Changeset - 77aa8e4cd0a2
[Not reviewed]
tip stable
0 1 0
Mads Kiilerich (mads) - 2 months ago 2026-01-06 20:44:44
mads@kiilerich.com
helpers: very partial support for pygments >= 2.12

setup.py makes it clear that Kallithea only supports "Pygments >= 2.2.0, < 2.8".
Later versions are known to cause minor breakage to tests, and it hasn't been
validated that they don't expose other problems.

Pawel (aka momat) reported a problem with pygments >= 2.12 when subclassing
HtmlFormatter. The wrap outfile parameter has been dropped, and he suggested
this workaround.

The workaround seems safe and backwards compatible and nice to have, but we
still don't recommend or support using Pygments >= 2.8 .
1 file changed with 1 insertions and 1 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/helpers.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# This program is free software: you can redistribute it and/or modify
 
# it under the terms of the GNU General Public License as published by
 
# the Free Software Foundation, either version 3 of the License, or
 
# (at your option) any later version.
 
#
 
# This program is distributed in the hope that it will be useful,
 
# but WITHOUT ANY WARRANTY; without even the implied warranty of
 
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 
# GNU General Public License for more details.
 
#
 
# You should have received a copy of the GNU General Public License
 
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
"""
 
Helper functions
 

	
 
Consists of functions to typically be used within templates, but also
 
available to Controllers. This module is available to both as 'h'.
 
"""
 
import hashlib
 
import logging
 
import re
 
import textwrap
 
import urllib.parse
 

	
 
from beaker.cache import cache_region
 
from pygments import highlight as code_highlight
 
from pygments.formatters.html import HtmlFormatter
 
from tg import tmpl_context as c
 
from tg.i18n import ugettext as _
 

	
 
import kallithea
 
from kallithea.lib.annotate import annotate_highlight
 
from kallithea.lib.auth import HasPermissionAny, HasRepoGroupPermissionLevel, HasRepoPermissionLevel
 
from kallithea.lib.diffs import BIN_FILENODE, CHMOD_FILENODE, DEL_FILENODE, MOD_FILENODE, NEW_FILENODE, RENAMED_FILENODE
 
from kallithea.lib.pygmentsutils import get_custom_lexer
 
from kallithea.lib.utils2 import AttributeDict, asbool, credentials_filter, link_to_ref, safe_bytes, safe_int, safe_str, time_to_datetime
 
from kallithea.lib.vcs.backends.base import BaseChangeset, EmptyChangeset
 
from kallithea.lib.vcs.exceptions import ChangesetDoesNotExistError
 
from kallithea.lib.vcs.utils import author_email, author_name
 
from kallithea.lib.webutils import (HTML, Option, age, canonical_url, checkbox, chop_at, end_form, escape, fmt_date, form, format_byte_size, hidden, js, jshtml,
 
                                    link_to, literal, password, pop_flash_messages, radio, render_w_mentions, reset, safeid, select, session_csrf_secret_name,
 
                                    session_csrf_secret_token, shorter, submit, text, textarea, url, urlify_text, wrap_paragraphs)
 
from kallithea.model import db
 
from kallithea.model.changeset_status import ChangesetStatusModel
 

	
 

	
 
# mute pyflakes "imported but unused"
 
# from webutils
 
assert Option
 
assert age
 
assert canonical_url
 
assert checkbox
 
assert chop_at
 
assert end_form
 
assert fmt_date
 
assert form
 
assert format_byte_size
 
assert hidden
 
assert js
 
assert jshtml
 
assert password
 
assert pop_flash_messages
 
assert radio
 
assert render_w_mentions
 
assert reset
 
assert safeid
 
assert select
 
assert session_csrf_secret_name
 
assert session_csrf_secret_token
 
assert shorter
 
assert submit
 
assert text
 
assert textarea
 
assert urlify_text
 
assert wrap_paragraphs
 
# from kallithea.lib.auth
 
assert HasPermissionAny
 
assert HasRepoGroupPermissionLevel
 
assert HasRepoPermissionLevel
 
# from utils2
 
assert credentials_filter
 
assert link_to_ref
 
assert time_to_datetime
 
# from vcs
 
assert EmptyChangeset
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
def FID(raw_id, path):
 
    """
 
    Creates a unique ID for filenode based on it's hash of path and revision
 
    it's safe to use in urls
 
    """
 
    return 'C-%s-%s' % (short_id(raw_id), hashlib.md5(safe_bytes(path)).hexdigest()[:12])
 

	
 

	
 
def get_ignore_whitespace_diff(GET):
 
    """Return true if URL requested whitespace to be ignored"""
 
    return bool(GET.get('ignorews'))
 

	
 

	
 
def ignore_whitespace_link(GET, anchor=None):
 
    """Return snippet with link to current URL with whitespace ignoring toggled"""
 
    params = dict(GET)  # ignoring duplicates
 
    if get_ignore_whitespace_diff(GET):
 
        params.pop('ignorews')
 
        title = _("Show whitespace changes")
 
    else:
 
        params['ignorews'] = '1'
 
        title = _("Ignore whitespace changes")
 
    params['anchor'] = anchor
 
    return link_to(
 
        literal('<i class="icon-strike"></i>'),
 
        url.current(**params),
 
        title=title,
 
        **{'data-toggle': 'tooltip'})
 

	
 

	
 
def get_diff_context_size(GET):
 
    """Return effective context size requested in URL"""
 
    return safe_int(GET.get('context'), default=3)
 

	
 

	
 
def increase_context_link(GET, anchor=None):
 
    """Return snippet with link to current URL with double context size"""
 
    context = get_diff_context_size(GET) * 2
 
    params = dict(GET)  # ignoring duplicates
 
    params['context'] = str(context)
 
    params['anchor'] = anchor
 
    return link_to(
 
        literal('<i class="icon-sort"></i>'),
 
        url.current(**params),
 
        title=_('Increase diff context to %(num)s lines') % {'num': context},
 
        **{'data-toggle': 'tooltip'})
 

	
 

	
 
def files_breadcrumbs(repo_name, rev, paths):
 
    url_l = [link_to(repo_name, url('files_home',
 
                                    repo_name=repo_name,
 
                                    revision=rev, f_path=''),
 
                     class_='ypjax-link')]
 
    paths_l = paths.split('/')
 
    for cnt, p in enumerate(paths_l):
 
        if p != '':
 
            url_l.append(link_to(p,
 
                                 url('files_home',
 
                                     repo_name=repo_name,
 
                                     revision=rev,
 
                                     f_path='/'.join(paths_l[:cnt + 1])
 
                                     ),
 
                                 class_='ypjax-link'
 
                                 )
 
                         )
 
    return literal('/'.join(url_l))
 

	
 

	
 
class CodeHtmlFormatter(HtmlFormatter):
 
    """
 
    My code Html Formatter for source codes
 
    """
 

	
 
    def wrap(self, source, outfile):
 
    def wrap(self, source, outfile=None):
 
        return self._wrap_div(self._wrap_pre(self._wrap_code(source)))
 

	
 
    def _wrap_code(self, source):
 
        for cnt, it in enumerate(source):
 
            i, t = it
 
            t = '<span id="L%s">%s</span>' % (cnt + 1, t)
 
            yield i, t
 

	
 
    def _wrap_tablelinenos(self, inner):
 
        inner_lines = []
 
        lncount = 0
 
        for t, line in inner:
 
            if t:
 
                lncount += 1
 
            inner_lines.append(line)
 

	
 
        fl = self.linenostart
 
        mw = len(str(lncount + fl - 1))
 
        sp = self.linenospecial
 
        st = self.linenostep
 
        la = self.lineanchors
 
        aln = self.anchorlinenos
 
        nocls = self.noclasses
 
        if sp:
 
            lines = []
 

	
 
            for i in range(fl, fl + lncount):
 
                if i % st == 0:
 
                    if i % sp == 0:
 
                        if aln:
 
                            lines.append('<a href="#%s%d" class="special">%*d</a>' %
 
                                         (la, i, mw, i))
 
                        else:
 
                            lines.append('<span class="special">%*d</span>' % (mw, i))
 
                    else:
 
                        if aln:
 
                            lines.append('<a href="#%s%d">%*d</a>' % (la, i, mw, i))
 
                        else:
 
                            lines.append('%*d' % (mw, i))
 
                else:
 
                    lines.append('')
 
            ls = '\n'.join(lines)
 
        else:
 
            lines = []
 
            for i in range(fl, fl + lncount):
 
                if i % st == 0:
 
                    if aln:
 
                        lines.append('<a href="#%s%d">%*d</a>' % (la, i, mw, i))
 
                    else:
 
                        lines.append('%*d' % (mw, i))
 
                else:
 
                    lines.append('')
 
            ls = '\n'.join(lines)
 

	
 
        # in case you wonder about the seemingly redundant <div> here: since the
 
        # content in the other cell also is wrapped in a div, some browsers in
 
        # some configurations seem to mess up the formatting...
 
        if nocls:
 
            yield 0, ('<table class="%stable">' % self.cssclass +
 
                      '<tr><td><div class="linenodiv">'
 
                      '<pre>' + ls + '</pre></div></td>'
 
                      '<td id="hlcode" class="code">')
 
        else:
 
            yield 0, ('<table class="%stable">' % self.cssclass +
 
                      '<tr><td class="linenos"><div class="linenodiv">'
 
                      '<pre>' + ls + '</pre></div></td>'
 
                      '<td id="hlcode" class="code">')
 
        yield 0, ''.join(inner_lines)
 
        yield 0, '</td></tr></table>'
 

	
 

	
 
_whitespace_re = re.compile(r'(\t)|( )(?=\n|</div>)')
 

	
 

	
 
def _markup_whitespace(m):
 
    groups = m.groups()
 
    if groups[0]:
 
        return '<u>\t</u>'
 
    if groups[1]:
 
        return ' <i></i>'
 

	
 

	
 
def markup_whitespace(s):
 
    return _whitespace_re.sub(_markup_whitespace, s)
 

	
 

	
 
def pygmentize(filenode, **kwargs):
 
    """
 
    pygmentize function using pygments
 

	
 
    :param filenode:
 
    """
 
    lexer = get_custom_lexer(filenode.extension) or filenode.lexer
 
    return literal(markup_whitespace(
 
        code_highlight(safe_str(filenode.content), lexer, CodeHtmlFormatter(**kwargs))))
 

	
 

	
 
def hsv_to_rgb(h, s, v):
 
    if s == 0.0:
 
        return v, v, v
 
    i = int(h * 6.0)  # XXX assume int() truncates!
 
    f = (h * 6.0) - i
 
    p = v * (1.0 - s)
 
    q = v * (1.0 - s * f)
 
    t = v * (1.0 - s * (1.0 - f))
 
    i = i % 6
 
    if i == 0:
 
        return v, t, p
 
    if i == 1:
 
        return q, v, p
 
    if i == 2:
 
        return p, v, t
 
    if i == 3:
 
        return p, q, v
 
    if i == 4:
 
        return t, p, v
 
    if i == 5:
 
        return v, p, q
 

	
 

	
 
def gen_color(n=10000):
 
    """generator for getting n of evenly distributed colors using
 
    hsv color and golden ratio. It always return same order of colors
 

	
 
    :returns: RGB tuple
 
    """
 

	
 
    golden_ratio = 0.618033988749895
 
    h = 0.22717784590367374
 

	
 
    for _unused in range(n):
 
        h += golden_ratio
 
        h %= 1
 
        HSV_tuple = [h, 0.95, 0.95]
 
        RGB_tuple = hsv_to_rgb(*HSV_tuple)
 
        yield [str(int(x * 256)) for x in RGB_tuple]
 

	
 

	
 
def pygmentize_annotation(repo_name, filenode, **kwargs):
 
    """
 
    pygmentize function for annotation
 

	
 
    :param filenode:
 
    """
 
    cgenerator = gen_color()
 
    color_dict = {}
 

	
 
    def get_color_string(cs):
 
        if cs in color_dict:
 
            col = color_dict[cs]
 
        else:
 
            col = color_dict[cs] = next(cgenerator)
 
        return "color: rgb(%s)! important;" % (', '.join(col))
 

	
 
    def url_func(changeset):
 
        author = escape(changeset.author)
 
        date = changeset.date
 
        message = escape(changeset.message)
 
        tooltip_html = ("<b>Author:</b> %s<br/>"
 
                        "<b>Date:</b> %s</b><br/>"
 
                        "<b>Message:</b> %s") % (author, date, message)
 

	
 
        lnk_format = show_id(changeset)
 
        uri = link_to(
 
                lnk_format,
 
                url('changeset_home', repo_name=repo_name,
 
                    revision=changeset.raw_id),
 
                style=get_color_string(changeset.raw_id),
 
                **{'data-toggle': 'popover',
 
                   'data-content': tooltip_html}
 
              )
 

	
 
        uri += '\n'
 
        return uri
 

	
 
    return literal(markup_whitespace(annotate_highlight(filenode, url_func, **kwargs)))
 

	
 

	
 
def capitalize(x):
 
    return x.capitalize()
 

	
 
def short_id(x):
 
    return x[:12]
 

	
 

	
 
def show_id(cs):
 
    """
 
    Configurable function that shows ID
 
    by default it's r123:fffeeefffeee
 

	
 
    :param cs: changeset instance
 
    """
 
    def_len = safe_int(kallithea.CONFIG.get('show_sha_length', 12))
 
    show_rev = asbool(kallithea.CONFIG.get('show_revision_number', False))
 

	
 
    raw_id = cs.raw_id[:def_len]
 
    if show_rev:
 
        return 'r%s:%s' % (cs.revision, raw_id)
 
    else:
 
        return raw_id
 

	
 

	
 
@cache_region('long_term', 'user_attr_or_none')
 
def user_attr_or_none(author, show_attr):
 
    """Try to match email part of VCS committer string with a local user and return show_attr
 
    - or return None if user not found"""
 
    email = author_email(author)
 
    if email:
 
        user = db.User.get_by_email(email)
 
        if user is not None:
 
            return getattr(user, show_attr)
 
    return None
 

	
 

	
 
def email_or_none(author):
 
    """Try to match email part of VCS committer string with a local user.
 
    Return primary email of user, email part of the specified author name, or None."""
 
    if not author:
 
        return None
 
    email = user_attr_or_none(author, 'email')
 
    if email is not None:
 
        return email # always use user's main email address - not necessarily the one used to find user
 

	
 
    # extract email from the commit string
 
    email = author_email(author)
 
    if email:
 
        return email
 

	
 
    # No valid email, not a valid user in the system, none!
 
    return None
 

	
 

	
 
def person(author, show_attr="username"):
 
    """Find the user identified by 'author' string, return one of the users attributes,
 
    default to the username attribute, None if there is no user"""
 
    value = user_attr_or_none(author, show_attr)
 
    if value is not None:
 
        return value
 

	
 
    # Still nothing?  Just pass back the author name if any, else the email
 
    return author_name(author) or author_email(author)
 

	
 

	
 
def person_by_id(id_, show_attr="username"):
 
    # maybe it's an ID ?
 
    if str(id_).isdigit() or isinstance(id_, int):
 
        id_ = int(id_)
 
        user = db.User.get(id_)
 
        if user is not None:
 
            return getattr(user, show_attr)
 
    return id_
 

	
 

	
 
def boolicon(value):
 
    """Returns boolean value of a value, represented as small html image of true/false
 
    icons
 

	
 
    :param value: value
 
    """
 

	
 
    if value:
 
        return HTML.tag('i', class_="icon-ok")
 
    else:
 
        return HTML.tag('i', class_="icon-minus-circled")
 

	
 

	
 
def action_parser(user_log, feed=False, parse_cs=False):
 
    """
 
    This helper will action_map the specified string action into translated
 
    fancy names with icons and links
 

	
 
    :param user_log: user log instance
 
    :param feed: use output for feeds (no html and fancy icons)
 
    :param parse_cs: parse Changesets into VCS instances
 
    """
 

	
 
    action = user_log.action
 
    action_params = ' '
 

	
 
    x = action.split(':')
 

	
 
    if len(x) > 1:
 
        action, action_params = x
 

	
 
    def get_cs_links():
 
        revs_limit = 3  # display this amount always
 
        revs_top_limit = 50  # show upto this amount of changesets hidden
 
        revs_ids = action_params.split(',')
 
        deleted = user_log.repository is None
 
        if deleted:
 
            return ','.join(revs_ids)
 

	
 
        repo_name = user_log.repository.repo_name
 

	
 
        def lnk(rev, repo_name):
 
            lazy_cs = False
 
            title_ = None
 
            url_ = '#'
 
            if isinstance(rev, BaseChangeset) or isinstance(rev, AttributeDict):
 
                if rev.op and rev.ref_name:
 
                    if rev.op == 'delete_branch':
 
                        lbl = _('Deleted branch: %s') % rev.ref_name
 
                    elif rev.op == 'tag':
 
                        lbl = _('Created tag: %s') % rev.ref_name
 
                    else:
 
                        lbl = 'Unknown operation %s' % rev.op
 
                else:
 
                    lazy_cs = True
 
                    lbl = rev.short_id[:8]
 
                    url_ = url('changeset_home', repo_name=repo_name,
 
                               revision=rev.raw_id)
 
            else:
 
                # changeset cannot be found - it might have been stripped or removed
 
                lbl = rev[:12]
 
                title_ = _('Changeset %s not found') % lbl
 
            if parse_cs:
 
                return link_to(lbl, url_, title=title_, **{'data-toggle': 'tooltip'})
 
            return link_to(lbl, url_, class_='lazy-cs' if lazy_cs else '',
 
                           **{'data-raw_id': rev.raw_id, 'data-repo_name': repo_name})
 

	
 
        def _get_op(rev_txt):
 
            _op = None
 
            _name = rev_txt
 
            if len(rev_txt.split('=>')) == 2:
 
                _op, _name = rev_txt.split('=>')
 
            return _op, _name
 

	
 
        revs = []
 
        if len([v for v in revs_ids if v != '']) > 0:
 
            repo = None
 
            for rev in revs_ids[:revs_top_limit]:
 
                _op, _name = _get_op(rev)
 

	
 
                # we want parsed changesets, or new log store format is bad
 
                if parse_cs:
 
                    try:
 
                        if repo is None:
 
                            repo = user_log.repository.scm_instance
 
                        _rev = repo.get_changeset(rev)
 
                        revs.append(_rev)
 
                    except ChangesetDoesNotExistError:
 
                        log.error('cannot find revision %s in this repo', rev)
 
                        revs.append(rev)
 
                else:
 
                    _rev = AttributeDict({
 
                        'short_id': rev[:12],
 
                        'raw_id': rev,
 
                        'message': '',
 
                        'op': _op,
 
                        'ref_name': _name
 
                    })
 
                    revs.append(_rev)
 
        cs_links = [" " + ', '.join(
 
            [lnk(rev, repo_name) for rev in revs[:revs_limit]]
 
        )]
 
        _op1, _name1 = _get_op(revs_ids[0])
 
        _op2, _name2 = _get_op(revs_ids[-1])
 

	
 
        _rev = '%s...%s' % (_name1, _name2)
 

	
 
        compare_view = (
 
            ' <div class="compare_view" data-toggle="tooltip" title="%s">'
 
            '<a href="%s">%s</a> </div>' % (
 
                _('Show all combined changesets %s->%s') % (
 
                    revs_ids[0][:12], revs_ids[-1][:12]
 
                ),
 
                url('changeset_home', repo_name=repo_name,
 
                    revision=_rev
 
                ),
 
                _('Compare view')
 
            )
 
        )
 

	
 
        # if we have exactly one more than normally displayed
 
        # just display it, takes less space than displaying
 
        # "and 1 more revisions"
 
        if len(revs_ids) == revs_limit + 1:
 
            cs_links.append(", " + lnk(revs[revs_limit], repo_name))
 

	
 
        # hidden-by-default ones
 
        if len(revs_ids) > revs_limit + 1:
 
            uniq_id = revs_ids[0]
 
            html_tmpl = (
 
                '<span> %s <a class="show_more" id="_%s" '
0 comments (0 inline, 0 general)