feat: single qoute str (#125)

Author: LytixDev

docs/syntax.md

@@ -11,7 +11,9 @@ var func return if elif else loop in true false as and or not str num bool list
 
 numbers: Can start with 0x (base16), 0b (base2) or nothing (base10) denoting the base. Then a set of number characters valid for the particular base in addition to '-' as a visual seperator. Base ten numbers can also have a decimal point.
 
-string: "" any characters enclosed inside double-qoutes
+string:
+    - "" any characters enclosed inside double-qoutes. Hanldes escape sequences (for example \n becomes newline).
+    - '' any characters enclosed inside single-qoutes Does not hanlde escape sequences.
 
 identifiers: Must start with a letter (a, b, .. z), underscore (_) or hyphen (-). Can then be any sequence of the aformentioned characters plus any number character (0, 1, .. 9).
 

src/interpreter/ast.c

@@ -234,11 +234,16 @@ Stmt *stmt_copy(Arena *arena, Stmt *to_copy)
     }
     case STMT_CMD: {
 	((CmdStmt *)copy)->cmd_name = str_view_arena_copy(arena, ((CmdStmt *)to_copy)->cmd_name);
-	((CmdStmt *)copy)->arg_exprs = arena_ll_alloc(arena);
-	LLItem *item;
-	ARENA_LL_FOR_EACH(((CmdStmt *)to_copy)->arg_exprs, item)
-	{
-	    arena_ll_append(((CmdStmt *)copy)->arg_exprs, expr_copy(arena, item->value));
+
+	if (((CmdStmt *)to_copy)->arg_exprs == NULL) {
+	    ((CmdStmt *)copy)->arg_exprs = NULL;
+	} else {
+	    ((CmdStmt *)copy)->arg_exprs = arena_ll_alloc(arena);
+	    LLItem *item;
+	    ARENA_LL_FOR_EACH(((CmdStmt *)to_copy)->arg_exprs, item)
+	    {
+		arena_ll_append(((CmdStmt *)copy)->arg_exprs, expr_copy(arena, item->value));
+	    }
 	}
 	break;
     }

src/interpreter/lexer.c

@@ -269,9 +269,9 @@ StateFn lex_any(Lexer *lexer)
 	case '\\':
 	    emit(lexer, t_backslash);
 	    break;
-	case '\'':
-	    emit(lexer, t_qoute);
-	    break;
+	// case '\'':
+	//     emit(lexer, t_qoute);
+	//     break;
 
 	/* one or two character tokens */
 	case '=':
@@ -342,6 +342,7 @@ StateFn lex_any(Lexer *lexer)
 	    return STATE_FN(lex_access);
 
 	case '"':
+	case '\'':
 	    return STATE_FN(lex_string);
 
 	case '#':
@@ -388,7 +389,7 @@ StateFn lex_shell_arg_list(Lexer *lexer)
      * Lexing rules for shell arg list:
      *  whitespace, tab           -> backup, emit, advance, ignore and continue
      *  $                         -> backup, emit, advance and lex access
-     *  "                         -> backup, emit, advance and lex string
+     *  ", '                      -> backup, emit, advance and lex string
      *  (                         -> backup, emit, advance, lex any until rparen and continue
      *  )                         -> backup, emit, advance, stop and return lex rparen
      *  \n, }, ;, |, >, <, &, EOF -> backup, emit, and stop
@@ -411,6 +412,7 @@ StateFn lex_shell_arg_list(Lexer *lexer)
 	    lex_access(lexer);
 	    break;
 	case '"':
+	case '\'':
 	    shell_arg_emit(lexer);
 	    lex_string(lexer);
 	    break;
@@ -524,30 +526,41 @@ StateFn lex_string(Lexer *lexer)
 {
     /* ignore starting qoute */
     ignore(lexer);
+
+    /* find which string type this is (' or ") */
+    char str_type = peek_ahead(lexer, -1);
     size_t str_start = lexer->pos_in_line;
     size_t str_end;
 
     StrBuilder sb;
     str_builder_init(&sb, lexer->arena);
     char c;
 continue_str_lexing:
-    while ((c = next(lexer)) != '"') {
+    while ((c = next(lexer)) != str_type) {
 	switch (c) {
 	case EOF:
 	case '\n':
 	    backup(lexer);
 	    report_lex_err(lexer, true, "Unterminated string literal");
 	    return STATE_FN(lex_any);
 	case '\\': {
-	    char next_char = next(lexer);
+	    /* For single qouted strings we do escaping */
+	    if (str_type == '\'') {
+		str_builder_append_char(&sb, '\\');
+		break;
+	    }
 
+	    char next_char = next(lexer);
 	    switch (next_char) {
 	    case '"':
 		str_builder_append_char(&sb, next_char);
 		break;
 	    case 'n':
 		str_builder_append_char(&sb, '\n');
 		break;
+	    case '\\':
+		str_builder_append_char(&sb, '\\');
+		break;
 	    default:
 		report_lex_err(lexer, true, "Unknown escape sequence");
 		/* Error occured. Continue parsing after string is terminated. */
@@ -588,9 +601,9 @@ StateFn lex_string(Lexer *lexer)
 	accept_run(lexer, " \t\v");
 	lexer->start = lexer->pos;
 
-	if (!match(lexer, '"')) {
+	if (!match(lexer, str_type)) {
 	    lexer->pos++; // Gives an underline where the double qoute should be in the error msg
-	    report_lex_err(lexer, true, "Expected double qoute to continue multiline string");
+	    report_lex_err(lexer, true, "Expected another string after '\'.");
 	    ignore(lexer);
 	    return STATE_FN(lex_any);
 	}

src/interpreter/value/slash_value.c

@@ -670,8 +670,6 @@ SlashValue str_to_str(Interpreter *interpreter, SlashValue self)
     return self;
 }
 
-
-// SlashValue range_item_get(Interpreter *interpreter, SlashValue self, SlashValue other)
 SlashValue str_item_get(Interpreter *interpreter, SlashValue self, SlashValue other)
 {
     assert(IS_STR(self));
@@ -705,6 +703,27 @@ SlashValue str_item_get(Interpreter *interpreter, SlashValue self, SlashValue ot
     return AS_VALUE(new);
 }
 
+void str_item_assign(Interpreter *interpreter, SlashValue self, SlashValue index, SlashValue other)
+{
+    assert(IS_STR(self));
+    assert(IS_STR(other));
+    assert(IS_NUM(index)); // TODO: implement for range
+    if (!NUM_IS_INT(index))
+	REPORT_RUNTIME_ERROR("Str index can not be a floating point number: '%f'", index.num);
+
+    SlashStr *str = AS_STR(self);
+    /* Ensure the index is valid */
+    int idx = (int)index.num;
+    if (idx < 0 || (size_t)idx >= str->len)
+	REPORT_RUNTIME_ERROR("Str index '%d' out of range for str with len '%zu'", idx, str->len);
+
+    SlashStr *str_other = AS_STR(other);
+    if (str_other->len != 1)
+	REPORT_RUNTIME_ERROR("Can only assign a string of length one, not length.");
+
+    str->str[idx] = str_other->str[0];
+}
+
 bool str_item_in(SlashValue self, SlashValue other)
 {
     assert(IS_STR(self) && IS_STR(other));
@@ -968,7 +987,7 @@ SlashTypeInfo str_type_info = { .name = "str",
 				.print = str_print,
 				.to_str = str_to_str,
 				.item_get = str_item_get,
-				.item_assign = NULL,
+				.item_assign = str_item_assign,
 				.item_in = str_item_in,
 				.truthy = str_truthy,
 				.eq = str_eq,

test/str.slash

@@ -4,28 +4,10 @@
     assert $a[0] == "h"
     assert $a[..5] == "hello"
     # item assign
-    # $a[0] = "H"
-    # assert $a[0] == "H"
-    # item in
-    # assert ("Hello" in $a) and ("World!" in $a)
+    $a[0] = "H"
+    assert $a[0] == "H"
 }
 
-# test every method
-# {
-#     var a = "a,b,c"
-#     var res = $a.split(",")
-#     assert $res.len() == 3
-#     assert $res[0] == "a" and $res[1] == "b" and $res[2] == "c"
-# 
-#     $a = "a!!b!!c!!"
-#     $res = $a.split("!!")
-#     assert $res.len() == 3
-#     assert $res[0] == "a" and $res[1] == "b" and $res[2] == "c"
-# }
-
-assert "hello"\
-       "world" == "helloworld"
-
 # string comparison
 assert "hello" == "hello"
 assert "b" > "a"
@@ -40,3 +22,12 @@ assert "something is true"
 assert "Hello" in "Hello World"
 assert not ("Foo" in "Bar")
 assert "" in "something"
+
+# multiline string 
+assert "hello"\
+       "world" == "helloworld"
+
+# escape sequences
+assert '\n' != "\n" # "\n" becomes newline
+assert '\n' == ("\\" + "n") # does not become newline
+assert '\' == "\\"