X-Git-Url: https://git.xandkar.net/?p=tiger.ml.git;a=blobdiff_plain;f=README.md;h=0678fa9ca0f505c0a5dcd3311ad1e45d67c32d08;hp=463c385a44e43aabd0805dd29d3c3d028146e3c3;hb=299dd05453cc3ce02261fe3f224d9f19a28ec551;hpb=188ac95004d45d6cd6eda1a613d3537693ccba14

diff --git a/README.md b/README.md
index 463c385..0678fa9 100644
--- a/README.md
+++ b/README.md
@@ -5,17 +5,9 @@ A Tiger-compiler implementation in (OCa)ML
 Status
 ------
 
-### Technical issues
-- [-] testing framework
-  - [x] run arbitrary code snippets
-  - [x] check non-failures
-  - [x] check expected output
-  - [-] check expected exceptions
-    - [x] semant stage
-    - [ ] generalized expect `Output ('a option) | Exception of (exn -> bool)`
-  - [x] run all book test case files 
-  - [ ] grid view (cols: lex, pars, semant, etc.; rows: test cases.) 
-- [ ] Travis CI
+![screenshot-tests-head](screenshots/tests-head.jpg)
+...
+![screenshot-tests-tail](screenshots/tests-tail.jpg)
 
 ### Features
 #### Done
@@ -23,10 +15,10 @@ Status
 - [x] ch 2: Lexer
 - [x] ch 3: Parser
 - [x] ch 4: AST
+- [x] ch 5: Semantic Analysis (type checking)
 #### In-progress
-- [-] ch 5: Semantic Analysis (type checking)
-#### TODO (short-term)
 - [ ] ch 6: Activation Records
+#### TODO (short-term)
 - [ ] ch 7: Translation to Intermediate Code
 - [ ] ch 08: Basic Blocks and Traces
 - [ ] ch 09: Instruction Selection
@@ -45,6 +37,22 @@ Status
 #### Maybe
 - [ ] ch 14: Object-Oriented Languages
 
+### Technical issues
+- [-] testing framework
+  - [x] run arbitrary code snippets
+  - [x] check non-failures
+  - [x] check expected output
+  - [-] check expected exceptions
+    - [x] semant stage
+    - [ ] generalized expect `Output ('a option) | Exception of (exn -> bool)`
+  - [x] run all book test case files 
+  - [-] grid view (cols: lex, pars, semant, etc.; rows: test cases.) 
+    - [x] implementation
+    - [ ] refactoring
+  - [ ] test time-outs (motive: cycle non-detection caused an infinite loop)
+    - [ ] parallel test execution
+- [ ] Travis CI
+
 Implementation Notes
 --------------------
 
@@ -52,11 +60,88 @@ Implementation Notes
 
 #### shift/reduce conflicts
 ##### grouping consecutive declarations
+In order to support mutual recursion, we need to group consecutive
+type and function declarations (see Tiger-book pages 97-99).
+
+Initially, I defined the rules to do so as:
+
+    decs:
+      | dec      { $1 :: [] }
+      | dec decs { $1 :: $2 }
+      ;
+    dec:
+      | var_dec  { $1 }
+      | typ_decs { Ast.TypeDecs $1 }
+      | fun_decs { Ast.FunDecs $1 }
+      ;
+
+which, while straightforward (and working, because `ocamlyacc` defaults to
+shift in case of a conflict), nonetheless caused a shift/reduce conflict in
+each of: `typ_decs` and `fun_decs`; where the parser did not know whether to
+shift and stay in `(typ|fun_)_dec` state or to reduce and get back to `dec`
+state.
+
+Sadly, tagging the rules with a lower precedence (to explicitly favor
+shifting) - does not help :(
+
+    %nonassoc LOWEST
+    ...
+    dec:
+      | var_dec                { $1 }
+      | typ_decs  %prec LOWEST { Ast.TypeDecs $1 }
+      | fun_decs  %prec LOWEST { Ast.FunDecs $1 }
+      ;
+
+The difficulty seems to be in the lack of a separator token which would be
+able to definitively mark the end of each sequence of consecutive
+`(typ_|fun_)` declarations.
+
+Keeping this in mind, another alternative is to manually capture the possible
+interspersion patterns in the rules like:
+
+    (N * foo) followed-by (N * not-foo)
+
+for the exception of `var_dec`, which, since we do not need to group its
+consecutive sequences, can be reduced upon first sighting.
+
 ##### lval
 
 ### AST
 
 #### print as M-exp
 
+I chose to pretty-print AST as an (indented)
+[M-expression](https://en.wikipedia.org/wiki/M-expression) - an underrated
+format, used in Mathematica and was intended for Lisp by McCarthy himself; it
+is nearly as flexible as S-expressions, but significantly more readable (IMO).
+
+As an example, here is what `test28.tig` looks like after parsing and
+pretty-printing:
+
+    LetExp[
+        [
+        TypeDecs[
+            TypeDec[
+                arrtype1,
+                ArrayTy[
+                int]],
+            TypeDec[
+                arrtype2,
+                ArrayTy[
+                int]]],
+        VarDec[
+            arr1,
+            arrtype1,
+            ArrayExp[
+                arrtype2,
+                IntExp[
+                    10],
+                IntExp[
+                    0]]]],
+        SeqExp[
+            VarExp[
+                SimpleVar[
+                    arr1]]]]
+
 ### Machine
 Will most-likely compile to RISC and execute using SPIM (as favored by Appel)