Author: paultcochrane
Date: Sat Feb 10 16:12:58 2007
New Revision: 16939

Modified:
   trunk/languages/PIR/README

Log:
[PIR] Converted dos line endings to unix line endings


Modified: trunk/languages/PIR/README
==============================================================================
--- trunk/languages/PIR/README  (original)
+++ trunk/languages/PIR/README  Sat Feb 10 16:12:58 2007
@@ -1,88 +1,88 @@
-This is the README file for a PIR implementation in PGE.

-

-INTRODUCTION

-------------

-

-A good test-case for the compiler tool suite of Parrot is

-to compile its own assembler. This project aims to bring

-a full PIR implementation using the Parrot compiler tools.

-

-Initially, I started this project for 2 reasons:

-       

-       1. learn about PGE and other compiler tools in Parrot

-       2. allow me to add high-level constructs to write code

-                       for Parrot faster. Currently, these high-level 
constructs

-                       are not implemented.

-

-                                       

-

-STATUS

-------

-

-Currently, only work is done on the parser. Only when

-the parser is completed should work be done on the further

-transformations (using TGE in the Parrot compiler tools).

-When the parser is changed, the generated parse tree will

-probably be different as well, so any work on the later

-transformation phases should be redone, which is a pain.

-

-The following is an overview of what is parsed (mostly) correctly:

-       

-       * most PIR constructs

-       * recognizes (almost) all Parrot instructions

-               (but doesn't handle their arguments yet)

-

-See the TODO file for current issues.

-

-The grammar files are split up into several .pg files. This makes

-the edit-compile-test cycle faster.

-       

-       

-Instead of .include-ing the generated files in the main file

-(pirc.pir), the generated .pir files are compiled to bytecode

-(.pbc) and loaded in pirc.pir. This prevends compiling all the

-included files when compiling pirc.pir (which would really be

-s l o w).

-

-

-

-INSTRUCTIONS

-------------

-

-Type

-

-       $ make

-       

-Easy huh? Please note that compiling may take a while.

-The compiler is called 'pir.pbc', but the main source file

-is called pirc.pir, otherwise it would be "pir.pir", which

-is kinda strange.

-

-After that, you can run the test suite:

-

-  $ make test

-  

-which will run a little perl script that runs all files in the

-examples directory.

-

-The 'examples' directory contains a number of test files

-that are working. Running these can be done by typing:

-

-       $ ../../parrot pir.pbc examples/hello.pir 

-       

-       

-

-ACKNOWLEDGEMENTS

-----------------

-

-I stole the Makefile from Punie. The compiler driver (main file)

-is assembled from code from Punie, ABC and Pheme (although I changed

-a lot, so some things taken might be gone already).

-

-Thanks to Jerry "particle" Gay for fixing numerous platform issues.

-Thanks to Will "Coke" Coleda for helping setting up the test harness stuff.

-

-AUTHOR

-------

-

-Klaas-Jan Stol <[EMAIL PROTECTED]>

+This is the README file for a PIR implementation in PGE.
+
+INTRODUCTION
+------------
+
+A good test-case for the compiler tool suite of Parrot is
+to compile its own assembler. This project aims to bring
+a full PIR implementation using the Parrot compiler tools.
+
+Initially, I started this project for 2 reasons:
+       
+       1. learn about PGE and other compiler tools in Parrot
+       2. allow me to add high-level constructs to write code
+                       for Parrot faster. Currently, these high-level 
constructs
+                       are not implemented.
+
+                                       
+
+STATUS
+------
+
+Currently, only work is done on the parser. Only when
+the parser is completed should work be done on the further
+transformations (using TGE in the Parrot compiler tools).
+When the parser is changed, the generated parse tree will
+probably be different as well, so any work on the later
+transformation phases should be redone, which is a pain.
+
+The following is an overview of what is parsed (mostly) correctly:
+       
+       * most PIR constructs
+       * recognizes (almost) all Parrot instructions
+               (but doesn't handle their arguments yet)
+
+See the TODO file for current issues.
+
+The grammar files are split up into several .pg files. This makes
+the edit-compile-test cycle faster.
+       
+       
+Instead of .include-ing the generated files in the main file
+(pirc.pir), the generated .pir files are compiled to bytecode
+(.pbc) and loaded in pirc.pir. This prevends compiling all the
+included files when compiling pirc.pir (which would really be
+s l o w).
+
+
+
+INSTRUCTIONS
+------------
+
+Type
+
+       $ make
+       
+Easy huh? Please note that compiling may take a while.
+The compiler is called 'pir.pbc', but the main source file
+is called pirc.pir, otherwise it would be "pir.pir", which
+is kinda strange.
+
+After that, you can run the test suite:
+
+  $ make test
+  
+which will run a little perl script that runs all files in the
+examples directory.
+
+The 'examples' directory contains a number of test files
+that are working. Running these can be done by typing:
+
+       $ ../../parrot pir.pbc examples/hello.pir 
+       
+       
+
+ACKNOWLEDGEMENTS
+----------------
+
+I stole the Makefile from Punie. The compiler driver (main file)
+is assembled from code from Punie, ABC and Pheme (although I changed
+a lot, so some things taken might be gone already).
+
+Thanks to Jerry "particle" Gay for fixing numerous platform issues.
+Thanks to Will "Coke" Coleda for helping setting up the test harness stuff.
+
+AUTHOR
+------
+
+Klaas-Jan Stol <[EMAIL PROTECTED]>

Reply via email to