Permalink
Switch branches/tags
Nothing to show
Commits on Oct 21, 2010
Commits on Oct 6, 2010
  1. Fixed an infinite-loop case in the lexer (the failure case involves a…

    spencertipping committed Oct 6, 2010
    … file that ends with a line-comment but no EOL)
Commits on Jul 12, 2010
  1. Added some more tests to reproduce a problem in Rhino. Apparently ena…

    spencertipping committed Jul 12, 2010
    …bling the inline macro can cause stack overflows.
Commits on Jun 9, 2010
Commits on Jun 2, 2010
Commits on Jun 1, 2010
Commits on May 31, 2010
  1. Fixed spaces for the dot operator. This doesn't address the lack of a…

    spencertipping committed May 31, 2010
    … good numerical lexer, but it's a step in the right direction.
Commits on May 30, 2010
  1. Added d.rebase.local() for scope-preserving functions, and converted …

    spencertipping committed May 30, 2010
    …the README to ReStructured Text.
Commits on May 29, 2010
  1. Fixed an infinite looping bug that arose when escape characters appea…

    spencertipping committed May 29, 2010
    …red in interpolated strings. There is still the limitation that you shouldn't use escapes for newlines, cr's, Unicode, etc. in your code sequences
Commits on May 28, 2010
Commits on May 27, 2010
  1. Removed support for cached functions, which were a fairly lame use of…

    spencertipping committed May 27, 2010
    … macros anyway. I may do something with nullary functions however.
  2. Referencing the guide

    spencertipping committed May 27, 2010
Commits on May 24, 2010
  1. Reversed the order of assignment expansion and sandwiching. Also noti…

    spencertipping committed May 24, 2010
    …ced that the tests don't pass anymore in node.js; it seems to be a problem with the cached-value macro.
Commits on May 23, 2010
  1. Added cached thunk generation to Rebase. This is invoked by using the…

    spencertipping committed May 23, 2010
    … left-unary $| operator -- see the unit tests for an example.
  2. Fixed a bug that arose if you followed a non-braced while by an else.…

    spencertipping committed May 23, 2010
    … The patch is a rather lame hack that involves looking to see whether we just wrote a block and omitting the semicolon iff that is the case.
Commits on May 22, 2010
  1. Updated the documentation to reflect changes to the way tokens are ma…

    spencertipping committed May 22, 2010
    …rked, and clarified point 2 of the caveats
  2. Added another caveat about do/while loops, and mentioned the caveats …

    spencertipping committed May 22, 2010
    …section on the first line.
  3. Generalized the workaround for if/else. Other cases include try/catch…

    spencertipping committed May 22, 2010
    …, try/catch/finally, try/finally, and do/while. Capturing the semantics of these constructs isn't really the point; rather, I'm more concerned with making sure that there isn't a semicolon in there (which is necessary in the expression model, since the if/else are separate statements.
  4. Added a workaround for if/else. Unfortanately Rebase still cannot qui…

    spencertipping committed May 22, 2010
    …te parse itself; there's something wrong with its detection/anticipation of unary operators.
  5. Cleaned up the operator aliasing logic. Now you can alias operators t…

    spencertipping committed May 22, 2010
    …o regular functions by using the alias_in method of d.rebase.
  6. Changed the array flat_map binding to monadic notation, since a monad…

    spencertipping committed May 22, 2010
    …ic array's bind operator is a flat map.