This is an archive of the discontinued Mercurial Phabricator instance.

Differential D377

simplemerge: add `filtereddata=False` to simplemerge()
AbandonedPublic

Authored by phillco on Aug 14 2017, 2:15 AM.

Details

Reviewers
None
Group Reviewers
hg-reviewers
Summary

In the next patch we'll make this function only take contexts, and also make
contrib/simplemerge work by having it pass fake context-like objects.

In order for those fake objects to work, simplemerge needs to know not to run
the repo decoding filters on them -- the data will already be decoded, coming
off the filesystem, and we won't have the repo object -- hence this flag.

Diff Detail

Repository
rHG Mercurial
Lint
Lint Skipped
Unit
Unit Tests Skipped

Event Timeline

phillco created this revision.Aug 14 2017, 2:15 AM
durin42 added a subscriber: durin42.Aug 15 2017, 5:39 PM

(from an IRC conversation)

I'm not crazy about the shape this makes the API, as I'd rather we didn't start growing awareness of filtering in layers this high.

I think we can probably kill contrib/simplemerge entirely - it's had nothing but cleanup changes and a feature removal since May of 2007, and I can't fathom a use for it today. Let's see if we can get consensus around that on the mailing lists and then hopefully simplify this a TON.

durin42 added a comment.Aug 15 2017, 5:40 PM

Also, timeless noticed abd66eb0889e, which reinforces my belief that we don't need contrib/simplemerge anymore.

phillco abandoned this revision.Aug 17 2017, 2:38 PM

Abandoned in favor of D434's approach.

Revision Contents
Changeset List

PathPackages
Mmercurial/simplemerge.py (31 lines)

Diff 858

mercurial/simplemerge.py

if len(overrides) > 2: if len(overrides) > 2:
name_base = overrides[2] name_base = overrides[2]
if len(overrides) > 3: if len(overrides) > 3:
raise error.Abort(_("can only specify three labels.")) raise error.Abort(_("can only specify three labels."))
return [name_a, name_b, name_base] return [name_a, name_b, name_base]
def simplemerge(ui, localfile, basefile, otherfile, def simplemerge(ui, localfile, basefile, otherfile,
localctx=None, basectx=None, otherctx=None, repo=None, **opts): localctx=None, basectx=None, otherctx=None, repo=None,
filtereddata=False, **opts):
"""Performs the simplemerge algorithm. """Performs the simplemerge algorithm.
{local|base|other}ctx are optional. If passed, they (local/base/other) will {local|base|other}ctx are optional. If passed, they (local/base/other) will
be read from and the merge result written to (local). You should pass be read from and the merge result written to (local). You should pass
explicit labels in this mode since the default is to use the file paths.""" explicit labels in this mode since the default is to use the file paths.
filtereddata should only be True if the data() in your context returns
decoded data.
"""
def readfile(filename): def readfile(filename):
f = open(filename, "rb") f = open(filename, "rb")
text = f.read() text = f.read()
f.close() f.close()
return _verifytext(text, filename, ui, opts) return _verifytext(text, filename, ui, opts)
def readctx(ctx): def readctx(ctx):
if not ctx: if not ctx:
return None return None
if filtereddata:
return _verifytext(ctx.data(), ctx.path(), ui, opts)
else:
if not repo: if not repo:
raise error.ProgrammingError('simplemerge: repo must be passed if ' raise error.ProgrammingError('simplemerge: repo must be passed '
'using contexts') 'if using contexts and '
'filtereddata is False.')
# `wwritedata` is used to get the post-filter data from `ctx` (i.e., # `wwritedata` is used to get the post-filter data from `ctx` (i.e.,
# what would have been in the working copy). Since merges were run in # what would have been in the working copy). Since merges were run
# the working copy, and thus used post-filter data, we do the same to # in the working copy, and thus used post-filter data, we do the
# maintain behavior. # same to maintain behavior.
return repo.wwritedata(ctx.path(), text = _verifytext(ctx.data(), ctx.path(), ui, opts)
_verifytext(ctx.data(), ctx.path(), ui, opts)) return repo.wwritedata(ctx.path(), text)
class ctxwriter(object): class ctxwriter(object):
def __init__(self, ctx): def __init__(self, ctx):
self.ctx = ctx self.ctx = ctx
self.text = "" self.text = ""
def write(self, text): def write(self, text):
self.text += text self.text += text