kallithea Changeset - 39a59e6915bb

Changeset - 39a59e6915bb

Parent rev.

Child rev.

[Not reviewed]

default

0 2 0

Mads Kiilerich - 9 years ago 2016-09-25 17:21:07
madski@unity3d.com

helpers: refactor and optimize urlify_issues

Avoid parsing the configuration and compiling regexps every time a string is
processed.

Instead, the configuration is parsed once and turned into a function that apply
the compiled regexps in a chain.

A next iteration can perhaps run them in parallel and integrate with the
general urlify machinery.

2 files changed with 56 insertions and 50 deletions:

kallithea/lib/helpers.py

kallithea/tests/functional/test_files.py

0 comments (0 inline, 0 general)

kallithea/lib/helpers.py

➞

Show inline comments

@@ @@ -1307,153 +1307,155 @@ def urlify_text(s, repo_name=None, link_ @@
             if tagtype:
                 tagvalue = match_obj.group('tagvalue')
                 return '<div class="metatag" tag="%s">%s =&gt; <a href="/%s">%s</a></div>' % (tagtype, tagtype, tagvalue, tagvalue)
             lang = match_obj.group('lang')
             if lang:
                 return '<div class="metatag" tag="lang">%s</div>' % lang
             tag = match_obj.group('tag')
             if tag:
                 return '<div class="metatag" tag="%s">%s</div>' % (tag, tag)
         return match_obj.group(0)
     def _urlify(s):
         """
         Extract urls from text and make html links out of them
         """
         return _URLIFY_RE.sub(_replace, s)
     if truncate is None:
         s = s.rstrip()
     else:
         s = truncatef(s, truncate, whole_word=True)
     s = html_escape(s)
     s = _urlify(s)
     if repo_name is not None:
         s = urlify_issues(s, repo_name)
     if link_ is not None:
         # make href around everything that isn't a href already
         s = linkify_others(s, link_)
     s = s.replace('\r\n', '<br/>').replace('\n', '<br/>')
     return literal(s)
 def linkify_others(t, l):
     """Add a default link to html with links.
     HTML doesn't allow nesting of links, so the outer link must be broken up
     in pieces and give space for other links.
     """
     urls = re.compile(r'(\<a.*?\<\/a\>)',)
     links = []
     for e in urls.split(t):
         if e.strip() and 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_issues_replace_f(repo_name, ISSUE_SERVER_LNK, ISSUE_PREFIX):
     def urlify_issues_replace(match_obj):
         pref = ''
         if match_obj.group().startswith(' '):
             pref = ' '
         issue_id = ''.join(match_obj.groups())
         issue_url = ISSUE_SERVER_LNK.replace('{id}', issue_id)
         issue_url = issue_url.replace('{repo}', repo_name)
         issue_url = issue_url.replace('{repo_name}', repo_name.split(URL_SEP)[-1])
         return (
             '%(pref)s<a class="%(cls)s" href="%(url)s">'
             '%(issue-prefix)s%(id-repr)s'
             '</a>'
             ) % {
              'pref': pref,
              'cls': 'issue-tracker-link',
              'url': issue_url,
              'id-repr': issue_id,
              'issue-prefix': ISSUE_PREFIX,
              'serv': ISSUE_SERVER_LNK,
+            }
     return urlify_issues_replace
 # Global variable that will hold the actual urlify_issues function body.
 # Will be set on first use when the global configuration has been read.
 _urlify_issues_f = None
 def urlify_issues(newtext, repo_name):
     from kallithea import CONFIG as conf
     """Urlify issue references according to .ini configuration"""
     global _urlify_issues_f
     if _urlify_issues_f is None:
         from kallithea import CONFIG
         assert CONFIG['sqlalchemy.url'] # make sure config has been loaded
         # Build chain of urlify functions, starting with not doing any transformation
         tmp_urlify_issues_f = lambda s: s
     # 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)
         issue_pat_re = re.compile(r'issue_pat(.*)')
         for k in CONFIG.keys():
             # Find all issue_pat* settings that also have corresponding server_link and prefix configuration
             m = issue_pat_re.match(k)
             if m is None:
                 continue
             suffix = m.group(1)
             issue_pat = CONFIG.get(k)
             issue_server_link = CONFIG.get('issue_server_link%s' % suffix)
             issue_prefix = CONFIG.get('issue_prefix%s' % suffix)
             if issue_pat and issue_server_link and issue_prefix:
                 log.debug('issue pattern %r: %r -> %r %r', suffix, issue_pat, issue_server_link, issue_prefix)
             else:
                 log.error('skipping incomplete issue pattern %r: %r -> %r %r', suffix, issue_pat, issue_server_link, issue_prefix)
                 continue
     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)
             # Wrap tmp_urlify_issues_f with substitution of this pattern, while making sure all loop variables (and compiled regexpes) are bound
             issue_re = re.compile(issue_pat)
             def issues_replace(match_obj,
                                issue_server_link=issue_server_link, issue_prefix=issue_prefix):
                 leadingspace = ' ' if match_obj.group().startswith(' ') else ''
                 issue_id = ''.join(match_obj.groups())
                 issue_url = issue_server_link.replace('{id}', issue_id)
                 issue_url = issue_url.replace('{repo}', repo_name)
                 issue_url = issue_url.replace('{repo_name}', repo_name.split(URL_SEP)[-1])
                 return (
                     '%(leadingspace)s<a class="issue-tracker-link" href="%(url)s">'
                     '%(issue-prefix)s%(id-repr)s'
                     '</a>'
                     ) % {
                      'leadingspace': leadingspace,
                      'url': issue_url,
                      'id-repr': issue_id,
                      'issue-prefix': issue_prefix,
                      'serv': issue_server_link,
+                    }
             tmp_urlify_issues_f = (lambda s,
                                           issue_re=issue_re, issues_replace=issues_replace, chain_f=tmp_urlify_issues_f:
                                    issue_re.sub(issues_replace, chain_f(s)))
         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(ISSUE_PATTERN)
         # Set tmp function globally - atomically
         _urlify_issues_f = tmp_urlify_issues_f
         urlify_issues_replace = _urlify_issues_replace_f(repo_name, ISSUE_SERVER_LNK, ISSUE_PREFIX)
         newtext = URL_PAT.sub(urlify_issues_replace, newtext)
         log.debug('processed prefix:`%s` => %s', pattern_index, newtext)
     return newtext
     return _urlify_issues_f(newtext)
 def render_w_mentions(source, repo_name=None):
     """
     Render plain text with revision hashes and issue references urlified
     and with @mention highlighting.
     """
     s = safe_unicode(source)
     s = urlify_text(s, repo_name=repo_name)
     return literal('<div class="formatted-fixed">%s</div>' % s)
 def short_ref(ref_type, ref_name):
     if ref_type == 'rev':
         return short_id(ref_name)
     return ref_name
 def link_to_ref(repo_name, ref_type, ref_name, rev=None):
     """
     Return full markup for a href to changeset_home for a changeset.
     If ref_type is branch it will link to changelog.
     ref_name is shortened if ref_type is 'rev'.
     if rev is specified show it too, explicitly linking to that revision.
     """
     txt = short_ref(ref_type, ref_name)
     if ref_type == 'branch':
         u = url('changelog_home', repo_name=repo_name, branch=ref_name)
     else:
         u = url('changeset_home', repo_name=repo_name, revision=ref_name)
     l = link_to(repo_name + '#' + txt, u)
     if rev and ref_type != 'rev':
         l = literal('%s (%s)' % (l, link_to(short_id(rev), url('changeset_home', repo_name=repo_name, revision=rev))))
     return l
 def changeset_status(repo, revision):
     return ChangesetStatusModel().get_status(repo, revision)
 def changeset_status_lbl(changeset_status):
     return ChangesetStatus.get_status_lbl(changeset_status)
 def get_permission_name(key):
     return dict(Permission.PERMS).get(key)
 def journal_filter_help():
     return _(textwrap.dedent('''

kallithea/tests/functional/test_files.py

➞

Show inline comments

@@ @@ -47,96 +47,100 @@ class TestFilesController(TestController @@
         response.mustcontain('<a class="browser-file ypjax-link" href="/%s/files/96507bd11ecc815ebc6270fdf6db110928c09c1e/setup.py"><i class="icon-doc"></i><span>setup.py</span></a>' % HG_REPO)
         response.mustcontain('<a class="browser-file ypjax-link" href="/%s/files/96507bd11ecc815ebc6270fdf6db110928c09c1e/test_and_report.sh"><i class="icon-doc"></i><span>test_and_report.sh</span></a>' % HG_REPO)
         response.mustcontain('<a class="browser-file ypjax-link" href="/%s/files/96507bd11ecc815ebc6270fdf6db110928c09c1e/tox.ini"><i class="icon-doc"></i><span>tox.ini</span></a>' % HG_REPO)
     def test_index_revision(self):
         self.log_user()
         response = self.app.get(
             url(controller='files', action='index',
                 repo_name=HG_REPO,
                 revision='7ba66bec8d6dbba14a2155be32408c435c5f4492',
                 f_path='/')
+        )
         #Test response...
         response.mustcontain('<a class="browser-dir ypjax-link" href="/%s/files/7ba66bec8d6dbba14a2155be32408c435c5f4492/docs"><i class="icon-folder-open"></i><span>docs</span></a>' % HG_REPO)
         response.mustcontain('<a class="browser-dir ypjax-link" href="/%s/files/7ba66bec8d6dbba14a2155be32408c435c5f4492/tests"><i class="icon-folder-open"></i><span>tests</span></a>' % HG_REPO)
         response.mustcontain('<a class="browser-file ypjax-link" href="/%s/files/7ba66bec8d6dbba14a2155be32408c435c5f4492/README.rst"><i class="icon-doc"></i><span>README.rst</span></a>' % HG_REPO)
         response.mustcontain('1.1 KiB')
     def test_index_different_branch(self):
         self.log_user()
         response = self.app.get(url(controller='files', action='index',
                                     repo_name=HG_REPO,
                                     revision='97e8b885c04894463c51898e14387d80c30ed1ee',
                                     f_path='/'))
         response.mustcontain("""<option selected="selected" value="97e8b885c04894463c51898e14387d80c30ed1ee">git at 97e8b885c048</option>""")
     def test_index_paging(self):
         self.log_user()
         for r in [(73, 'a066b25d5df7016b45a41b7e2a78c33b57adc235'),
                   (92, 'cc66b61b8455b264a7a8a2d8ddc80fcfc58c221e'),
                   (109, '75feb4c33e81186c87eac740cee2447330288412'),
                   (1, '3d8f361e72ab303da48d799ff1ac40d5ac37c67e'),
                   (0, 'b986218ba1c9b0d6a259fac9b050b1724ed8e545')]:
             response = self.app.get(url(controller='files', action='index',
                                     repo_name=HG_REPO,
                                     revision=r[1],
                                     f_path='/'))
             response.mustcontain("""@ r%s:%s""" % (r[0], r[1][:12]))
     def test_file_source(self):
         # Force the global cache to be populated now when we know the right .ini has been loaded.
         # (Without this, the test would fail.)
         import kallithea.lib.helpers
         kallithea.lib.helpers._urlify_issues_f = None
         self.log_user()
         response = self.app.get(url(controller='files', action='index',
                                     repo_name=HG_REPO,
                                     revision='8911406ad776fdd3d0b9932a2e89677e57405a48',
                                     f_path='vcs/nodes.py'))
         response.mustcontain("""<div class="commit">Partially implemented <a class="issue-tracker-link" href="https://issues.example.com/vcs_test_hg/issue/16">#16</a>. filecontent/commit message/author/node name are safe_unicode now.<br/>"""
 """In addition some other __str__ are unicode as well<br/>"""
 """Added test for unicode<br/>"""
 """Improved test to clone into uniq repository.<br/>"""
 """removed extra unicode conversion in diff.</div>
 """)
         response.mustcontain("""<option selected="selected" value="8911406ad776fdd3d0b9932a2e89677e57405a48">default at 8911406ad776</option>""")
     def test_file_source_history(self):
         self.log_user()
         response = self.app.get(url(controller='files', action='history',
                                     repo_name=HG_REPO,
                                     revision='tip',
                                     f_path='vcs/nodes.py'),
                                 extra_environ={'HTTP_X_PARTIAL_XHR': '1'},)
         assert response.body == HG_NODE_HISTORY
     def test_file_source_history_git(self):
         self.log_user()
         response = self.app.get(url(controller='files', action='history',
                                     repo_name=GIT_REPO,
                                     revision='master',
                                     f_path='vcs/nodes.py'),
                                 extra_environ={'HTTP_X_PARTIAL_XHR': '1'},)
         assert response.body == GIT_NODE_HISTORY
     def test_file_annotation(self):
         self.log_user()
         response = self.app.get(url(controller='files', action='index',
                                     repo_name=HG_REPO,
                                     revision='tip',
                                     f_path='vcs/nodes.py',
                                     annotate=True))
         response.mustcontain("""r356:25213a5fbb04""")
     def test_file_annotation_git(self):
         self.log_user()
         response = self.app.get(url(controller='files', action='index',
                                     repo_name=GIT_REPO,
                                     revision='master',

0 comments (0 inline, 0 general)