作为学校项目的一部分，我必须识别 .dot 文件并生成相应的解析树。为此，我必须使用我遇到困难的 ocamllex 和 ocamlyacc ......
这是我的 ocaml .mli 类型文件：
type id = string
and node = id
and attr = id * id
and attr_list = attr list list
and attr_stmt =
Graphs of (attr_list)
| Nodes of (attr_list)
| Edge of (attr_list)
and node_stmt = node * attr_list
and node_subgraph =
Node of (node)
| Subgraphs of (graph)
and edge_stmt = node_subgraph * (node_subgraph list) * attr_list
and stmt =
Node_stmt of (node_stmt)
| Edge_stmt of (edge_stmt)
| Attr_stmt of (attr_stmt)
| Attr of (attr)
| Subgraph of graph
and graph =
Graph_node of (node * stmt list)
| Graph of (stmt list);;
这是我的词法分析器文件：
{
open Parser
}
rule token = parse
(* Espacements *)
| [ ' ' '\t' '\n']+ {token lexbuf}
(* *)
| "graph" {GRAPH}
| "subgraph" {SUBGRAPH}
| "--" {EDGE}
(* Délimiteurs *)
| "{" {LEFT_ACC}
| "}" {RIGHT_ACC}
| "(" {LEFT_PAR}
| ")" {RIGHT_PAR}
| "[" {LEFT_BRA}
| "]" {RIGHT_BRA}
| "," {COMMA}
| ";" {SEMICOLON}
| "=" {EQUAL}
| ":" {TWOPOINT}
| ['a'-'z''A'-'Z''0'-'9']* as id {ID (id)}
| eof { raise End_of_file }
这是我未完成的 yacc 文件：
%{
open Types
%}
%token <string> ID
%token <string> STR
%token GRAPH SUBGRAPH EDGE
%token LEFT_ACC RIGHT_ACC LEFT_PAR RIGHT_PAR LEFT_BRA RIGHT_BRA
%token COMME SEMICOLON EQUAL TWOPOINT EOF
%start main
%type <graph> main
%%
main:
graph EOF { $1 }
graph:
GRAPH ID LEFT_ACC content RIGHT_ACC {$4}
| GRAPH LEFT_ACC content RIGHT_ACC {$3}
subgraph:
SUBGRAPH ID LEFT_ACC content RIGHT_ACC {$4}
| SUBGRAPH LEFT_ACC content RIGHT_ACC {$3}
content:
| ID EDGE ID SEMICOLON {[($1,$3)]}
识别一个简单的点文件似乎就足够了
graph D {
A -- B ;
}
但是当我尝试编译我的解析器接口时，我得到了这个错误：这个表达式的类型是'a list，但是一个表达式应该是 Types.graph 类型的（指ID EDGE ID SEMICOLON {[$1,$3)]}line ）
我不明白，因为 {[$1,$3]} 有一个 (string * string) 列表类型。如果我们正在寻找可以是图表的 types.mli。
否则，我是否正确理解 ocamllex 和 ocamlyacc 的运行？