From c5d88b890ba2273cb8bd9d7d16468708feb7d17e Mon Sep 17 00:00:00 2001
From: Miguel Lechon <miguel.lechon@gmail.com>
Date: Wed, 17 Feb 2021 12:05:56 +0100
Subject: [PATCH] [grammar test] Extra set delimiters and separators.

Delimiters: Square brackets and parentheses.
Separators: Commas and semicolons
---
 tests/grammar.c  | 14 +++++---------
 tests/grammar.md | 23 ++++++++++++-----------
 2 files changed, 17 insertions(+), 20 deletions(-)

diff --git a/tests/grammar.c b/tests/grammar.c
index b1105cb..f60e4b6 100644
--- a/tests/grammar.c
+++ b/tests/grammar.c
@@ -212,9 +212,10 @@ static void ExpandRule(MD_Node *rule, MD_String8List *out_strings, MD_Node *cur_
                     if(rule_element->string.size == 2 && rule_element->string.str[0] == '\\')
                     {
                         switch(rule_element->string.str[1]){
-                          case '\\': c = '\\';  break;
-                          case '\'': c = '\'';  break;
-                          case '"':  c = '\"';  break;
+                          case '\\': c = '\\'; break;
+                          case '\'': c = '\''; break;
+                          case '"':  c = '\"'; break;
+                          case 'n':  c = '\n'; break;
                         }
                     }
                     else
@@ -349,8 +350,6 @@ struct Test
     MD_Node *file_control_node;
     MD_String8List expanded_list;
 };
-
-
 int TestCompare(const void *a_, const void *b_)
 {
     int result = 1;
@@ -365,8 +364,6 @@ int TestCompare(const void *a_, const void *b_)
     return result;
 }
 
-
-
 int main(int argument_count, char **arguments)
 {
     MD_Node *grammar = MD_ParseWholeFile(MD_S8Lit("tests/grammar.md"));
@@ -418,7 +415,6 @@ int main(int argument_count, char **arguments)
     for(MD_EachNode(production, productions->first_child))
     {
         printf("%.*s: \n", MD_StringExpand(production->string));
-
         for(MD_EachNode(rule, production->first_child))
         {
             printf("    ");
@@ -502,6 +498,7 @@ int main(int argument_count, char **arguments)
             MD_Node *file_node = MD_ParseWholeString(MD_S8Lit(""), expanded);
             file_node->string = file_node->whole_string = (MD_String8){0};
 
+            // printf("> %.*s <\n", MD_StringExpand(expanded));
             if(!EqualTrees(file_node, tests[i].file_control_node))
             {
                 printf("\nFailed test %d\n", i_test);
@@ -510,7 +507,6 @@ int main(int argument_count, char **arguments)
                 MD_OutputTree(stdout, file_node);
                 printf("Grammar:\n");
                 MD_OutputTree(stdout, tests[i].file_control_node); printf("\n");
-                BP;
                 return -1;
             }
 
diff --git a/tests/grammar.md b/tests/grammar.md
index a337f1c..616ae53 100644
--- a/tests/grammar.md
+++ b/tests/grammar.md
@@ -9,14 +9,21 @@
  *   and miscellaneous semantics (@fill, @markup)
  */
 
-file            : [@child set_list]
-set_list        : { [tag_list] set [' ' @sibling set_list] }
+file                 : [@child file_set_list]
+file_set_list        : { [tag_list] set [file_set_separator @sibling file_set_list] }
+file_set_separator   : ' ' | '\n'
+
+// TODO(mal): Unify file_set_separator and set_separator by allowing ',' and ';' everywhere
+set_list        : { [tag_list] set [set_separator @sibling file_set_list] }
+set_separator   : ' ' | '\n' | ',' | ';'
                                         // TODO(mal): Accept other open/close tokens
 tag_list        : '@' @tag tag ' ' [tag_list]
 tag             : identifier [@markup '(' [@child set_list] @markup ')']
-set             : @fill leaf | @fill identifier ':' @child @fill leaf | [@fill identifier ':'] '{' [@child set_list] '}'
+set             : @fill leaf | @fill identifier ':' @child @fill leaf | [@fill identifier ':'] set_open [@child set_list] set_close
+set_open        : '{' | '[' | '('
+set_close       : '}' | ']' | ')'
 leaf            : identifier | integer_literal | char_literal | string_literal  // TODO(mal): Also symbol_label
-identifier      : alpha [alphanumeric]  // TODO(mal): I think we should allow leading underscores
+identifier      : alpha [alphanumeric] | '_' [alphanumeric]
 alphanumeric    : alpha [alphanumeric] | digit [alphanumeric] | '_' [alphanumeric]
 
 integer_literal : { ['-'] natural_literal }
@@ -60,13 +67,7 @@ symbol_colon    : ':'
 
 
 /*
-    // NOTE(mal): I think this one should work too, but MD only allows newlines as top-level construct separators
-    file            : [@child set_list]
-    set_list        : set [',' @sibling set_list]
-    set             : @fill element | '{' [@child set_list] '}'
-    element         : 'A'
-
-    // NOTE(mal): This is the simplest grammar that works
+    // NOTE(mal): This is the simplest subset of the grammar that works
     file            : [@child set_list]
     set_list        : set ['\n' @sibling set_list]
     set             : @fill element | '{' [@child set_list] '}'