Changeset - 82198c193c11
[Not reviewed]
default
0 1 0
Mads Kiilerich - 10 years ago 2015-06-19 18:00:42
madski@unity3d.com
helpers: improve html_escape

Repeated string replace is faster than iterating over each character in Python
and doing a dict lookup.
1 file changed with 13 insertions and 15 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/helpers.py
Show inline comments
 
@@ -44,108 +44,107 @@ from webhelpers.html.tools import auto_l
 
from webhelpers.number import format_byte_size, format_bit_size
 
from webhelpers.pylonslib import Flash as _Flash
 
from webhelpers.pylonslib.secure_form import secure_form as form, authentication_token
 
from webhelpers.text import chop_at, collapse, convert_accented_entities, \
 
    convert_misc_entities, lchop, plural, rchop, remove_formatting, \
 
    replace_whitespace, urlify, truncate, wrap_paragraphs
 
from webhelpers.date import time_ago_in_words
 
from webhelpers.paginate import Page as _Page
 
from webhelpers.html.tags import _set_input_attrs, _set_id_attr, \
 
    convert_boolean_attrs, NotGiven, _make_safe_id_component
 

	
 
from kallithea.lib.annotate import annotate_highlight
 
from kallithea.lib.utils import repo_name_slug, get_custom_lexer
 
from kallithea.lib.utils2 import str2bool, safe_unicode, safe_str, \
 
    get_changeset_safe, datetime_to_time, time_to_datetime, AttributeDict,\
 
    safe_int
 
from kallithea.lib.markup_renderer import MarkupRenderer, url_re
 
from kallithea.lib.vcs.exceptions import ChangesetDoesNotExistError
 
from kallithea.lib.vcs.backends.base import BaseChangeset, EmptyChangeset
 
from kallithea.config.conf import DATE_FORMAT, DATETIME_FORMAT
 
from kallithea.model.changeset_status import ChangesetStatusModel
 
from kallithea.model.db import URL_SEP, Permission
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
def canonical_url(*args, **kargs):
 
    '''Like url(x, qualified=True), but returns url that not only is qualified
 
    but also canonical, as configured in canonical_url'''
 
    from kallithea import CONFIG
 
    try:
 
        parts = CONFIG.get('canonical_url', '').split('://', 1)
 
        kargs['host'] = parts[1].split('/', 1)[0]
 
        kargs['protocol'] = parts[0]
 
    except IndexError:
 
        kargs['qualified'] = True
 
    return url(*args, **kargs)
 

	
 
def canonical_hostname():
 
    '''Return canonical hostname of system'''
 
    from kallithea import CONFIG
 
    try:
 
        parts = CONFIG.get('canonical_url', '').split('://', 1)
 
        return parts[1].split('/', 1)[0]
 
    except IndexError:
 
        parts = url('home', qualified=True).split('://', 1)
 
        return parts[1].split('/', 1)[0]
 

	
 
def html_escape(text, html_escape_table=None):
 
    """Produce entities within text."""
 
    if not html_escape_table:
 
        html_escape_table = {
 
            "&": "&",
 
            '"': """,
 
            "'": "'",
 
            ">": ">",
 
            "<": "&lt;",
 
        }
 
    return "".join(html_escape_table.get(c, c) for c in text)
 

	
 
def html_escape(text):
 
    """Return string with all html escaped.
 
    This is also safe for javascript in html but not necessarily correct.
 
    """
 
    return (text
 
        .replace('&', '&amp;')
 
        .replace(">", "&gt;")
 
        .replace("<", "&lt;")
 
        .replace('"', "&quot;")
 
        .replace("'", "&apos;")
 
        )
 

	
 
def shorter(text, size=20):
 
    postfix = '...'
 
    if len(text) > size:
 
        return text[:size - len(postfix)] + postfix
 
    return text
 

	
 

	
 
def _reset(name, value=None, id=NotGiven, type="reset", **attrs):
 
    """
 
    Reset button
 
    """
 
    _set_input_attrs(attrs, type, name, value)
 
    _set_id_attr(attrs, id, name)
 
    convert_boolean_attrs(attrs, ["disabled"])
 
    return HTML.input(**attrs)
 

	
 
reset = _reset
 
safeid = _make_safe_id_component
 

	
 

	
 
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
 

	
 
    :param raw_id:
 
    :param path:
 
    """
 

	
 
    return 'C-%s-%s' % (short_id(raw_id), md5(safe_str(path)).hexdigest()[:12])
 

	
 

	
 
class _GetError(object):
 
    """Get error from form_errors, and represent it as span wrapped error
 
    message
 

	
 
    :param field_name: field to fetch errors for
 
    :param form_errors: form errors dict
 
    """
 

	
 
    def __call__(self, field_name, form_errors):
 
        tmpl = """<span class="error_msg">%s</span>"""
 
        if form_errors and field_name in form_errors:
 
            return literal(tmpl % form_errors.get(field_name))
 

	
 
get_error = _GetError()
 

	
 
@@ -1275,101 +1274,100 @@ def urlify_text(text_, safe=True):
 
    def url_func(match_obj):
 
        url_full = match_obj.groups()[0]
 
        return '<a href="%(url)s">%(url)s</a>' % ({'url': url_full})
 
    _newtext = url_re.sub(url_func, text_)
 
    if safe:
 
        return literal(_newtext)
 
    return _newtext
 

	
 

	
 
def urlify_changesets(text_, repository):
 
    """
 
    Extract revision ids from changeset and make link from them
 

	
 
    :param text_:
 
    :param repository: repo name to build the URL with
 
    """
 
    from pylons import url  # doh, we need to re-import url to mock it later
 

	
 
    def url_func(match_obj):
 
        rev = match_obj.group(0)
 
        return '<a class="revision-link" href="%(url)s">%(rev)s</a>' % {
 
         'url': url('changeset_home', repo_name=repository, revision=rev),
 
         'rev': rev,
 
        }
 

	
 
    return re.sub(r'(?:^|(?<=[\s(),]))([0-9a-fA-F]{12,40})(?=$|\s|[.,:()])', url_func, text_)
 

	
 
def linkify_others(t, l):
 
    urls = re.compile(r'(\<a.*?\<\/a\>)',)
 
    links = []
 
    for e in urls.split(t):
 
        if not urls.match(e):
 
            links.append('<a class="message-link" href="%s">%s</a>' % (l, e))
 
        else:
 
            links.append(e)
 

	
 
    return ''.join(links)
 

	
 
def urlify_commit(text_, repository, link_=None):
 
    """
 
    Parses given text message and makes proper links.
 
    issues are linked to given issue-server, and rest is a changeset link
 
    if link_ is given, in other case it's a plain text
 

	
 
    :param text_:
 
    :param repository:
 
    :param link_: changeset link
 
    """
 
    def escaper(string):
 
        return string.replace('<', '&lt;').replace('>', '&gt;')
 
    newtext = html_escape(text_)
 

	
 
    # urlify changesets - extract revisions and make link out of them
 
    newtext = urlify_changesets(escaper(text_), repository)
 
    newtext = urlify_changesets(newtext, repository)
 

	
 
    # extract http/https links and make them real urls
 
    newtext = urlify_text(newtext, safe=False)
 

	
 
    newtext = urlify_issues(newtext, repository, link_)
 

	
 
    return literal(newtext)
 

	
 
def urlify_issues(newtext, repository, link_=None):
 
    from kallithea import CONFIG as conf
 

	
 
    # allow multiple issue servers to be used
 
    valid_indices = [
 
        x.group(1)
 
        for x in map(lambda x: re.match(r'issue_pat(.*)', x), conf.keys())
 
        if x and 'issue_server_link%s' % x.group(1) in conf
 
        and 'issue_prefix%s' % x.group(1) in conf
 
    ]
 

	
 
    if valid_indices:
 
        log.debug('found issue server suffixes `%s` during valuation of: %s'
 
                  % (','.join(valid_indices), newtext))
 

	
 
    for pattern_index in valid_indices:
 
        ISSUE_PATTERN = conf.get('issue_pat%s' % pattern_index)
 
        ISSUE_SERVER_LNK = conf.get('issue_server_link%s' % pattern_index)
 
        ISSUE_PREFIX = conf.get('issue_prefix%s' % pattern_index)
 

	
 
        log.debug('pattern suffix `%s` PAT:%s SERVER_LINK:%s PREFIX:%s'
 
                  % (pattern_index, ISSUE_PATTERN, ISSUE_SERVER_LNK,
 
                     ISSUE_PREFIX))
 

	
 
        URL_PAT = re.compile(r'%s' % ISSUE_PATTERN)
 

	
 
        def url_func(match_obj):
 
            pref = ''
 
            if match_obj.group().startswith(' '):
 
                pref = ' '
 

	
 
            issue_id = ''.join(match_obj.groups())
 
            issue_url = ISSUE_SERVER_LNK.replace('{id}', issue_id)
 
            if repository:
 
                issue_url = issue_url.replace('{repo}', repository)
 
                repo_name = repository.split(URL_SEP)[-1]
 
                issue_url = issue_url.replace('{repo_name}', repo_name)
 

	
 
            return (
 
                '%(pref)s<a class="%(cls)s" href="%(url)s">'
0 comments (0 inline, 0 general)