Skip to content

Definition of <bit string literal> not found.  #24

Open

Description

@GCer-Hidenori

At line 220 of sql-2003-2.bnf,<bit string literal> is used.
But there is no definition.

Activity

  1. ronsavage commented on Mar 15, 2020

    @ronsavage
    Owner
  2. GCer-Hidenori commented on Mar 15, 2020

    @GCer-Hidenori
    ContributorAuthor

    I found strange thing.
    <bit string literal> is used in <nondelimiter token> definition.
    And <nondelimiter token> is used in <token> definition.
    But <token> is used nowhere.

  3. ronsavage commented on Mar 16, 2020

    @ronsavage
    Owner
  4. GCer-Hidenori commented on Mar 23, 2020

    @GCer-Hidenori
    ContributorAuthor

    Hi

    Yes,you are right.
    <token> is defined at line 213

    But there is no reference to it from anywhere.

  5. ronsavage commented on Mar 25, 2020

    @ronsavage
    Owner
  6. GCer-Hidenori commented on Mar 28, 2020

    @GCer-Hidenori
    ContributorAuthor

    I’m very happy to hear your operation was successful.
    I hope you could leave the hospital soon.

  7. ronsavage commented on Mar 28, 2020

    @ronsavage
    Owner
  8. GCer-Hidenori commented on Mar 28, 2020

    @GCer-Hidenori
    ContributorAuthor

    There are times like that.
    Have a good rest.

  9. Schultzer commented on Jun 26, 2025

    @Schultzer

    <token> is a root rule. If the rule is not defined any where then they are the root of the three. E.g when you build a lexer you would use <token> to create each tokens, then you can parse them into an AST and validate them based on the rest of the rules.

  10. ronsavage commented on Jun 26, 2025

    @ronsavage
    Owner

    Hi GCer-Hidenori

    The email I received from Benjamin via github had a couple of references to '' missing. I will copy Benjamin's reply here just to see what happens. I suspect that string will vanish again:

    is a root rule. If the rule is not defined anywhere else then it is the root of a tree. E.g when you build a lexer you would use to create each token, then you can parse them into an AST and validate them based on the rest of the rules.

    And here are some tutorials on ebnf:

    1. https://en.wikipedia.org/wiki/Extended_Backus%E2%80%93Naur_form
    2. https://tomassetti.me/ebnf/

    Lastly note that we are now using sql-2016.ebnf.

  11. ronsavage commented on Jun 26, 2025

    @ronsavage
    Owner

    As I feared. After I click the green Comment button in github, it echos my text with '<token>' reduced to ''.

  12. ronsavage commented on Jun 26, 2025

    @ronsavage
    Owner

    So, one more try, using \<token to stop github gobbling up the <token> token!

    <token> is a root rule. If the rule is not defined anywhere else then it is the root of a tree. E.g when you build a lexer you would use to create each token, then you can parse them into an AST and validate them based on the rest of the rules.

  13. ronsavage commented on Jun 26, 2025

    @ronsavage
    Owner

    Found another typo:
    you would use to create each token
    should read:
    you would use it to create each token

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions

      ⌂ Proxy