additional delimiters / delimiter commands

src/parser/commands_data.jl

@@ -107,7 +107,43 @@ combining_accents = [
 ]
 
 punctuation_symbols = split(raw", ; . !")
-delimiter_symbols = split(raw"| / ( ) [ ] < >")
+delimiter_symbols = split(raw"| / \ ( ) [ ] ⟨ ⟩ ‖ ⌈ ⌉ ⌊ ⌋ ⌜ ⌝ ⌞ ⌟")
+    ## NOTE `<` and `>` not included in `delimiter_symbols` because 
+    ##      they should not be used as such. In math mode, they are relation symbols and
+    ##      we don't want to change that by overwriting the `symbol_to_canonical` entries.
+    ##      **However**, `\left<` and `\right>` should work to produce `\left\langle` and 
+    ##      `\right\rangle`, respectively, due to intercepting `delimiter(...)`.
+    ## NOTE `{` and `}` not included because they are group delimiters that the tokenizer
+    ##      should recognize as `lcurly` and `rcurly`. 
+    ##      The symbols have to be typed by command.
+
+## some delimiter symbols can also be typed with commands;
+### instead of relying on `get_symbol_char` (like with `space_commands` in `commands_registration.jl`)
+### we define them explicitly, because `latex_symbols` misses some
+delimiter_commands = Dict(
+    raw"\vert" => '|',
+    raw"\slash" => '/',         # NOTE seems to work with LaTeX, but not MathJax
+    raw"\backslash" => '\\',
+    raw"\lbrack" => '[',
+    raw"\rbrack" => ']',
+    raw"\langle" => '⟨',
+    raw"\rangle" => '⟩',
+    raw"\|" => '‖',
+    raw"\Vert" => '‖',
+    raw"\lceil" => '⌈',
+    raw"\rceil" => '⌉',
+    raw"\lfloor" => '⌊',
+    raw"\lfloor" => '⌊',
+    raw"\ulcorner" => '⌜',
+    raw"\urcorner" => '⌝',
+    raw"\llcorner" => '⌞',
+    raw"\lrcorner" => '⌟',
+    raw"\{" => '{',
+    raw"\}" => '}',
+    raw"\lbrace" => '{',
+    raw"\rbrace" => '}',
+)
+
 font_names = split(raw"rm cal it tt sf bf default bb frak scr regular")
 
 # TODO Add to the parser what come below, if needed

src/parser/commands_registration.jl

@@ -58,8 +58,6 @@ const command_definitions = Dict(
     raw"\frac" => (TeXExpr(:frac), 2),
     raw"\sqrt" => (TeXExpr(:sqrt), 1),
     raw"\overline" => (TeXExpr(:overline), 1),
-    raw"\{" => (TeXExpr(:delimiter, '{'), 0),
-    raw"\}" => (TeXExpr(:delimiter, '}'), 0),
     raw"\_" => (TeXExpr(:symbol, '_'), 0),
     raw"\%" => (TeXExpr(:symbol, '%'), 0),
     raw"\$" => (TeXExpr(:symbol, '$'), 0),
@@ -150,11 +148,20 @@ for symbol in punctuation_symbols
     symbol_to_canonical[symbol] = TeXExpr(:punctuation, symbol)
 end
 
+# Delimiters
 for symbol in delimiter_symbols
     symbol = first(symbol)
     symbol_to_canonical[symbol] = TeXExpr(:delimiter, symbol)
 end
 
+for (com_str, symbol) in pairs(delimiter_commands)
+    delim_expr = TeXExpr(:delimiter, symbol)
+    if !haskey(symbol_to_canonical, symbol)
+        symbol_to_canonical[symbol] = delim_expr
+    end
+    command_definitions[com_str] = (delim_expr, 0)
+end
+
 ##
 ## Default behavior
 ##

src/parser/parser.jl

@@ -106,7 +106,13 @@ end
 function delimiter(com_str, str)
     str = str[length(com_str)+1:end]
     if length(str) == 1
-        return TeXExpr(:delimiter, only(str))
+        char = only(str)
+        if char == '<'
+            char = '⟨'
+        elseif char == '>'
+            char = '⟩'
+        end
+        return TeXExpr(:delimiter, char)
     else
         return only(texparse(str).args)
     end

src/parser/tokenizer.jl

@@ -1,13 +1,14 @@
+const token_command = re"\\[a-zA-Z]+" | re"\\."
 tex_tokens = [
     :char => re".",
     :primes => re"'+",
     :caret => re"\^",
     :underscore => re"_",
     :rcurly => re"}",
     :lcurly => re"{",
-    :command => re"\\[a-zA-Z]+" | re"\\.",
-    :right => re"\\right.",
-    :left => re"\\left.",
+    :command => token_command, 
+    :right => re"\\right." | re"\\right" * token_command,
+    :left => re"\\left." | re"\\left" * token_command,
     :newline => (re"\\" * re"\\") | re"\\n",
     :dollar => re"$"
 ]

test/parser.jl

@@ -55,6 +55,65 @@ end
 
         @test_throws TeXParseError texparse(raw"\left( x")
         @test_throws TeXParseError texparse(raw"x \right)")
+
+        ## all `delimiter_symbols`
+        test_parse(raw"|", (:delimiter, '|'))
+        test_parse(raw"/", (:delimiter, '/'))
+        test_parse(raw"\\", (:delimiter, '\\'))
+        test_parse(raw"(", (:delimiter, '('))
+        test_parse(raw")", (:delimiter, ')'))
+        test_parse(raw"[", (:delimiter, '['))
+        test_parse(raw"]", (:delimiter, ']'))
+        test_parse(raw"⟨", (:delimiter, '⟨'))
+        test_parse(raw"⟩", (:delimiter, '⟩'))
+        test_parse(raw"‖", (:delimiter, '‖'))
+        test_parse(raw"⌈", (:delimiter, '⌈'))
+        test_parse(raw"⌉", (:delimiter, '⌉'))
+        test_parse(raw"⌊", (:delimiter, '⌊'))
+        test_parse(raw"⌋", (:delimiter, '⌋'))
+        test_parse(raw"⌜", (:delimiter, '⌜'))
+        test_parse(raw"⌝", (:delimiter, '⌝'))
+        test_parse(raw"⌞", (:delimiter, '⌞'))
+        test_parse(raw"⌟", (:delimiter, '⌟'))
+
+        test_parse(raw"<", (:space, '<'))   # formerly, this was a delimiter
+        test_parse(raw">", (:space, '>'))
+
+        ## all `delimiter_commands`
+        test_parse(raw"\vert", (:delimiter, '|'))
+        test_parse(raw"\slash", (:delimiter, '/'))
+        test_parse(raw"\backslash", (:delimiter, '\\'))
+        test_parse(raw"\lbrack", (:delimiter, '['))
+        test_parse(raw"\rbrack", (:delimiter, ']'))
+        test_parse(raw"\langle", (:delimiter, '⟨'))
+        test_parse(raw"\rangle", (:delimiter, '⟩'))
+        test_parse(raw"\|", (:delimiter, '‖'))
+        test_parse(raw"\Vert", (:delimiter, '‖'))
+        test_parse(raw"\lceil", (:delimiter, '⌈'))
+        test_parse(raw"\rceil", (:delimiter, '⌉'))
+        test_parse(raw"\lfloor", (:delimiter, '⌊'))
+        test_parse(raw"\lfloor", (:delimiter, '⌊'))
+        test_parse(raw"\ulcorner", (:delimiter, '⌜'))
+        test_parse(raw"\urcorner", (:delimiter, '⌝'))
+        test_parse(raw"\llcorner", (:delimiter, '⌞'))
+        test_parse(raw"\lrcorner", (:delimiter, '⌟'))
+        test_parse(raw"\{", (:delimiter, '{'))
+        test_parse(raw"\}", (:delimiter, '}'))
+        test_parse(raw"\lbrace", (:delimiter, '{'))
+        test_parse(raw"\rbrace", (:delimiter, '}'))
+
+        ### test commands as arguments to a delimited group
+        for (cmd_str, delim_symb) in pairs(MathTeXEngine.delimiter_commands)
+            ## NOTE this does not check for "correct" left right pairs like `\lbrack` and `\rbrack`
+            test_parse(
+                "\\left$(cmd_str)\\right$(cmd_str)", 
+                (:delimited,
+                    (:delimiter, delim_symb),
+                    (:group,),
+                    (:delimiter, delim_symb)
+                )
+            )
+        end
     end
 
     @testset "Fonts" begin