Skip to content
Commits on Jun 20, 2013
  1. @gasche

    Merge pull request #398 from UnixJunkie/master

    gasche committed Jun 20, 2013
    fix broken test
  2. new version of combine + unit test for it

    Francois Berenger committed Jun 20, 2013
    this version does not use Obj.magic anymore
  3. one unit test in nsplit was not reflecting the implementation

    Francois Berenger committed Jun 20, 2013
Commits on Jun 19, 2013
  1. @gasche

    start to use Acc.accum everywhere possible

    Francois Berenger committed with gasche Jun 19, 2013
  2. @gasche

    accum is now a function of the Acc module

    Francois Berenger committed with gasche Jun 19, 2013
  3. @gasche

    start to use the Acc module

    Francois Berenger committed with gasche Jun 19, 2013
  4. @gasche

    all calls to dummy_node are the same now

    Francois Berenger committed with gasche Jun 19, 2013
  5. @gasche
  6. @gasche

    whitespace fix

    gasche committed Jun 19, 2013
  7. @gasche
  8. @gasche

    nsplit function for lists

    Francois Berenger committed with gasche Jun 19, 2013
  9. @gasche

    proprify group_consecutive

    gasche committed Jun 19, 2013
Commits on Jun 12, 2013
  1. @gasche

    Merge pull request #389 from UnixJunkie/tail_rec_group_consecutive

    gasche committed Jun 11, 2013
    Tail rec version of group_consecutive
  2. added spaces in the unit tests for group_consecutive

    Francois Berenger committed Jun 12, 2013
  3. tail rec version of group_consecutive

    Francois Berenger committed Jun 12, 2013
  4. @rgrinberg

    Merge pull request #387 from UnixJunkie/batList_group_unit_test

    rgrinberg committed Jun 11, 2013
    added unit tests for group
  5. added unit tests for group

    Francois Berenger committed Jun 12, 2013
Commits on Jun 8, 2013
  1. @vincent-hugot

    list: group_nosort -> group_consecutive.

    vincent-hugot committed Jun 8, 2013
    Documentation notes. Cf. issue #384.
  2. @gasche

    add an Array1.of_enum function to Bigarray, strengthen Array1.enum sp…

    gasche committed Jun 8, 2013
    …ecification
    
    In an ideal world we should also provide enum / of_enum pair for
    higher-dimensional cases, but in practice most people only use Array1
    anyway (to talk with C bindings). Note that enum already exists in the
    generic-dimension case GenArray, but the new implementation is better
    specified as it guarantees the expected sequential ordering.
  3. @gasche
Commits on Jun 5, 2013
  1. @vincent-hugot

    list: added group_nosort function, which performs a non-sorting group…

    vincent-hugot committed Jun 5, 2013
    …ing.
    
    Similar to Haskell's groupBy. Cf. issue #384.
  2. @vincent-hugot

    list: added span function.

    vincent-hugot committed Jun 5, 2013
    Similar to Haskell's, but using BatList's mutable implementation.
    Useful for group.
  3. @thelema
  4. @thelema

    Merge pull request #379 from UnixJunkie/avg_for_BatArray

    thelema committed Jun 4, 2013
    added avg, favg + corresponding tests and an internal flength function u...
  5. @thelema
  6. @thelema

    Merge pull request #382 from UnixJunkie/min_max_for_BatList

    thelema committed Jun 4, 2013
    added min_max + unit test and doc
  7. micro fix asked by a maintainer

    Francois Berenger committed Jun 5, 2013
  8. added min_max + unit test and doc

    Francois Berenger committed Jun 5, 2013
  9. added avg, favg + corresponding tests and an internal flength functio…

    Francois Berenger committed Jun 5, 2013
    …n used by both
Commits on Jun 4, 2013
  1. @rixed @thelema

    Faster version of BatString.nreplace

    rixed committed with thelema Jun 4, 2013
    With an accompanying benchmark against other possible implementations
    showing this new version being consistently faster than current one
    by around 20%.
    
    Also introduce BatSubstring.unsafe_substring (used by a benchmark but
    useful in itself).
  2. @rixed @thelema
  3. @thelema

    Faster version of BatString.replace

    Glyn Webster committed with thelema Jun 4, 2013
    This version of 'replace' only allocates a single string. On short
    strings it takes 60% of the time of the old version. Perhaps most the
    time is spent in 'find'.
Commits on Jun 3, 2013
  1. @thelema
  2. @thelema

    Fix Int32.pack/unpack functions; invalid_argument test for being past…

    thelema committed Jun 3, 2013
    … string length was incorrect direction
Commits on May 28, 2013
  1. @gasche

    Merge pull request #376 from rgrinberg/string_reverse

    gasche committed May 28, 2013
    String.rev
Something went wrong with that request. Please try again.