> it'd be nice to genericize MultiLevelSkipListWriter so that it could index
arbitrary files

+1 on this idea.  Using skip lists for the term index would be an
improvement.

On Tue, Nov 18, 2008 at 12:27 PM, Michael McCandless (JIRA) <jira@apac...
> wrote:

>
>    [
> https://issues.apache.org/jira/browse/LUCENE-1458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12648739#action_12648739]
>
> Michael McCandless commented on LUCENE-1458:
> --------------------------------------------
>
> bq. Can we design a format that allows us rely upon the operating system's
> virtual memory and avoid caching in process memory altogether?
>
> Interesting!  I've been wondering what you're up to over on KS, Marvin :)
>
> I'm not sure it'll be a win in practice: I'm not sure I'd trust the
> OS's IO cache to "make the right decisions" about what to cache.  Plus
> during that binary search the IO system is loading whole pages into
> the IO cache, even though you'll only peak at the first few bytes of
> each.
>
> We could also explore something in-between, eg it'd be nice to
> genericize MultiLevelSkipListWriter so that it could index arbitrary
> files, then we could use that to index the terms dict.  You could
> choose to spend dedicated process RAM on the higher levels of the skip
> tree, and then tentatively trust IO cache for the lower levels.
>
> I'd like to eventually make the TermsDict index pluggable so one could
> swap in different indexers like this (it's not now).
>
>
> > Further steps towards flexible indexing
> > ---------------------------------------
> >
> >                 Key: LUCENE-1458
> >                 URL: https://issues.apache.org/jira/browse/LUCENE-1458
> >             Project: Lucene - Java
> >          Issue Type: New Feature
> >          Components: Index
> >    Affects Versions: 2.9
> >            Reporter: Michael McCandless
> >            Assignee: Michael McCandless
> >            Priority: Minor
> >             Fix For: 2.9
> >
> >         Attachments: LUCENE-1458.patch, LUCENE-1458.patch
> >
> >
> > I attached a very rough checkpoint of my current patch, to get early
> > feedback.  All tests pass, though back compat tests don't pass due to
> > changes to package-private APIs plus certain bugs in tests that
> > happened to work (eg call TermPostions.nextPosition() too many times,
> > which the new API asserts against).
> > [Aside: I think, when we commit changes to package-private APIs such
> > that back-compat tests don't pass, we could go back, make a branch on
> > the back-compat tag, commit changes to the tests to use the new
> > package private APIs on that branch, then fix nightly build to use the
> > tip of that branch?o]
> > There's still plenty to do before this is committable! This is a
> > rather large change:
> >   * Switches to a new more efficient terms dict format.  This still
> >     uses tii/tis files, but the tii only stores term & long offset
> >     (not a TermInfo).  At seek points, tis encodes term & freq/prox
> >     offsets absolutely instead of with deltas delta.  Also, tis/tii
> >     are structured by field, so we don't have to record field number
> >     in every term.
> > .
> >     On first 1 M docs of Wikipedia, tii file is 36% smaller (0.99 MB
> >     -> 0.64 MB) and tis file is 9% smaller (75.5 MB -> 68.5 MB).
> > .
> >     RAM usage when loading terms dict index is significantly less
> >     since we only load an array of offsets and an array of String (no
> >     more TermInfo array).  It should be faster to init too.
> > .
> >     This part is basically done.
> >   * Introduces modular reader codec that strongly decouples terms dict
> >     from docs/positions readers.  EG there is no more TermInfo used
> >     when reading the new format.
> > .
> >     There's nice symmetry now between reading & writing in the codec
> >     chain -- the current docs/prox format is captured in:
> > {code}
> > FormatPostingsTermsDictWriter/Reader
> > FormatPostingsDocsWriter/Reader (.frq file) and
> > FormatPostingsPositionsWriter/Reader (.prx file).
> > {code}
> >     This part is basically done.
> >   * Introduces a new "flex" API for iterating through the fields,
> >     terms, docs and positions:
> > {code}
> > FieldProducer -> TermsEnum -> DocsEnum -> PostingsEnum
> > {code}
> >     This replaces TermEnum/Docs/Positions.  SegmentReader emulates the
> >     old API on top of the new API to keep back-compat.
> >
> > Next steps:
> >   * Plug in new codecs (pulsing, pfor) to exercise the modularity /
> >     fix any hidden assumptions.
> >   * Expose new API out of IndexReader, deprecate old API but emulate
> >     old API on top of new one, switch all core/contrib users to the
> >     new API.
> >   * Maybe switch to AttributeSources as the base class for TermsEnum,
> >     DocsEnum, PostingsEnum -- this would give readers API flexibility
> >     (not just index-file-format flexibility).  EG if someone wanted
> >     to store payload at the term-doc level instead of
> >     term-doc-position level, you could just add a new attribute.
> >   * Test performance & iterate.
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-dev-unsubscribe@luce...
> For additional commands, e-mail: java-dev-help@luce...
>
>