A brand-new version 0.9.2 of the "readable" package is now available for your Lisping pleasure. You can get the tarball here: https://sourceforge.net/projects/readable/files/
Here's a selected list of the goodness that awaits you, straight from the ChangeLog: - Modify grammar and implementation to support "#;"+space|tab the same way as abbreviation+space|tab. Now "#;", followed by space or tab, comments out the next *sweet-expression* (not the next *neoteric-expression*). This is per a comment by Mark H. Weaver. - Add code+grammar rules, and supporting procedures, to handle empty cases (for #;+space|tab handling) - Better error detection: Forbid lines with >1 n-expr on a line where the last expression is "." - Better error detection: In guile implementation forbid unescaped FF/VT when not on their own line (when outside a collecting list). This has been forbidden in the grammar for a long time, but now the guile implementation detects it and complains about it - Better Common Lisp performance. Several tail recursions were turned into loops, which speeds up some implementations. The switch from tail recursion to "loop" in neoteric.lisp's symbol reading, all by itself, produced a 5% performance improvement when using clisp. - Lots of changes to SRFI-110 to make it simpler, clearer, and even more precise. In several places tried to create consistent terminology, e.g., rename "empty tag" to "empty value", and change "empty line" to "blank line" (so that an empty line is not confused with other "empty" things). The phrase "returns [Object v]" is no longer repeated in every rule that produces a result, which makes it easier to read. Many of these changes originate in comments by Mark H. Weaver and John Cowan. - SRFI-110 now *REQUIRES* SRFI-30 (#|...|#) and SRFI-62 (#;datum), instead of being a SHOULD. The sample implementations implement them anyway, but by requiring them, SRFI-110 users can depend on them. - Added code to work around clisp REPL bug in initial-indent mode (the code has no effect, but also causes no harm, in other implementations like SBCL). - Modified sweet-clisp to patch its REPL (by default) to work around other clisp REPL bugs, and added an option to inhibit this behavior. Otherwise clisp will throw away a second t-expr if there is no blank line between the first and second t-expr. - Changed code in several places to make them easier to follow. In particular, new macro let-splitter simplifies code. - Modified sweet.g to eliminate spurious ANTLR warning message - Overall simpler grammar (in draft SRFI-110 and in implementation). This is in part a response to comments by Mark H. Weaver. There are more production rules, but each one is simpler, and "hspaces*" became "hs" (good because it happens often) - Change production names in a few cases to make them clearer. E.G., "head" -> "line_exprs", "rest" -> "rest_of_line", "collecting_tail" -> "collecting_content". - Rename production "ignorable" into "skippable". Common Lisp uses "ignorable" for something else, and having 2 meanings for one phrase - while possible - was confusing. It's easier to just rename things, so that one term works everywhere. - Added "sweet-sbcl" convenience script - Include final version of SRFI-105 - "make distcheck" bugs fixed - Add sample *.spec files for rpm-based distributions - Add new "configure" options: * --without-guile disable support for guile * --without-common-lisp disable support for Common Lisp * --without-clisp disable support for clisp - As always, more tests. In particular, there are now a number of tests for error conditions, to make sure they produce errors --- David A. Wheeler ------------------------------------------------------------------------------ Introducing Performance Central, a new site from SourceForge and AppDynamics. Performance Central is your source for news, insights, analysis and resources for efficient Application Performance Management. Visit us today! http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk _______________________________________________ Readable-discuss mailing list Readable-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/readable-discuss
