无法识别的字符的 ANTLR 打印索引（在行中）

Question

在我的 ANTLR 语法中，我有以下语句：

@lexer::apifuncs
{
    LEXER->rec->displayRecognitionError = recognition_error;
}

@parser::apifuncs
{
    RECOGNIZER->displayRecognitionError = recognition_error;
}

C函数"recognition_error"（处理ANTLR抛出的识别错误）定义如下：

void recognition_error(pANTLR3_BASE_RECOGNIZER recognizer, pANTLR3_UINT8 *token_list)
{

    pANTLR3_EXCEPTION           ex;
    pANTLR3_LEXER               lexer;
    pANTLR3_PARSER              parser;
    pANTLR3_TREE_PARSER         tparser;
    pANTLR3_INPUT_STREAM        ins;
    pANTLR3_INT_STREAM          is;
    pANTLR3_COMMON_TOKEN_STREAM cts;
    pANTLR3_TREE_NODE_STREAM    tns;

    if (recognizer->type == ANTLR3_TYPE_LEXER)
    {
        lexer = (pANTLR3_LEXER) (recognizer->super);
        ins = lexer->input;
        is = ins->istream;
    }
    else if (recognizer->type == ANTLR3_TYPE_PARSER)
    {
        parser  = (pANTLR3_PARSER) (recognizer->super);
        cts = (pANTLR3_COMMON_TOKEN_STREAM)(parser->tstream->super);
        is = parser->tstream->istream;
    }
    else if (recognizer->type == ANTLR3_TYPE_TREE_PARSER)
    {
        tparser = (pANTLR3_TREE_PARSER) (recognizer->super);
        tns = tparser->ctnstream->tnstream;
        is = tns->istream;
    }
    else
    {
        printf("UNKNOWN ERROR!\n");
        return;
}

    printf("INFO <%d> * <%d> * <%d> * <%d>\n", recognizer->state->exception->line, recognizer->state->exception->charPositionInLine, recognizer->state->exception->index, ((pANTLR3_COMMON_TOKEN) recognizer->state->exception->token)->getCharPositionInLine);

}

每次我使用此语法解析一些字符错误的文本时，函数 "recognition_error" 不会打印发现错误的字符的索引（它只是打印垃圾，如 -1 或0).

如何固定此语法 and/or C 函数以打印正确的索引？

Answer 1

可用的错误信息取决于异常的类型和它出现的令牌类型（例如，树重写生成的虚拟节点）。在 MySQL Workbench 中，我编写了一些错误处理代码，这些代码花费了一些时间来找到要报告的好信息。这里重要的是检查 exception->type 值并采取相应的行动。

无法识别的字符的 ANTLR 打印索引（在行中）

ANTLR print index (in the line) of unrecognized char

antlr

antlr3