Changeset - 51af759ef518
[Not reviewed]
default
0 2 0
Mads Kiilerich (mads) - 5 years ago 2021-02-16 22:56:53
mads@kiilerich.com
diffs: generalize name of DiffProcessor argument inline_diff to "html"
2 files changed with 7 insertions and 7 deletions:
0 comments (0 inline, 0 general)
kallithea/controllers/feed.py
Show inline comments
 
@@ -26,97 +26,97 @@ Original author and date, and relevant c
 
"""
 

	
 

	
 
import logging
 

	
 
from beaker.cache import cache_region
 
from tg import response
 
from tg import tmpl_context as c
 
from tg.i18n import ugettext as _
 

	
 
import kallithea
 
import kallithea.lib.helpers as h
 
from kallithea.controllers import base
 
from kallithea.lib import feeds, webutils
 
from kallithea.lib.auth import HasRepoPermissionLevelDecorator, LoginRequired
 
from kallithea.lib.diffs import DiffProcessor
 
from kallithea.lib.utils2 import asbool, safe_int, safe_str
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class FeedController(base.BaseRepoController):
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    def _before(self, *args, **kwargs):
 
        super(FeedController, self)._before(*args, **kwargs)
 

	
 
    def _get_title(self, cs):
 
        return webutils.shorter(cs.message, 160)
 

	
 
    def __get_desc(self, cs):
 
        desc_msg = [(_('%s committed on %s')
 
                     % (h.person(cs.author), webutils.fmt_date(cs.date))) + '<br/>']
 
        # branches, tags, bookmarks
 
        for branch in cs.branches:
 
            desc_msg.append('branch: %s<br/>' % 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(kallithea.CONFIG.get('rss_cut_off_limit', 32 * 1024))
 
        raw_diff = cs.diff()
 
        diff_processor = DiffProcessor(raw_diff,
 
                                       diff_limit=diff_limit,
 
                                       inline_diff=False)
 
                                       html=False)
 

	
 
        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 diff_processor.limited_diff:
 
            changes = changes + ['\n ' +
 
                                 _('Changeset was too big and was cut off...')]
 

	
 
        # rev link
 
        _url = webutils.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(webutils.urlify_text(cs.message))
 
        desc_msg.append('\n')
 
        desc_msg.extend(changes)
 
        if asbool(kallithea.CONFIG.get('rss_include_diff', False)):
 
            desc_msg.append('\n\n')
 
            desc_msg.append(safe_str(raw_diff))
 
        desc_msg.append('</pre>')
 
        return desc_msg
 

	
 
    def _feed(self, repo_name, feeder):
 
        """Produce a simple feed"""
 

	
 
        @cache_region('long_term_file', '_get_feed_from_cache')
 
        def _get_feed_from_cache(*_cache_keys):  # parameters are not really used - only as caching key
 
            header = dict(
 
                title=_('%s %s feed') % (c.site_name, repo_name),
 
                link=webutils.canonical_url('summary_home', repo_name=repo_name),
 
                description=_('Changes on %s repository') % repo_name,
 
            )
 

	
 
            rss_items_per_page = safe_int(kallithea.CONFIG.get('rss_items_per_page', 20))
 
            entries=[]
 
            for cs in reversed(list(c.db_repo_scm_instance[-rss_items_per_page:])):
 
                entries.append(dict(
 
                    title=self._get_title(cs),
 
                    link=webutils.canonical_url('changeset_home', repo_name=repo_name, revision=cs.raw_id),
 
                    author_email=cs.author_email,
 
                    author_name=cs.author_name,
 
                    description=''.join(self.__get_desc(cs)),
 
                    pubdate=cs.date,
 
                ))
kallithea/lib/diffs.py
Show inline comments
 
@@ -145,196 +145,196 @@ def as_html(parsed_lines, table_class='c
 
                    'cc': code_class,
 
                })
 
                _html.append('''\n\t\t<div class="add-bubble"><div>&nbsp;</div></div><pre>%(code)s</pre>\n''' % {
 
                    'code': change['line']
 
                })
 

	
 
                _html.append('''\t</td>''')
 
                _html.append('''\n</tr>\n''')
 
    _html.append('''</table>''')
 
    if _html_empty:
 
        return None
 
    return ''.join(_html)
 

	
 

	
 
def wrap_to_table(html):
 
    """Given a string with html, return it wrapped in a table, similar to what
 
    as_html returns."""
 
    return '''\
 
              <table class="code-difftable">
 
                <tr class="line">
 
                <td class="lineno new"></td>
 
                <td class="code"><pre>%s</pre></td>
 
                </tr>
 
              </table>''' % html
 

	
 

	
 
def html_diff(filenode_old, filenode_new, diff_limit=None,
 
                ignore_whitespace=True, line_context=3):
 
    """
 
    Returns a file diff as HTML wrapped into a table.
 
    Checks for diff_limit and presents a message if the diff is too big.
 
    """
 
    if filenode_old is None:
 
        filenode_old = FileNode(filenode_new.path, '', EmptyChangeset())
 

	
 
    op = None
 
    a_path = filenode_old.path # default, might be overriden by actual rename in diff
 
    if filenode_old.is_binary or filenode_new.is_binary:
 
        html_diff = wrap_to_table(_('Binary file'))
 
        stats = (0, 0)
 

	
 
    elif diff_limit != -1 and (
 
            diff_limit is None or
 
            (filenode_old.size < diff_limit and filenode_new.size < diff_limit)):
 

	
 
        raw_diff = get_gitdiff(filenode_old, filenode_new,
 
                                ignore_whitespace=ignore_whitespace,
 
                                context=line_context)
 
        diff_processor = DiffProcessor(raw_diff)
 
        diff_processor = DiffProcessor(raw_diff, html=True)
 
        if diff_processor.parsed: # there should be exactly one element, for the specified file
 
            f = diff_processor.parsed[0]
 
            op = f['operation']
 
            a_path = f['old_filename']
 

	
 
        html_diff = as_html(parsed_lines=diff_processor.parsed)
 
        stats = diff_processor.stat()
 

	
 
    else:
 
        html_diff = wrap_to_table(_('Changeset was too big and was cut off, use '
 
                               'diff menu to display this diff'))
 
        stats = (0, 0)
 

	
 
    if not html_diff:
 
        submodules = [o for o in [filenode_new, filenode_old] if isinstance(o, SubModuleNode)]
 
        if submodules:
 
            html_diff = wrap_to_table(webutils.escape('Submodule %r' % submodules[0]))
 
        else:
 
            html_diff = wrap_to_table(_('No changes detected'))
 

	
 
    cs1 = filenode_old.changeset.raw_id
 
    cs2 = filenode_new.changeset.raw_id
 

	
 
    return cs1, cs2, a_path, html_diff, stats, op
 

	
 

	
 
def get_gitdiff(filenode_old, filenode_new, ignore_whitespace=True, context=3):
 
    """
 
    Returns git style diff between given ``filenode_old`` and ``filenode_new``.
 
    """
 
    # make sure we pass in default context
 
    context = context or 3
 
    submodules = [o for o in [filenode_new, filenode_old] if isinstance(o, SubModuleNode)]
 
    if submodules:
 
        return b''
 

	
 
    for filenode in (filenode_old, filenode_new):
 
        if not isinstance(filenode, FileNode):
 
            raise VCSError("Given object should be FileNode object, not %s"
 
                % filenode.__class__)
 

	
 
    repo = filenode_new.changeset.repository
 
    old_raw_id = getattr(filenode_old.changeset, 'raw_id', repo.EMPTY_CHANGESET)
 
    new_raw_id = getattr(filenode_new.changeset, 'raw_id', repo.EMPTY_CHANGESET)
 

	
 
    vcs_gitdiff = get_diff(repo, old_raw_id, new_raw_id, filenode_new.path,
 
                           ignore_whitespace, context)
 
    return vcs_gitdiff
 

	
 

	
 
def get_diff(scm_instance, rev1, rev2, path=None, ignore_whitespace=False, context=3):
 
    """
 
    A thin wrapper around vcs lib get_diff.
 
    """
 
    try:
 
        return scm_instance.get_diff(rev1, rev2, path=path,
 
                                     ignore_whitespace=ignore_whitespace, context=context)
 
    except MemoryError:
 
        webutils.flash('MemoryError: Diff is too big', category='error')
 
        return b''
 

	
 

	
 
NEW_FILENODE = 1
 
DEL_FILENODE = 2
 
MOD_FILENODE = 3
 
RENAMED_FILENODE = 4
 
COPIED_FILENODE = 5
 
CHMOD_FILENODE = 6
 
BIN_FILENODE = 7
 

	
 

	
 
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.
 
    can be used to render it in a HTML template or as text.
 
    """
 
    _diff_git_re = re.compile(b'^diff --git', re.MULTILINE)
 

	
 
    def __init__(self, diff, vcs='hg', diff_limit=None, inline_diff=True):
 
    def __init__(self, diff, vcs='hg', diff_limit=None, html=True):
 
        """
 
        :param diff:   a text in diff format
 
        :param vcs: type of version control hg or git
 
        :param diff_limit: define the size of diff that is considered "big"
 
            based on that parameter cut off will be triggered, set to None
 
            to show full diff
 
        """
 
        if not isinstance(diff, bytes):
 
            raise Exception('Diff must be bytes - got %s' % type(diff))
 

	
 
        self._diff = memoryview(diff)
 
        self.adds = 0
 
        self.removes = 0
 
        self.diff_limit = diff_limit
 
        self.limited_diff = False
 
        self.vcs = vcs
 
        self.parsed = self._parse_gitdiff(inline_diff=inline_diff)
 
        self.parsed = self._parse_gitdiff(html=html)
 

	
 
    def _parse_gitdiff(self, inline_diff):
 
    def _parse_gitdiff(self, html):
 
        """Parse self._diff and return a list of dicts with meta info and chunks for each file.
 
        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
 

	
 
        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:]):
 
            if self.diff_limit and end > self.diff_limit:
 
                self.limited_diff = True
 
                continue
 

	
 
            head, diff_lines = _get_header(self.vcs, self._diff[start:end])
 

	
 
            op = None
 
            stats = {
 
                'added': 0,
 
                'deleted': 0,
 
                'binary': False,
 
                'ops': {},
 
            }
 

	
 
            if head['deleted_file_mode']:
 
                op = 'removed'
 
                stats['binary'] = True
 
                stats['ops'][DEL_FILENODE] = 'deleted file'
 

	
 
            elif head['new_file_mode']:
 
                op = 'added'
 
                stats['binary'] = True
 
                stats['ops'][NEW_FILENODE] = 'new file %s' % head['new_file_mode']
 
            else:  # modify operation, can be cp, rename, chmod
 
                # CHMOD
 
                if head['new_mode'] and head['old_mode']:
 
                    op = 'modified'
 
                    stats['binary'] = True
 
                    stats['ops'][CHMOD_FILENODE] = ('modified file chmod %s => %s'
 
                                        % (head['old_mode'], head['new_mode']))
 
                # RENAME
 
                if (head['rename_from'] and head['rename_to']
 
                      and head['rename_from'] != head['rename_to']):
 
                    op = 'renamed'
 
                    stats['binary'] = True
 
                    stats['ops'][RENAMED_FILENODE] = ('file renamed from %s to %s'
 
                                    % (head['rename_from'], head['rename_to']))
 
                # COPY
 
@@ -354,97 +354,97 @@ class DiffProcessor(object):
 
                        op = 'removed'
 
                        stats['binary'] = True
 
                        stats['ops'][DEL_FILENODE] = 'deleted file'
 

	
 
                # it's not ADD not DELETE
 
                if op is None:
 
                    op = 'modified'
 
                    stats['binary'] = True
 
                    stats['ops'][MOD_FILENODE] = 'modified file'
 

	
 
            # a real non-binary diff
 
            if head['a_file'] or head['b_file']:
 
                chunks, added, deleted = _parse_lines(diff_lines)
 
                stats['binary'] = False
 
                stats['added'] = added
 
                stats['deleted'] = deleted
 
                # explicit mark that it's a modified file
 
                if op == 'modified':
 
                    stats['ops'][MOD_FILENODE] = 'modified file'
 
            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 == 'removed' and chunks:
 
                # a way of seeing deleted content could perhaps be nice - but
 
                # not with the current UI
 
                chunks = []
 

	
 
            chunks.insert(0, [{
 
                'old_lineno': '',
 
                'new_lineno': '',
 
                'action':     'context',
 
                'line':       msg,
 
                } for _op, msg in stats['ops'].items()
 
                  if _op not in [MOD_FILENODE]])
 

	
 
            _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:
 
        if not html:
 
            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 = next(lineiter)
 
                    while True:
 
                        # find a first del line
 
                        while peekline['action'] != 'del':
 
                            peekline = next(lineiter)
 
                        delline = peekline
 
                        peekline = next(lineiter)
 
                        # if not followed by add, eat all following del lines
 
                        if peekline['action'] != 'add':
 
                            while peekline['action'] == 'del':
 
                                peekline = next(lineiter)
 
                            continue
 
                        # found an add - make sure it is the only one
 
                        addline = peekline
 
                        try:
 
                            peekline = next(lineiter)
 
                        except StopIteration:
 
                            # add was last line - ok
 
                            _highlight_inline_diff(delline, addline)
 
                            raise
 
                        if peekline['action'] != 'add':
 
                            # there was only one add line - ok
 
                            _highlight_inline_diff(delline, addline)
 
                except StopIteration:
 
                    pass
 

	
 
        return _files
 

	
 
    def stat(self):
 
        """
 
        Returns tuple of added, and removed lines for this instance
 
        """
 
        return self.adds, self.removes
 

	
 

	
 
_escape_re = re.compile(r'(&)|(<)|(>)|(\t)($)?|(\r)|( $)')
 

	
 

	
 
def _escaper(diff_line):
 
    r"""
 
    Do HTML escaping/markup of a single diff line (including first +/- column)
0 comments (0 inline, 0 general)