Re: [PATCH 10 of 12 V3] bundle2: support a 'records' mode for the 'bookmarks' part

Potential bug mentioned inline On 11/20/17 8:52 AM, Boris Feld wrote: # HG changeset patch # User Boris Feld # Date 1508246776 -7200 # Tue Oct 17 15:26:16 2017 +0200 # Node ID 7e610585998e3eff1d0498a6ce024350bb04fc23 # Parent

D2108: infinitepush: drop the `--to` flag to push and use `-B` instead

durham added a comment. > There are things which I am not sure whether to keep or not: > > - the --bundle-store flag to push command This is useful for scripts or tools that want to upload a commit to the cloud without having to give it a name. For instance, you can use it to push a

D1814: rebase: add experimental.inmemory.nomergedriver to turn off IMM

durham added inline comments. INLINE COMMENTS > rebase.py:812 > +elif (ui.config('experimental', 'mergedriver') and > + ui.configbool('rebase', > 'experimental.inmemory.nomergedriver')): > +whynotimm = 'mergedriver enabled' I think we probably want IMM disabled

D1806: filemerge: fix backing up an in-memory file to a custom location

durham accepted this revision. durham added a comment. Accepting, since it seems more correct than before. Should we even be calling _makebackup in the case of an inmemory merge? Like, maybe the makebackup should be conditional based on if the source file context is actually a

D1564: worker: make windows workers daemons

durham added inline comments. INLINE COMMENTS > worker.py:286 > +threads.remove(t) > +except KeyboardInterrupt: > +trykillworkers() Why only do it on keyboard interrupt? What if there's another exception? If you did it for all exceptions, you could drop the

D821: unamend: move fb extension unamend to core

durham added inline comments. INLINE COMMENTS > pulkit wrote in uncommit.py:265 > Okay while trying to add this condition, I found we cannot refuse to unamend > a changeset on the basis of unamend_source, for e.g > `a -amend-> b -unamend-> a' -amend-> c -unamend-> a''` > > But if we refuse on

D1358: remotenames: store journal entry for bookmarks if journal is loaded

durham added a comment. There seems like two use cases for remotenames: knowledge about the most recently seen location of each remote bookmark, and knowledge about where the remote bookmarks have been over time. I think 99% of the benefit of remotenames comes from the first part.

Re: Upstreaming Facebook extensions

On 11/14/17 8:26 AM, Augie Fackler wrote: On Nov 13, 2017, at 22:49, Matt Harbison wrote: On Mon, 13 Nov 2017 22:12:31 -0500, Jun Wu wrote: Excerpts from Matt Harbison's message of 2017-11-13 21:50:29 -0500: For LFS, it has been used in a repo

D1431: sshpeer: making the ssh error message configurable

durham added a comment. I'd also update the commit summary with an example of what a better message might be. Like "there was an ssh error, please see http://company/internalwiki/ssh.html; REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D1431 To: zuza,

D1431: sshpeer: making the ssh error message configurable

durham requested changes to this revision. durham added inline comments. This revision now requires changes to proceed. INLINE COMMENTS > sshpeer.py:210 > +"errormessage", > +_("no suitable response from remote hg" > The standard Mercurial way to format

D1358: remotenames: store journal entry for bookmarks if journal is loaded

durham added a comment. I'm a little confused here, in the same way Pulkit has expressed. remotenames has a couple parts related to storage: one is the storing of the remotenames themselves, and the other is tracking how they change over time. The remotename storage itself doesn't seem

D821: unamend: move fb extension unamend to core

durham accepted this revision. durham added a comment. Overall looks good to me. My one comment is probably not enough to block this going in. INLINE COMMENTS > uncommit.py:260 > +prednode = markers[0].prednode() > +predctx = unfi[prednode] > + Might be worth doing the predecessor

D1329: bundle: allow bundlerepo to support alternative manifest implementations

This revision was automatically updated to reflect the committed changes. Closed by commit rHGa2dfc723b6b5: bundle: allow bundlerepo to support alternative manifest implementations (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D1329: bundle: allow bundlerepo to support alternative manifest implementations

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY With our treemanifest logic, the manifests are no longer transported as part of the changegroup and are no longer stored in a revlog. This means the

Re: Storage format for remotenames.

I wish we had some easily reusable serializer/deserializer instead of having to reinvent these every time. What's our reasoning for not using json? I forget. If there are some weird characters, like control characters or something, that break json, I'd say we just use json and prevent users

D1257: dirstate: remove excess attribute lookups for dirstate.status (issue5714)

This revision was automatically updated to reflect the committed changes. Closed by commit rHGffeea2406276: dirstate: remove excess attribute lookups for dirstate.status (issue5714) (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D1257: dirstate: remove excess attribute lookups for dirstate.status (issue5714)

durham added a comment. Fixed by using _map.clear() instead of replacing the _map. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D1257 To: durham, #hg-reviewers, yuja Cc: mercurial-devel ___ Mercurial-devel mailing

D1257: dirstate: remove excess attribute lookups for dirstate.status (issue5714)

durham updated this revision to Diff 3147. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D1257?vs=3136=3147 REVISION DETAIL https://phab.mercurial-scm.org/D1257 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D1253: dirstate: avoid reading the map when possible (issue5713) (issue5717)

This revision was automatically updated to reflect the committed changes. Closed by commit rHG6e66033f91cc: dirstate: avoid reading the map when possible (issue5713) (issue5717) (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D1252: dirstate: move clear onto dirstatemap class

This revision was automatically updated to reflect the committed changes. Closed by commit rHG0217f75b6e59: dirstate: move clear onto dirstatemap class (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D1252?vs=3128=3145

D1257: dirstate: remove excess attribute lookups for dirstate.status (issue5714)

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY A recent refactor added a layer of abstraction to the dirstate which makes doing things like 'foo in dirstate' now require some extra Python attribute lookups.

D1144: directaccess: add support to export and tests to demonstrate things

durham added a comment. @lothiraldan Are pinned revs deleted when the cache is rebuilt? I would kinda not expect them to be. If they are deleted, then yea moving these exceptions to another attribute that is persisted across cache clears makes sense. For your concern about changing

D1201: dirstate: clean up when restoring identical backups

durham accepted this revision. durham added a comment. lgtm. Is this a recent regression? Seems like this would be polluting lots of peoples repositories. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D1201 To: mbthomas, #hg-reviewers, krbullock, durham Cc:

D1253: dirstate: avoid reading the map when possible (issue5713) (issue5717)

durham added a comment. @mharbison72 your test output looks like the sort order changed? Does it consistently repro before and after this patch? The sort order is from the sort -u in the test, so I can't imagine this affecting it. REPOSITORY rHG Mercurial REVISION DETAIL

D1253: dirstate: avoid reading the map when possible (issue5713) (issue5717)

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Before the recent refactor, we would not load the entire map until it was accessed. As part of the refactor, that got lost and even just trying to load the

D1252: dirstate: move clear onto dirstatemap class

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY A future diff will move the lazy loading aspect of dirstate to the dirstatemap class. This means it requires a slightly different strategy of clearing than just

Re: Deploying registrar configs in a BC safe way

On 10/25/17 5:48 AM, Augie Fackler wrote: On Oct 25, 2017, at 07:51, Yuya Nishihara <y...@tcha.org> wrote: On Tue, 24 Oct 2017 18:49:50 -0700, Gregory Szorc wrote: On Oct 24, 2017, at 10:25, Durham Goode <dur...@fb.com> wrote: I've been trying to update our extensions to sup

Re: [PATCH remotenames] configs: update configs to use registrar

I didn't fill in the commit message on this one. V2 incoming On 10/24/17 2:15 PM, Durham Goode wrote: # HG changeset patch # User Durham Goode <dur...@fb.com> # Date 1508879615 25200 # Tue Oct 24 14:13:35 2017 -0700 # Node ID 9ead5c33dd22a9a66cea009bcf76ffd801d5677e #

[PATCH remotenames V2] configs: update configs to use registrar

# HG changeset patch # User Durham Goode <dur...@fb.com> # Date 1508879721 25200 # Tue Oct 24 14:15:21 2017 -0700 # Node ID c8d9ee2123baadd3ca1b571b3e7acf86ca1e6783 # Parent 5fb97b6b3d46fcd4dc1b3cb6be137762eb9e60e7 configs: update configs to use registrar Upstream now gives devel wa

[PATCH remotenames] configs: update configs to use registrar

# HG changeset patch # User Durham Goode <dur...@fb.com> # Date 1508879615 25200 # Tue Oct 24 14:13:35 2017 -0700 # Node ID 9ead5c33dd22a9a66cea009bcf76ffd801d5677e # Parent 5fb97b6b3d46fcd4dc1b3cb6be137762eb9e60e7 configs: update configs to use registrar diff --git a/remotename

Deploying registrar configs in a BC safe way

I've been trying to update our extensions to support the new registrar class, and one thing I've noticed is that it's not really possible to have code that works with old versions of mercurial and new versions of mercurial at the same time, without having devel warnings. I can easily stub out

D983: dirstate: move the _dirfoldmap to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHGe8a89ed7ce96: dirstate: move the _dirfoldmap to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D983?vs=2513=2713

D981: dirstate: remove _filefoldmap property cache

This revision was automatically updated to reflect the committed changes. Closed by commit rHGbfddc3d678ae: dirstate: remove _filefoldmap property cache (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D981?vs=2511=2710

D982: dirstate: remove _dirs property cache

This revision was automatically updated to reflect the committed changes. Closed by commit rHG014bd2a555c8: dirstate: remove _dirs property cache (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D982?vs=2512=2711

D979: dirstate: move nonnormal and otherparent sets to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG60927b19ed65: dirstate: move nonnormal and otherparent sets to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D978: dirstate: move write into dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHGe2214632c3a2: dirstate: move write into dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D978?vs=2508=2707

D977: dirstate: move _read into dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHGe159f217230e: dirstate: move _read into dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D977?vs=2507=2706

D980: dirstate: move identity to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHGc6ef9a2498a5: dirstate: move identity to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D980?vs=2510=2709

D960: bundle2: immediate exit for ctrl+c (issue5692)

durham added inline comments. INLINE COMMENTS > yuja wrote in bundle2.py:380 > SignalInterrupt is a subclass of KeyboardInterrupt, so there might > be deeper problem. > > FWIW, I think it's probably a good idea to replace this blacklist > with `isinstance(exc, Exception)`. Before my refactor

D945: fsmonitor: update to match new dirstate refactor

This revision was automatically updated to reflect the committed changes. Closed by commit rHG7259f0ddfc0f: fsmonitor: update to match new dirstate refactor (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D981: dirstate: remove _filefoldmap property cache

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Now that the filefoldmap is source of truthed on the dirstatemap, let's get rid of the property cache on the dirstate. REPOSITORY rHG Mercurial REVISION

D982: dirstate: remove _dirs property cache

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Now that dirs is source of truthed on the dirstatemap, let's get rid of the _dirs propertycache on the dirstate. REPOSITORY rHG Mercurial REVISION DETAIL

D983: dirstate: move the _dirfoldmap to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Now that dirstatemap is the source of truth for the list of directories, let's move _dirfoldmap on to it. This pattern of moving cached variables onto the

D979: dirstate: move nonnormal and otherparent sets to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of separating dirstate business logic from storage, let's move the nonnormal and otherparent storage to the dirstatemap class. This will allow

D980: dirstate: move identity to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Moving the identity function to the dirstatemap class will allow alternative dirstate implementations to replace the implementation. REPOSITORY rHG Mercurial

D978: dirstate: move write into dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of separating the dirstate business logic from the dirstate storage logic, let's move the serialization code down into dirstatemap. REPOSITORY rHG

D960: bundle2: immediate exit for ctrl+c (issue5692)

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY https://phab.mercurial-scm.org/rHG21c2df59a1dad534bfac45acc0bbfb6cb2afe9b4 regressed bundle2 by catching all exceptions and trying to handle them. The old

Re: [PATCH phase] exchange: fix test for remote support of binary phases

On 9/30/17 6:26 AM, Yuya Nishihara wrote: On Sat, 30 Sep 2017 16:01:36 +0530, Pulkit Goyal wrote: I don't see the traceback after applying this patch. If possible, this patch should be folded with the previous series. On Sat, Sep 30, 2017 at 3:05 PM, Boris Feld wrote:

D945: fsmonitor: update to match new dirstate refactor

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY The dirstate was refactored so dirstate._map is now at dirstate._map._map. Same for _copymap, is not _map.copymap. It seems none of the mercurial tests cover

D759: dirstate: move parents source of truth to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG5fabba9b3d9c: dirstate: move parents source of truth to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D754: dirstate: move filefoldmap to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG19a5ad535cb1: dirstate: move filefoldmap to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D754?vs=2078=2157

D758: dirstate: move parent reading to the dirstatemap class

This revision was automatically updated to reflect the committed changes. Closed by commit rHG949276a8cd76: dirstate: move parent reading to the dirstatemap class (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D752: dirstate: create new dirstatemap class

This revision was automatically updated to reflect the committed changes. Closed by commit rHG42b68c9c2742: dirstate: create new dirstatemap class (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D752?vs=1947=2155

D757: dirstate: move opendirstatefile to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG0407c611b7be: dirstate: move opendirstatefile to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D757?vs=2081=2160

D756: dirstate: move _copymap to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG894cd88815ca: dirstate: move _copymap to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D756?vs=2080=2159

D753: dirstate: move nonnormalentries to dirstatemap

This revision was automatically updated to reflect the committed changes. Closed by commit rHG362ed91ca00c: dirstate: move nonnormalentries to dirstatemap (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D753?vs=1948=2156

D821: unamend: move fb extension unamend to core

durham requested changes to this revision. durham added a comment. This revision now requires changes to proceed. Generally looks good. Just need to fix the transaction thing. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D821 To: pulkit, #hg-reviewers, durham

D821: unamend: move fb extension unamend to core

durham added inline comments. INLINE COMMENTS > uncommit.py:260 > + > +tr = repo.transaction('unamend') > +with dirstate.parentchange(): This should be in a with statement probably? Can we just have it be as part of the top level lock with statement? REPOSITORY rHG

D738: directaccess: add support to export and tests to demonstrate things

durham accepted this revision. durham added inline comments. INLINE COMMENTS > test-directaccess.t:11 > + > [extensions] > + > uncommit = > + > amend = Do we need uncommit? It doesn't seem used. > test-directaccess.t:64 > + > + $ hg exp 7 > + # HG changeset patch I'm still not sure we

D737: directaccess: add support for accessing hidden commits if command is read only

durham requested changes to this revision. durham added a comment. This revision now requires changes to proceed. Probably needs a simple test, unless that comes later in the series. INLINE COMMENTS > dispatch.py:898 > +if cmdtype == readonly: > +

D736: directaccess: add support for storing the type of command in func object

durham accepted this revision. durham added a comment. Minor comments, but lgtm INLINE COMMENTS > registrar.py:143 > +unrecoverablewrite = "unrecoverable" > +recoverablewrite = "recoverable" > +readonly = "readonly" Might be worth a comment describing what these mean, since it's

D728: rebase: move working parent movement logic to scmutil.cleanupnodes

durham requested changes to this revision. durham added inline comments. This revision now requires changes to proceed. INLINE COMMENTS > scmutil.py:603 > +changes will be discarded, the caller is responsible for checking the > +cleanness of working copy. > """ This seems like a

D758: dirstate: move parent reading to the dirstatemap class

durham added inline comments. INLINE COMMENTS > indygreg wrote in dirstate.py:1370 > Nit: this should ideally be in a context manager or try..finally block. But > it existed this way before, so not worth blocking over. Left this alone for now to avoid code churn during the refactor.

D757: dirstate: move opendirstatefile to dirstatemap

durham marked an inline comment as done. durham added inline comments. INLINE COMMENTS > indygreg wrote in dirstate.py:1373-1374 > This (pre-existing) error message is pretty bad because typical end users > won't have a clue what it means or what to do if they see it. Agreed. I'm not even

D756: dirstate: move _copymap to dirstatemap

durham updated this revision to Diff 2080. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D756?vs=1951=2080 REVISION DETAIL https://phab.mercurial-scm.org/D756 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D759: dirstate: move parents source of truth to dirstatemap

durham updated this revision to Diff 2083. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D759?vs=1954=2083 REVISION DETAIL https://phab.mercurial-scm.org/D759 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D757: dirstate: move opendirstatefile to dirstatemap

durham updated this revision to Diff 2081. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D757?vs=1952=2081 REVISION DETAIL https://phab.mercurial-scm.org/D757 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D758: dirstate: move parent reading to the dirstatemap class

durham updated this revision to Diff 2082. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D758?vs=1953=2082 REVISION DETAIL https://phab.mercurial-scm.org/D758 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D755: dirstate: move _dirs to dirstatemap

durham updated this revision to Diff 2079. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D755?vs=1950=2079 REVISION DETAIL https://phab.mercurial-scm.org/D755 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D754: dirstate: move filefoldmap to dirstatemap

durham updated this revision to Diff 2078. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D754?vs=1949=2078 REVISION DETAIL https://phab.mercurial-scm.org/D754 AFFECTED FILES mercurial/dirstate.py CHANGE DETAILS diff --git a/mercurial/dirstate.py

D752: dirstate: create new dirstatemap class

durham added inline comments. INLINE COMMENTS > indygreg wrote in dirstate.py:1345 > Do we want this inherit from `collections.MutableMapping`? > > A good reason to say "no" is KISS, especially if we'll be having multiple > backends. For now I'd say no, until we've done at least one other

D752: dirstate: create new dirstatemap class

durham added a comment. This series is the first 8 commits of a 20 patch series. At the end all the storage for the dirstate goes through the dirstatemap class. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D752 To: durham, #hg-reviewers Cc: mercurial-devel

D758: dirstate: move parent reading to the dirstatemap class

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving dirstate storage logic to a separate class, let's move the function that reads the parents from the file. This will allow extensions to write

D754: dirstate: move filefoldmap to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving the dirstate storage logic to a separate class, lets move the filfoldmap computation to that class. This will allow extensions to replace the

D759: dirstate: move parents source of truth to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving dirstate storage to its own class, let's move the source of truth for the dirstate parents to dirstatemap. This requires that dirstate._pl no

D752: dirstate: create new dirstatemap class

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY This is part of a larger refactor to move the dirstate storage logic to a separate class, so it's easier to rewrite the dirstate storage layer without having to

D757: dirstate: move opendirstatefile to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving the dirstate storage logic to another class, let's move opendirstatefile to dirstatemap. This will allow extensions to replace the pending

D756: dirstate: move _copymap to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving all dirstate storage to a new class, let's move the copymap onto that class. In a future patch this will let us move the read/write functions

D753: dirstate: move nonnormalentries to dirstatemap

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY As part of moving dirstate storage to its own class, let's move the nonnormalentries logic onto the dirstatemap class. This will let extensions replace the

D730: revlog: add revmap back to revlog.addgroup

This revision was automatically updated to reflect the committed changes. durham marked an inline comment as done. Closed by commit rHG1db9abf407c5: revlog: add revmap back to revlog.addgroup (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D746: changegroup: remove dictionary creation from deltachunk

This revision was automatically updated to reflect the committed changes. Closed by commit rHG05131c963767: changegroup: remove dictionary creation from deltachunk (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D745: bundlerepo: update to use new deltaiter api

This revision was automatically updated to reflect the committed changes. Closed by commit rHG0fe62d8bdd50: bundlerepo: update to use new deltaiter api (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D745?vs=1933=1944

D744: debug: update debugbundle to use new deltaiter api

This revision was automatically updated to reflect the committed changes. Closed by commit rHG311f6ccf8f23: debug: update debugbundle to use new deltaiter api (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D730: revlog: add revmap back to revlog.addgroup

durham marked an inline comment as done. durham added a comment. I've updated the message and attached 3 more commits to the stack that update debugbundle, bundlerepo, and the deltachunk fix you suggested. INLINE COMMENTS > martinvonz wrote in changegroup.py:191-192 > I'll repeat my comment

D745: bundlerepo: update to use new deltaiter api

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D745 AFFECTED FILES mercurial/bundlerepo.py CHANGE DETAILS diff --git a/mercurial/bundlerepo.py

D746: changegroup: remove dictionary creation from deltachunk

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Previously delta chunk returned a dictionary. Now that we consume deltachunk within changegroup (instead of outside in revlog) we can just return a tuple and

D744: debug: update debugbundle to use new deltaiter api

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY Changegroup now has a deltaiter api for easy iteration over a series of deltas. Let's use that in the debugbundle command. REPOSITORY rHG Mercurial REVISION

D708: bundle2: remove unnecessary try finally

This revision was automatically updated to reflect the committed changes. Closed by commit rHGcc7b37c90616: bundle2: remove unnecessary try finally (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D708?vs=1820=1875

D709: bundle2: move part processing to a separate function

This revision was automatically updated to reflect the committed changes. Closed by commit rHGf010097c885c: bundle2: move part processing to a separate function (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D707: bundle2: move handler validation out of processpart

This revision was automatically updated to reflect the committed changes. Closed by commit rHG07e4170f02f3: bundle2: move handler validation out of processpart (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D477: revlog: add option to mmap revlog index

durham added a comment. Also, the improvements should become more pronounced if we fix some algorithms (like phases) that access really old commits. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D477 To: mbthomas, #fbhgext, indygreg, #hg-reviewers, durin42,

D696: registrar: add a enum 'cmdtype' for the type of the command

durham accepted this revision. durham added a comment. I think the name could be better, but that can be bikeshed. Stamping my approval for the concept and pattern. INLINE COMMENTS > registrar.py:148 > > +class cmdtype(object): > +""" enum for the type of command which will tell

D706: bundle2: move processpart stream maintenance into part iterator

durham updated this revision to Diff 1818. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D706?vs=1802=1818 REVISION DETAIL https://phab.mercurial-scm.org/D706 AFFECTED FILES mercurial/bundle2.py CHANGE DETAILS diff --git a/mercurial/bundle2.py

D707: bundle2: move handler validation out of processpart

durham updated this revision to Diff 1819. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D707?vs=1803=1819 REVISION DETAIL https://phab.mercurial-scm.org/D707 AFFECTED FILES mercurial/bundle2.py CHANGE DETAILS diff --git a/mercurial/bundle2.py

D708: bundle2: remove unnecessary try finally

durham updated this revision to Diff 1820. REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D708?vs=1804=1820 REVISION DETAIL https://phab.mercurial-scm.org/D708 AFFECTED FILES mercurial/bundle2.py CHANGE DETAILS diff --git a/mercurial/bundle2.py

D705: bundle2: move exception handling into part iterator

This revision was automatically updated to reflect the committed changes. Closed by commit rHG3d8fb0c37e12: bundle2: move exception handling into part iterator (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE

D704: bundle2: move part counter to partiterator

This revision was automatically updated to reflect the committed changes. Closed by commit rHG550343626bb2: bundle2: move part counter to partiterator (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D704?vs=1800=1808

D703: bundle2: move part iterator a separate class

This revision was automatically updated to reflect the committed changes. Closed by commit rHGe9e0e1143fc5: bundle2: move part iterator a separate class (authored by durham, committed by ). REPOSITORY rHG Mercurial CHANGES SINCE LAST UPDATE https://phab.mercurial-scm.org/D703?vs=1799=1807

D706: bundle2: move processpart stream maintenance into part iterator

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY The processpart function also did some stream maintenance, so let's move it to the part iterator as well, as part of moving all part iteration logic into the

D708: bundle2: remove unnecessary try finally

durham created this revision. Herald added a subscriber: mercurial-devel. Herald added a reviewer: hg-reviewers. REVISION SUMMARY This is no longer needed. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D708 AFFECTED FILES mercurial/bundle2.py CHANGE DETAILS

  1   2   3   4   5   >