limitations have been removed).
 
 There is a restriction on the names of the arguments: for instance,
 if try to call an argument ``_call_`` or ``_func_``
diff --git a/setup.py b/setup.py
index eedae53..98ff0f7 100644
--- a/setup.py
+++ b/setup.py
@@ -1,8 +1,4 @@
-import sys
-if sys.version >= '3':
-    from setuptools import setup
-else:
-    from distutils.core import setup
+from setuptools import setup
 
 dic = {}
 exec(open('src/decorator.py').read(), dic)
diff --git a/src/tests/documentation.py b/src/tests/documentation.py
index 1952d6f..e4d99b6 100644
--- a/src/tests/documentation.py
+++ b/src/tests/documentation.py
@@ -680,8 +680,8 @@ Now the XMLWriter is able to serialize floats:
 I could give a down-to-earth example of situations in which it is desiderable
 to dispatch on more than one argument (for instance once I implemented
 a database-access library where the first dispatching argument was the
-the database driver and the second the database record), but here I prefer
-to follow the tradition and show the time-honored
+the database driver and the second one was the database record),
+but here I prefer to follow the tradition and show the time-honored
 Rock-Paper-Scissor example:
 
 $$Rock
@@ -773,8 +773,8 @@ The implementation of generic functions in the decorator module is
 still experimental. In this initial phase implicity was preferred
 over consistency with the way ``functools.singledispatch`` works in
 the standard library. So there some subtle differences in special
-case. I will only show an example.
-Suppose with are using a third party set-like class like
+cases. I will only show an example.
+Suppose you are using a third party set-like class like
 the following:
 
 $$SomeSet
@@ -812,7 +812,11 @@ Generic functions implemented via ``functools.singledispatch`` use
 a more sophisticated lookup algorithm; in particular they are able
 to discern that a ``Set`` is a ``Sized`` object, so the
 implementation for ``Set`` is taken and the result is 1, not 0.
-Also, the functions implemented via ``functools.singledispatch``
+Still, the implementation in the decorator module is easy to
+undestand, once one declare that real ancestors take the precedence
+over virtual ancestors.
+
+The functions implemented via ``functools.singledispatch``
 are smarter when there are conflicting implementations and are
 able to solve more potential conflicts. Just to have an idea
 of what I am talking about, here is a situation with a conflict:
@@ -829,10 +833,11 @@ of what I am talking about, here is a situation with a conflict:
  RuntimeError: Ambiguous dispatch for WithLength instance: Sized or Iterable?
 
 Since ``WithLength`` is both a (virtual) subclass
-of ``collections.Iterable`` and of ``collections.Sized``, it is impossible
+of ``collections.Iterable`` and of ``collections.Sized``, which are
+not related by subclassing, it is impossible
 to decide which implementation should be taken. Consistently with
 the *refuse the temptation to guess* philosophy, an error is raised.
-``functools.singledispatch`` works exactly the same in this case.
+``functools.singledispatch`` would work exactly the same in this case.
 
 Finally let me notice that the decorator module implementation does
 not use any cache, whereas the one in ``singledispatch`` has a cache.
@@ -875,8 +880,9 @@ performance penalty could be completely negligible.  As always, the
 only way to know if there is
 a penalty in your specific use case is to measure it.
 
-You should be aware that decorators will make your tracebacks
-longer and more difficult to understand. Consider this example:
+More importantly, you should be aware that decorators will make your
+tracebacks longer and more difficult to understand. Consider this
+example:
 
 .. code-block:: python
 
@@ -912,17 +918,25 @@ would require to change the CPython implementation of functions and
 add an hook to make it possible to change their signature directly.
 That could happen in future versions of Python (see PEP 362_) and
 then the decorator module would become obsolete. However, at present,
-even in Python 3.5 it is impossible to change the function signature
+even in Python 3.4 it is impossible to change the function signature
 directly, therefore the ``decorator`` module is still useful.
 Actually, this is the main reasons why I keep maintaining
 the module and releasing new versions.
+It should be noticed that in Python 3.5 a lot of improvements
+have been made: in that version you can decorated a function
+with ``func_tools.update_wrapper`` and ``pydoc`` will see the correct
+signature; still internally the function will have an incorrect
+signature, as you can see by using ``inspect.getfullargspec``:
+all documentation tools using such function (which has been
+correctly deprecated) will see the wrong signature.
 
 .. _362: http://www.python.org/dev/peps/pep-0362
 
 In the present implementation, decorators generated by ``decorator``
 can only be used on user-defined Python functions or methods, not on generic
 callable objects, nor on built-in functions, due to limitations of the
-``inspect`` module in the standard library, especially for Python 2.X.
+``inspect`` module in the standard library, especially for Python 2.X
+(in Python 3.5 a lot of such limitations have been removed).
 
 There is a restriction on the names of the arguments: for instance,
 if try to call an argument ``_call_`` or ``_func_``
@@ -1414,7 +1428,7 @@ class SomeSet(collections.Sized):
     # methods that make SomeSet set-like
     # not shown ...
     def __len__(self):
-        return 0
+        return 0  # in reality one would return more than zero
 
 
 @dispatch_on('obj')