nomsu/nomsu.6.peg

-- Nomsu version 6
file <-
    {:curr_indent: ' '* :}
    (((methodchain / action / expression / inline_block) eol !.)
     / file_chunks / comment? blank_lines?)
    {:curr_indent: %nil :}
    !.

shebang <- "#!" (!"nomsu" [^%nl])* "nomsu" ws+ "-V" ws* [0-9.]+ [^%nl]* (%nl / !.)

file_chunks (FileChunks) <-
    {:shebang: shebang :}?
    (top_block (nl_nodent section_division top_block)*)
    blank_lines?
    unexpected_indent? unexpected_chunk?

section_division <- ("~")^+3 eol

eof <- !.
eol <- ws* (&%nl / !.)
nodent <- (unexpected_indent [^%nl]* / =curr_indent)
indent <- { =curr_indent "    " }
blank_lines <- %nl ((nodent comment / ws*) %nl)*
nl_nodent <- blank_lines nodent
nl_indent <- blank_lines tab_error? {:curr_indent: indent :} (comment nl_nodent)*


comment (Comment) <-
    "#" {~ [^%nl]* (%nl+ (indent -> '') [^%nl]*)* (%nl &%nl)* ~}


top_block (Block) <-
    ((blank_lines nodent) / (comment nl_nodent))? statement (nl_nodent statement)*

inline_block (Block) <-
    ":" ws* (inline_statement (ws* ";" ws* inline_statement)*)?
    (&eol !nl_indent / &(ws* ([)},;] / "]")))

indented_block (Block) <-
    ":" eol nl_indent statement (nl_nodent statement)*
    (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
    {:curr_indent: %nil :}

statement_block (Block) <-
    ":" ws* (methodchain / action / expression)


statement <-
    (methodchain / action / expression / statement_block) (eol / unexpected_code)

inline_statement <-
    inline_methodchain / inline_action / inline_expression

noindex_inline_expression <-
    number / variable / inline_text / inline_list / inline_dict / inline_nomsu /
     "("
       ws* (inline_block / inline_methodchain / inline_action / inline_expression) ws*
    (")" / eof / missing_paren_err / unexpected_code)

inline_expression <- inline_index_chain / noindex_inline_expression / inline_index / inline_block

indented_expression <-
    indented_text / indented_nomsu / indented_list / indented_dict / indented_block /
    indented_parens

indented_parens <-
    "(" indented_naked_expression (nl_nodent ")" / missing_paren_err / unexpected_code)

indented_naked_expression <-
    ({| nl_indent
        (methodchain / action / expression) (eol / unexpected_code)
        (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
        {:curr_indent: %nil :}
    |} -> unpack)

expression <-
    inline_expression / indented_index_chain / indented_expression / inline_index / indented_index

inline_nomsu (EscapedNomsu) <- "\" inline_expression
indented_nomsu (EscapedNomsu) <- "\" expression


inline_index (Index) <-
    "." (hex_integer / integer / text_word / noindex_inline_expression)
_inline_index (IndexChain) <- inline_index
inline_index_chain <-
    (noindex_inline_expression _inline_index+) -> foldr

indented_index (Index) <- "." indented_expression
_indented_index (IndexChain) <- indented_index
indented_index_chain <-
    (noindex_inline_expression inline_index* (inline_index / _indented_index)) -> foldr


-- Actions need 1 argument and either another argument or a word.
inline_action (Action) <-
    !section_division
    (  word (ws* inline_arg)*
     / inline_expression ((ws* inline_arg)+ / "(" ws* ")"))
inline_arg <- inline_expression / word

action (Action) <-
    !section_division
    (  word ((linesplit / ws*) arg)*
     / !statement_block arg (((linesplit / ws*) arg)+ / "(" ws* ")"))
arg <- expression / indented_naked_expression / word
linesplit <- eol nl_nodent ".." ws* 


inline_methodsuffix (MethodCall) <-
    inline_action
    / "(" ws* inline_action (ws* ";" ws* inline_action)* ")"
inline_methodchain <-
    ((inline_action / inline_expression) (ws* "," ws* inline_methodsuffix)+) -> foldr

methodsuffix (MethodCall) <-
      action
    / "(" ws* inline_action (ws* ";" ws* inline_action)* ws* ")"
    / eol ({| nl_indent
        (action eol) (nl_nodent action eol)* 
        (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
        {:curr_indent: %nil :}
    |} -> unpack)
methodchain <-
    ((inline_action / expression) ((linesplit / ws*) "," ws* methodsuffix)+) -> foldr

word <- !number { operator_char+ / ident_char+ }


text_word (Text) <- word

inline_text (Text) <-
    '"' !eol _inline_text* ('"' / eof / missing_quote_err / unexpected_code)
_inline_text <-
    {~ (('\"' -> '"') / ('\\' -> '\') / escaped_char / text_char+)+ ~}
    / inline_text_interpolation / illegal_char
inline_text_interpolation <-
    "\" (
        variable / inline_list / inline_dict
        / ("("
            ws* ((inline_methodchain / inline_action / inline_expression) ws*)?
            (")" / eof / missing_paren_err / unexpected_code))
    )

text_char <- %utf8_char / !["\] %print / %tab

indented_text (Text) <-
    '("' %nl {%nl*} ({|
        {:curr_indent: indent :}
            (indented_plain_text / text_interpolation / illegal_char / blank_text_lines)*
        {:curr_indent: %nil :}
    |} -> unpack)
    (nl_nodent '")' / eof / missing_indented_quote_err)

-- Tracking text-lines-within-indented-text as separate objects allows for better debugging line info
indented_plain_text (Text) <-
    {~ 
        ((("\" blank_lines =curr_indent "..") -> "") / ('\\' -> '\') / ('\"' -> '"') / ('\;' -> '')
        / (!text_interpolation ((!("\n") escaped_char) / '\'))
        / ('"' / text_char)+)+
        blank_text_lines?
    ~}
blank_text_lines <- 
    {~ (%nl ((ws* -> '') (&%nl / !.) / (=curr_indent -> '') &[^%nl]))+ ~}

text_interpolation <-
    ({|
    -- %indentation will backtrack and match the actual indentation of the current line
        "\" {:curr_indent: %indentation :}
        (indented_block (blank_lines =curr_indent "..")? / indented_expression)
    |} -> unpack)
    / inline_text_interpolation


number <-
    hex_integer / real_number / integer

integer (Number) <-
    (("-"? [0-9]+)-> tonumber)

hex_integer (Number) <-
    (("-"? "0x" [0-9a-fA-F]+)-> tonumber)
    {:hex: '' -> 'yes' :}

real_number (Number) <-
    (("-"? [0-9]+ "." [0-9]+)-> tonumber)


variable (Var) <- "$" ({ident_char+} / "(" ws* inline_action ws* ")" / {''})


inline_list (List) <-
    "[" ws* !eol
        (inline_list_item (ws* ',' ws* inline_list_item)* (ws* ',')?)? ws*
    ("]" / eof / (","? (missing_bracket_error / unexpected_code)))
inline_list_item <- inline_action / inline_expression

indented_list (List) <-
    ({|
        "[" eol nl_indent
            list_line (nl_nodent list_line)*
        {:curr_indent: %nil :}
    |} -> unpack)
    (nl_nodent "]" / eof / missing_bracket_error / unexpected_code)
list_line <-
      (inline_list_item ws* "," ws*)+ eol
    / (inline_list_item ws* "," ws*)* (action / statement_block / expression) eol


inline_dict (Dict) <-
    "{" ws* !eol
        ((inline_action / inline_expression) (ws* ',' ws* (inline_action / inline_expression))*)? ws*
    ("}" / eof / (","? (missing_brace_error / unexpected_code)))

indented_dict (Dict) <-
    ({|
        "{" eol nl_indent
            dict_line (nl_nodent dict_line)*
            (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
        {:curr_indent: %nil :}
    |} -> unpack)
    (nl_nodent "}" / eof / missing_brace_error / unexpected_code)
dict_line <-
      ((inline_action / inline_expression) ws* "," ws*)+ eol
    / ((inline_action / inline_expression) ws* "," ws*)* (action / statement_block / expression) eol

operator_char <- [#'`~@^&*+=<>?/%!|-]
ident_char <- [a-zA-Z0-9_] / %utf8_char
ws <- " "

escaped_char <-
    ("\"->'') (
        (([xX]->'') ((({[0-9a-fA-F]^2} %number_16) -> tonumber) -> tochar))
      / ((([0-9] [0-9]^-2) -> tonumber) -> tochar)
      / ("a"->ascii_7)  / ("b"->ascii_8)  / ("t"->ascii_9)  / ("n"->ascii_10)
      / ("v"->ascii_11) / ("f"->ascii_12) / ("r"->ascii_13)
    )


-- Errors
unexpected_code <- ws* _unexpected_code
_unexpected_code (Error) <-
    {:error: {~ [^%nl]+ -> "Couldn't parse this code." ~} :}
unexpected_chunk (Error) <-
    {:error: {~ .+ -> "Couldn't parse this chunk of code." ~} :}
unexpected_indent (Error) <-
    {:error: {~ (=curr_indent ws+) -> "This indentation is messed up." ~} :}
    {:hint: {~ '' -> 'This line should either have the same indentation as the line above it, or exactly 4 spaces more.' ~} :}
missing_paren_err (Error) <-
    {:error: {~ eol -> 'This expression is missing a closing )-parenthesis.' ~} :}
    {:hint: {~ '' -> 'Put a ")" here' ~} :}
missing_quote_err (Error) <-
    {:error: {~ eol -> "This text is missing a closing quotation mark." ~} :}
    {:hint: {~ "" -> "Put a quotation mark here." ~} :}
missing_indented_quote_err (Error) <-
    {:error: {~ '' -> 'This text is missing a closing ")-quotation mark.' ~} :}
    {:hint: {~ "" -> 'Put a ") after this line, at the same level of indentation as the opening (".' ~} :}
missing_bracket_error (Error) <-
    {:error: {~ eol -> "This list is missing a closing ]-bracket" ~} :}
    {:hint: {~ '' -> 'Put a "]" here' ~} :}
missing_brace_error (Error) <-
    {:error: {~ eol -> "This dict is missing a closing }-brace" ~} :}
    {:hint: {~ '' -> 'Put a "}" here' ~} :}
tab_error (Error) <-
    &(=curr_indent %tab)
    {:error: {~ '' -> 'Tabs are not allowed for indentation.' ~} :}
    {:hint: {~ '' -> 'Use spaces instead of tabs.' ~} :}
illegal_char (Error) <-
    {:error: {~ (!(%nl / %tab / %print) .) -> "Illegal unprintable character here (it may not be visible, but it's there)" ~} :}
    {:hint: {~ '' -> "This sort of thing can happen when copying and pasting code. Try deleting and retyping the code." ~} :}