Actually, yes, your code is exactly what I was looking for. Your "countTrailingZeros" is what I'm calling "mask-offset". Very helpful, thank you. Although, I gotta say, I'm still gonna do it in Clojure. This is supposed to be programming for enjoyment! :)
Best, Dan On Mon, Feb 23, 2015 at 8:39 PM, Mikera <mike.r.anderson...@gmail.com> wrote: > Bit operations is one area where I'd actually recommend writing in Java > before wrapping in Clojure for several reasons: > 1) You can port / compare directly with canonical C implementation > 2) You can be pretty sure of avoiding boxing / other overheads by sticking > to Java primitive maths > 3) Clojure still doesn't quite have the full range of bitwise > functionality as far as I'm aware > > Feel free to use any of my Java bitwise code, if helpful: > > https://github.com/mikera/mikera/blob/master/src/main/java/mikera/util/Bits.java > > > On Tuesday, 24 February 2015 02:59:32 UTC+8, danl...@gmail.com wrote: >> >> So, much of the pain involved in handling UUID's correctly on the JVM >> relates to the fact that there is no primitive unsigned numeric type >> that can represent the full range of possible values of the msb and lsb. >> Ie., we need to always deal with the unpleasant "am I negative?" approach >> to >> reading (writing) that 64th bit. To avoid the complexity of all the >> edge cases, we encapsulate the basic primitives of working with >> unsigned numbers entirely within the abstraction of "mask" and >> "mask offset". Using these, we built the two fundamental bitwise >> operations >> that are used for most of the UUID calculation: ldb (load-byte) and >> dpb (deposit-byte). >> >> This scrap of code from my clj-uuid.bitmop library is extremely useful >> for working >> with "unsigned" long/binary values (analogously to how one might using >> the common-lisp >> functions by the same name). And, it has been "good enough" to do pretty >> well >> so far in terms of performance. But I'm sure that there are gifted >> binariticians >> in the audience that can improve this. (Note, the namespace uses >> ztellman/primitive-math >> which changes the semantics of some arithmetic operations and some type >> hinting. Also >> some of the 'let's are there for that reason. It may be helpful to refer >> to the link. >> >> ;;; https://github.com/danlentz/clj-uuid/blob/master/src/clj_ >> uuid/bitmop.clj >> >> >> (defn ^long expt2 [^long pow] >> (bit-set 0 pow)) >> >> (defn ^long mask [^long width ^long offset] >> (if (< (+ width offset) 64) >> (bit-shift-left (dec (bit-shift-left 1 width)) offset) >> (let [x (expt2 offset)] >> (bit-and-not -1 (dec ^long x))))) >> >> (declare ^long mask-offset ^long mask-width) >> >> (defn ^long mask-offset [^long m] >> (cond >> (zero? m) 0 >> (neg? m) (- 64 ^long (mask-width m)) >> :else (loop [c 0] >> (if (pos? (bit-and 1 (bit-shift-right m c))) >> c >> (recur (inc c)))))) >> >> (defn ^long mask-width [^long m] >> (if (neg? m) >> (let [x (mask-width (- (inc m)))] >> (- 64 ^long x)) >> (loop [m (bit-shift-right m (mask-offset m)) c 0] >> (if (zero? (bit-and 1 (bit-shift-right m c))) >> c >> (recur m (inc c)))))) >> >> (defn ^long ldb >> "Load Byte" >> [^long bitmask ^long num] >> (let [off (mask-offset bitmask)] >> (bit-and (>>> bitmask ^long off) >> (bit-shift-right num off)))) >> >> (defn ^long dpb >> "Deposit Byte" >> [^long bitmask ^long num ^long value] >> (bit-or (bit-and-not num bitmask) >> (bit-and bitmask >> (bit-shift-left value (mask-offset bitmask))))) >> > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to a topic in the > Google Groups "Clojure" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/clojure/XIN2ZerhIcI/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
