nomsu/nomsu.2.peg

-- Nomsu version 2
file <-
    {:curr_indent: ' '* :}
    (((action / expression / inline_block / indented_block) eol !.)
     / file_chunks / empty_block)
    {:curr_indent: %nil :}
    !.

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

file_chunks (FileChunks) <-
    shebang? comment? blank_lines?
    (top_block (nl_nodent section_division top_block)*)
    blank_lines?
    unexpected_indent? unexpected_chunk?

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

empty_block (Block) <-
    comment? blank_lines?

nodent <- (unexpected_indent [^%nl]* / =curr_indent)
indent <- =curr_indent "    "
blank_lines <- %nl ((nodent comment / ws*) %nl)*
eol <- ws* eol_comment? (!. / &%nl)

nl_nodent <- blank_lines nodent
nl_indent <- blank_lines {:curr_indent: indent :} (comment nl_nodent)*

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

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 -> 'Line ended without finding a closing )-parenthesis' ~} :}
    {:hint: {~ '' -> 'Put a ")" here' ~} :}
missing_quote_err (Error) <-
    {:error: {~ eol -> 'Line ended before finding a closing double quotation mark' ~} :}
    {:hint: {~ "" -> "Put a quotation mark here" ~} :}
missing_bracket_error (Error) <-
    {:error: {~ eol -> "Line ended before finding a closing ]-bracket" ~} :}
    {:hint: {~ '' -> 'Put a "]" here' ~} :}
missing_brace_error (Error) <-
    {:error: {~ eol -> "Line ended before finding a closing }-brace" ~} :}
    {:hint: {~ '' -> 'Put a "}" here' ~} :}

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

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

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

inline_statement <- (inline_action / inline_expression)

noindex_inline_expression <-
    number / variable / inline_text / inline_list / inline_dict / inline_nomsu
    / ( "("
            ws* (inline_action / inline_expression) ws*
            (ws* ',' ws* (inline_action / inline_expression) ws*)*
        (")" / missing_paren_err / unexpected_code)
      )
inline_expression <- index_chain / noindex_inline_expression 
indented_expression <-
    indented_text / indented_nomsu / indented_list / indented_dict / ({|
        "(..)" nl_indent
            (action / expression) (eol / unexpected_code)
            (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
        {:curr_indent: %nil :}
    |} -> unpack)
expression <-
    inline_expression / indented_expression

inline_nomsu (EscapedNomsu) <- "\" (inline_expression / inline_block)
indented_nomsu (EscapedNomsu) <-
    "\" (noindex_inline_expression / inline_block / indented_expression / indented_block)

index_chain (IndexChain) <-
    noindex_inline_expression ("." (text_word / noindex_inline_expression))+

-- Actions need either at least 1 word, or at least 2 tokens
inline_action (Action) <-
    !section_division
    (   (inline_arg (ws* (inline_arg / word))+)
      / (word       (ws* (inline_arg / word))*))
    (ws* inline_block)?
inline_arg <- inline_expression / inline_block
action (Action) <-
    !section_division
    (  (arg  ((nl_nodent "..")? ws* (arg / word))+)
     / (word ((nl_nodent "..")? ws* (arg / word))*))
arg <- expression / inline_block / indented_block

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

text_word (Text) <- word

inline_text (Text) <-
    !(indented_text)
    ('"' _inline_text* ('"' / missing_quote_err / unexpected_code))
_inline_text <-
    {~ (('\"' -> '"') / ('\\' -> '\') / escaped_char / [^%nl\"]+)+ ~}
    / inline_text_interpolation
inline_text_interpolation <-
    "\" (
        variable / inline_list / inline_dict / inline_text
        / ("("
            ws* (inline_action / inline_expression) ws*
            (ws* ',' ws* (inline_action / inline_expression) ws*)*
        (")" / missing_paren_err / unexpected_code))
    )

indented_text (Text) <-
    '".."' eol %nl {%nl+}? {:curr_indent: indent :}
        (indented_plain_text / text_interpolation / {~ %nl+ (=curr_indent -> "") ~})*
    unexpected_code?
    {:curr_indent: %nil :}
-- 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 "\") / [^%nl\]+)+
        (%nl+ (=curr_indent -> ""))* ~}
text_interpolation <-
    inline_text_interpolation / ("\" indented_expression (blank_lines =curr_indent "..")?)

number (Number) <- (("-"? (([0-9]+ "." [0-9]+) / ("." [0-9]+) / "0x" [0-9a-fA-F]+ / ([0-9]+)))-> tonumber)

-- Variables can be nameless (i.e. just %) and can only contain identifier chars.
-- This ensures you don't get weird parsings of `%x+%y` or `%'s thing`.
variable (Var) <- "%" {ident_char*}

inline_list (List) <-
    !('[..]')
    "[" ws*
        (inline_list_item (ws* ',' ws* inline_list_item)* (ws* ',')?)? ws*
    ("]" / (","? (missing_bracket_error / unexpected_code)))
indented_list (List) <-
    "[..]" eol nl_indent
        list_line (nl_nodent list_line)*
        (%nl (ws* %nl)* nodent (comment / eol / unexpected_code))*
    (","? unexpected_code)?
list_line <-
      (inline_list_item ws* "," ws*)+ eol
    / (inline_list_item ws* "," ws*)* (action / expression) eol
inline_list_item <- inline_action / inline_expression

inline_dict (Dict) <-
    !('{..}')
    "{" ws*
        (inline_dict_entry (ws* ',' ws* inline_dict_entry)*)? ws*
    ("}" / (","? (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))*
    (","? unexpected_code)?
dict_line <-
      (inline_dict_entry ws* "," ws*)+ eol
    / (inline_dict_entry ws* "," ws*)* dict_entry eol
dict_entry(DictEntry) <-
    dict_key (ws* ":" ws* (action / expression))?
inline_dict_entry(DictEntry) <-
    dict_key (ws* ":" ws* (inline_action / inline_expression)?)?
dict_key <-
    text_word / inline_expression

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

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)
    )