Changeset - e85f08375dc6
[Not reviewed]
default
0 5 0
Mads Kiilerich - 8 years ago 2017-10-03 00:14:40
mads@kiilerich.com
diffs: drop the DiffLimitExceeded container - just make it a flag available as property

Keep it simple.
5 files changed with 10 insertions and 40 deletions:
0 comments (0 inline, 0 general)
kallithea/controllers/changeset.py
Show inline comments
 
@@ -39,25 +39,24 @@ from kallithea.lib.vcs.exceptions import
 
import kallithea.lib.helpers as h
 
from kallithea.lib.auth import LoginRequired, HasRepoPermissionLevelDecorator, \
 
    NotAnonymous
 
from kallithea.lib.base import BaseRepoController, render, jsonify
 
from kallithea.lib.utils import action_logger
 
from kallithea.lib.compat import OrderedDict
 
from kallithea.lib import diffs
 
from kallithea.model.db import ChangesetComment, ChangesetStatus
 
from kallithea.model.comment import ChangesetCommentsModel
 
from kallithea.model.changeset_status import ChangesetStatusModel
 
from kallithea.model.meta import Session
 
from kallithea.model.repo import RepoModel
 
from kallithea.lib.diffs import LimitedDiffContainer
 
from kallithea.lib.exceptions import StatusChangeOnClosedPullRequestError
 
from kallithea.lib.vcs.backends.base import EmptyChangeset
 
from kallithea.lib.utils2 import safe_unicode
 
from kallithea.lib.graphmod import graph_data
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
def _update_with_GET(params, GET):
 
    for k in ['diff1', 'diff2', 'diff']:
 
        params[k] += GET.getall(k)
 

	
 
@@ -270,27 +269,25 @@ class ChangesetController(BaseRepoContro
 
            cs1 = changeset.parents[0].raw_id if changeset.parents else EmptyChangeset().raw_id
 
            context_lcl = get_line_ctx('', request.GET)
 
            ign_whitespace_lcl = get_ignore_ws('', request.GET)
 

	
 
            raw_diff = c.db_repo_scm_instance.get_diff(cs1, cs2,
 
                ignore_whitespace=ign_whitespace_lcl, context=context_lcl)
 
            diff_limit = None if c.fulldiff else self.cut_off_limit
 
            file_diff_data = []
 
            if method == 'show':
 
                diff_processor = diffs.DiffProcessor(raw_diff,
 
                                                     vcs=c.db_repo_scm_instance.alias,
 
                                                     diff_limit=diff_limit)
 
                c.limited_diff = False
 
                if isinstance(diff_processor.parsed, LimitedDiffContainer):
 
                    c.limited_diff = True
 
                c.limited_diff = diff_processor.limited_diff
 
                for f in diff_processor.parsed:
 
                    st = f['stats']
 
                    c.lines_added += st['added']
 
                    c.lines_deleted += st['deleted']
 
                    filename = f['filename']
 
                    fid = h.FID(changeset.raw_id, filename)
 
                    url_fid = h.FID('', filename)
 
                    diff = diff_processor.as_html(enable_comments=enable_comments,
 
                                                  parsed_lines=[f])
 
                    file_diff_data.append((fid, url_fid, f['operation'], f['old_filename'], filename, diff, st))
 
            else:
 
                # downloads/raw we only need RAW diff nothing else
kallithea/controllers/compare.py
Show inline comments
 
@@ -33,25 +33,24 @@ import re
 
from tg import request, tmpl_context as c
 
from tg.i18n import ugettext as _
 
from webob.exc import HTTPFound, HTTPBadRequest, HTTPNotFound
 

	
 
from kallithea.config.routing import url
 
from kallithea.lib.utils2 import safe_str, safe_int
 
from kallithea.lib.vcs.utils.hgcompat import unionrepo
 
from kallithea.lib import helpers as h
 
from kallithea.lib.base import BaseRepoController, render
 
from kallithea.lib.auth import LoginRequired, HasRepoPermissionLevelDecorator
 
from kallithea.lib import diffs
 
from kallithea.model.db import Repository
 
from kallithea.lib.diffs import LimitedDiffContainer
 
from kallithea.controllers.changeset import _ignorews_url, _context_url
 
from kallithea.lib.graphmod import graph_data
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class CompareController(BaseRepoController):
 

	
 
    def _before(self, *args, **kwargs):
 
        super(CompareController, self)._before(*args, **kwargs)
 

	
 
        # The base repository has already been retrieved.
 
@@ -263,29 +262,25 @@ class CompareController(BaseRepoControll
 
                raise HTTPBadRequest
 
            rev1 = c.a_rev
 

	
 
        diff_limit = None if fulldiff else self.cut_off_limit
 

	
 
        log.debug('running diff between %s and %s in %s',
 
                  rev1, c.cs_rev, org_repo.scm_instance.path)
 
        raw_diff = org_repo.scm_instance.get_diff(rev1=rev1, rev2=c.cs_rev,
 
                                      ignore_whitespace=ignore_whitespace,
 
                                      context=line_context)
 

	
 
        diff_processor = diffs.DiffProcessor(raw_diff or '', diff_limit=diff_limit)
 

	
 
        c.limited_diff = False
 
        if isinstance(diff_processor.parsed, LimitedDiffContainer):
 
            c.limited_diff = True
 

	
 
        c.limited_diff = diff_processor.limited_diff
 
        c.file_diff_data = []
 
        c.lines_added = 0
 
        c.lines_deleted = 0
 
        for f in diff_processor.parsed:
 
            st = f['stats']
 
            c.lines_added += st['added']
 
            c.lines_deleted += st['deleted']
 
            filename = f['filename']
 
            fid = h.FID('', filename)
 
            diff = diff_processor.as_html(enable_comments=False,
 
                                          parsed_lines=[f])
 
            c.file_diff_data.append((fid, None, f['operation'], f['old_filename'], filename, diff, st))
kallithea/controllers/feed.py
Show inline comments
 
@@ -29,25 +29,25 @@ Original author and date, and relevant c
 
import logging
 

	
 
from tg import response, tmpl_context as c
 
from tg.i18n import ugettext as _
 

	
 
from beaker.cache import cache_region, region_invalidate
 
from webhelpers.feedgenerator import Atom1Feed, Rss201rev2Feed
 

	
 
from kallithea import CONFIG
 
from kallithea.lib import helpers as h
 
from kallithea.lib.auth import LoginRequired, HasRepoPermissionLevelDecorator
 
from kallithea.lib.base import BaseRepoController
 
from kallithea.lib.diffs import DiffProcessor, LimitedDiffContainer
 
from kallithea.lib.diffs import DiffProcessor
 
from kallithea.model.db import CacheInvalidation
 
from kallithea.lib.utils2 import safe_int, str2bool, safe_unicode
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
language = 'en-us'
 
ttl = "5"
 

	
 

	
 
class FeedController(BaseRepoController):
 

	
 
@@ -67,35 +67,32 @@ class FeedController(BaseRepoController)
 
            desc_msg.append('branch: %s<br/>' % cs.branch)
 
        for book in cs.bookmarks:
 
            desc_msg.append('bookmark: %s<br/>' % book)
 
        for tag in cs.tags:
 
            desc_msg.append('tag: %s<br/>' % tag)
 

	
 
        changes = []
 
        diff_limit = safe_int(CONFIG.get('rss_cut_off_limit', 32 * 1024))
 
        raw_diff = cs.diff()
 
        diff_processor = DiffProcessor(raw_diff,
 
                                       diff_limit=diff_limit,
 
                                       inline_diff=False)
 
        limited_diff = False
 
        if isinstance(diff_processor.parsed, LimitedDiffContainer):
 
            limited_diff = True
 

	
 
        for st in diff_processor.parsed:
 
            st.update({'added': st['stats']['added'],
 
                       'removed': st['stats']['deleted']})
 
            changes.append('\n %(operation)s %(filename)s '
 
                           '(%(added)s lines added, %(removed)s lines removed)'
 
                            % st)
 
        if limited_diff:
 
        if diff_processor.limited_diff:
 
            changes = changes + ['\n ' +
 
                                 _('Changeset was too big and was cut off...')]
 

	
 
        # rev link
 
        _url = h.canonical_url('changeset_home', repo_name=c.db_repo.repo_name,
 
                   revision=cs.raw_id)
 
        desc_msg.append('changeset: <a href="%s">%s</a>' % (_url, cs.raw_id[:8]))
 

	
 
        desc_msg.append('<pre>')
 
        desc_msg.append(h.urlify_text(cs.message))
 
        desc_msg.append('\n')
 
        desc_msg.extend(changes)
kallithea/controllers/pullrequests.py
Show inline comments
 
@@ -30,25 +30,24 @@ import traceback
 
import formencode
 

	
 
from tg import request, tmpl_context as c
 
from tg.i18n import ugettext as _
 
from webob.exc import HTTPFound, HTTPNotFound, HTTPForbidden, HTTPBadRequest
 

	
 
from kallithea.config.routing import url
 
from kallithea.lib import helpers as h
 
from kallithea.lib import diffs
 
from kallithea.lib.auth import LoginRequired, HasRepoPermissionLevelDecorator, \
 
    NotAnonymous
 
from kallithea.lib.base import BaseRepoController, render, jsonify
 
from kallithea.lib.diffs import LimitedDiffContainer
 
from kallithea.lib.page import Page
 
from kallithea.lib.utils import action_logger
 
from kallithea.lib.vcs.exceptions import EmptyRepositoryError, ChangesetDoesNotExistError
 
from kallithea.lib.vcs.utils import safe_str
 
from kallithea.lib.vcs.utils.hgcompat import unionrepo
 
from kallithea.model.db import PullRequest, ChangesetStatus, ChangesetComment, \
 
    PullRequestReviewer, Repository, User
 
from kallithea.model.pull_request import CreatePullRequestAction, CreatePullRequestIterationAction, PullRequestModel
 
from kallithea.model.meta import Session
 
from kallithea.model.repo import RepoModel
 
from kallithea.model.comment import ChangesetCommentsModel
 
from kallithea.model.changeset_status import ChangesetStatusModel
 
@@ -589,29 +588,25 @@ class PullrequestsController(BaseRepoCon
 
        diff_limit = None if fulldiff else self.cut_off_limit
 

	
 
        # we swap org/other ref since we run a simple diff on one repo
 
        log.debug('running diff between %s and %s in %s',
 
                  c.a_rev, c.cs_rev, org_scm_instance.path)
 
        try:
 
            raw_diff = org_scm_instance.get_diff(rev1=safe_str(c.a_rev), rev2=safe_str(c.cs_rev),
 
                                                ignore_whitespace=ignore_whitespace,
 
                                                context=line_context)
 
        except ChangesetDoesNotExistError:
 
            raw_diff = _("The diff can't be shown - the PR revisions could not be found.")
 
        diff_processor = diffs.DiffProcessor(raw_diff or '', diff_limit=diff_limit)
 

	
 
        c.limited_diff = False
 
        if isinstance(diff_processor.parsed, LimitedDiffContainer):
 
            c.limited_diff = True
 

	
 
        c.limited_diff = diff_processor.limited_diff
 
        c.file_diff_data = []
 
        c.lines_added = 0
 
        c.lines_deleted = 0
 

	
 
        for f in diff_processor.parsed:
 
            st = f['stats']
 
            c.lines_added += st['added']
 
            c.lines_deleted += st['deleted']
 
            filename = f['filename']
 
            fid = h.FID('', filename)
 
            diff = diff_processor.as_html(enable_comments=True,
 
                                          parsed_lines=[f])
kallithea/lib/diffs.py
Show inline comments
 
@@ -132,36 +132,24 @@ NEW_FILENODE = 1
 
DEL_FILENODE = 2
 
MOD_FILENODE = 3
 
RENAMED_FILENODE = 4
 
COPIED_FILENODE = 5
 
CHMOD_FILENODE = 6
 
BIN_FILENODE = 7
 

	
 

	
 
class DiffLimitExceeded(Exception):
 
    pass
 

	
 

	
 
class LimitedDiffContainer(object):
 

	
 
    def __init__(self, diff_limit, cur_diff_size, diff):
 
        self.diff = diff
 
        self.diff_limit = diff_limit
 
        self.cur_diff_size = cur_diff_size
 

	
 
    def __iter__(self):
 
        for l in self.diff:
 
            yield l
 

	
 

	
 
class DiffProcessor(object):
 
    """
 
    Give it a unified or git diff and it returns a list of the files that were
 
    mentioned in the diff together with a dict of meta information that
 
    can be used to render it in a HTML template.
 
    """
 
    _diff_git_re = re.compile('^diff --git', re.MULTILINE)
 
    _chunk_re = re.compile(r'^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@(.*)')
 
    _newline_marker = re.compile(r'^\\ No newline at end of file')
 
    _git_header_re = re.compile(r"""
 
        ^diff[ ]--git
 
            [ ]a/(?P<a_path>.+?)[ ]b/(?P<b_path>.+?)\n
 
@@ -210,24 +198,25 @@ class DiffProcessor(object):
 
            based on that parameter cut off will be triggered, set to None
 
            to show full diff
 
        """
 
        if not isinstance(diff, basestring):
 
            raise Exception('Diff must be a basestring got %s instead' % type(diff))
 

	
 
        self._diff = diff
 
        self.adds = 0
 
        self.removes = 0
 
        # calculate diff size
 
        self.diff_limit = diff_limit
 
        self.cur_diff_size = 0
 
        self.limited_diff = False
 
        self.vcs = vcs
 
        self.parsed = self._parse_gitdiff(inline_diff=inline_diff)
 

	
 
    def _escaper(self, string):
 
        """
 
        Do HTML escaping/markup and check the diff limit
 
        """
 
        self.cur_diff_size += len(string)
 

	
 
        # escaper gets iterated on each .next() call and it checks if each
 
        # parsed line doesn't exceed the diff limit
 
        if self.diff_limit is not None and self.cur_diff_size > self.diff_limit:
 
@@ -297,29 +286,28 @@ class DiffProcessor(object):
 
            match = self._hg_header_re.match(diff_chunk)
 
        if match is None:
 
            raise Exception('diff not recognized as valid %s diff' % self.vcs)
 
        meta_info = match.groupdict()
 
        rest = diff_chunk[match.end():]
 
        if rest and not rest.startswith('@') and not rest.startswith('literal ') and not rest.startswith('delta '):
 
            raise Exception('cannot parse %s diff header: %r followed by %r' % (self.vcs, diff_chunk[:match.end()], rest[:1000]))
 
        diff_lines = (self._escaper(m.group(0)) for m in re.finditer(r'.*\n|.+$', rest)) # don't split on \r as str.splitlines do
 
        return meta_info, diff_lines
 

	
 
    def _parse_gitdiff(self, inline_diff):
 
        """Parse self._diff and return a list of dicts with meta info and chunks for each file.
 
        If diff is truncated, wrap it in LimitedDiffContainer.
 
        Might set limited_diff.
 
        Optionally, do an extra pass and to extra markup of one-liner changes.
 
        """
 
        _files = [] # list of dicts with meta info and chunks
 
        diff_container = lambda arg: arg
 

	
 
        starts = [m.start() for m in self._diff_git_re.finditer(self._diff)]
 
        starts.append(len(self._diff))
 

	
 
        for start, end in zip(starts, starts[1:]):
 
            head, diff_lines = self._get_header(buffer(self._diff, start, end - start))
 

	
 
            op = None
 
            stats = {
 
                'added': 0,
 
                'deleted': 0,
 
                'binary': False,
 
@@ -376,27 +364,25 @@ class DiffProcessor(object):
 
            # a real non-binary diff
 
            if head['a_file'] or head['b_file']:
 
                try:
 
                    chunks, added, deleted = self._parse_lines(diff_lines)
 
                    stats['binary'] = False
 
                    stats['added'] = added
 
                    stats['deleted'] = deleted
 
                    # explicit mark that it's a modified file
 
                    if op == 'M':
 
                        stats['ops'][MOD_FILENODE] = 'modified file'
 

	
 
                except DiffLimitExceeded:
 
                    diff_container = lambda _diff: \
 
                        LimitedDiffContainer(self.diff_limit,
 
                                            self.cur_diff_size, _diff)
 
                    self.limited_diff = True
 
                    break
 
            else:  # Git binary patch (or empty diff)
 
                # Git binary patch
 
                if head['bin_patch']:
 
                    stats['ops'][BIN_FILENODE] = 'binary diff not shown'
 
                chunks = []
 

	
 
            if op == 'D' and chunks:
 
                # a way of seeing deleted content could perhaps be nice - but
 
                # not with the current UI
 
                chunks = []
 

	
 
@@ -410,25 +396,25 @@ class DiffProcessor(object):
 

	
 
            _files.append({
 
                'old_filename':     head['a_path'],
 
                'filename':         head['b_path'],
 
                'old_revision':     head['a_blob_id'],
 
                'new_revision':     head['b_blob_id'],
 
                'chunks':           chunks,
 
                'operation':        op,
 
                'stats':            stats,
 
            })
 

	
 
        if not inline_diff:
 
            return diff_container(_files)
 
            return _files
 

	
 
        # highlight inline changes when one del is followed by one add
 
        for diff_data in _files:
 
            for chunk in diff_data['chunks']:
 
                lineiter = iter(chunk)
 
                try:
 
                    peekline = lineiter.next()
 
                    while True:
 
                        # find a first del line
 
                        while peekline['action'] != 'del':
 
                            peekline = lineiter.next()
 
                        delline = peekline
 
@@ -443,25 +429,25 @@ class DiffProcessor(object):
 
                        try:
 
                            peekline = lineiter.next()
 
                        except StopIteration:
 
                            # add was last line - ok
 
                            self._highlight_inline_diff(delline, addline)
 
                            raise
 
                        if peekline['action'] != 'add':
 
                            # there was only one add line - ok
 
                            self._highlight_inline_diff(delline, addline)
 
                except StopIteration:
 
                    pass
 

	
 
        return diff_container(_files)
 
        return _files
 

	
 
    def _parse_lines(self, diff_lines):
 
        """
 
        Given an iterator of diff body lines, parse them and return a dict per
 
        line and added/removed totals.
 
        """
 
        added = deleted = 0
 
        old_line = old_end = new_line = new_end = None
 

	
 
        try:
 
            chunks = []
 
            line = diff_lines.next()
0 comments (0 inline, 0 general)