- Lex vs straight parse? Bison uses a lexer first
grammatical classification symbol terminal, or token type: UPPERCASE conventionally nonterminal: non-UPPERCASE conventionally
three types of terminals: named token type: a C-like identifier, defined with %token character token type: 'a' '\n' literal string token: "foo" "bar", can be aliased to a named token with %token
Symbol names (named token type) can contain letters, underscores, periods, and non-initial digits and dashes.
{...} braced code, or an action, used for determining the semantic value of a rule
%type defines a new type %printer declares a block of c code to print the type %define a variable to adjust bison's behavior
=========================
========================= Goals:
- Use the bison grammar scanner defined in src/scan-gram.l/c/h to tokenize the input - token types defined in parse-gram.h - should bison output a stream of tokens on stdout? or should this be a custom tokenizer
- Use the bison grammar parser defined in src/parse-gram.y/c/h to parse the stream of input tokens - the token types and value types (declared using angle brackets, ) are generated with bison -d gram.y, see #1.
- With the productions, rules, symbols (terminals and non terminals), and token types, provide a DSL for mapping/sorting/partitioning/filtering and finally emitting (rendering) a particular node to a template file.
Refs: ** 1. https://www.gnu.org/software/bison/manual/bison.html#Token-Type-Names ** 2. https://www.gnu.org/software/bison/manual/bison.html#https://www.gnu.org/software/bison/manual/bison.html
=========================
%token GRAM_EOF 0 "end of file" %token STRING "string" %token PERCENT_TOKEN "%token" %token PERCENT_NTERM "%nterm" %token PERCENT_TYPE "%type" %token PERCENT_DESTRUCTOR "%destructor" %token PERCENT_PRINTER "%printer" %token PERCENT_LEFT "%left" %token PERCENT_RIGHT "%right" %token PERCENT_NONASSOC "%nonassoc" %token PERCENT_PRECEDENCE "%precedence" %token PERCENT_PREC "%prec" %token PERCENT_DPREC "%dprec" %token PERCENT_MERGE "%merge" %token %token BRACED_CODE "{...}" %token BRACED_PREDICATE "%?{...}" %token BRACKETED_ID "[identifier]" %token CHAR "char" %token COLON ":" %token EPILOGUE "epilogue" %token EQUAL "=" %token ID "identifier" %token ID_COLON "identifier:" %token PERCENT_PERCENT "%%" %token PIPE "|" %token PROLOGUE "%{...%}" %token SEMICOLON ";" %token TAG "" %token TAG_ANY "<*>" %token TAG_NONE "<>" %token INT "integer" %token <param_type> PERCENT_PARAM "%param"; %token PERCENT_UNION "%union"; %token PERCENT_EMPTY "%empty";