trac/wiki/macros.py
# -*- coding: utf-8 -*-
#
# Copyright (C) 2005-2023 Edgewall Software
# Copyright (C) 2005-2006 Christopher Lenz <cmlenz@gmx.de>
# All rights reserved.
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution. The terms
# are also available at https://trac.edgewall.org/wiki/TracLicense.
#
# This software consists of voluntary contributions made by many
# individuals. For the exact contribution history, see the revision
# history and logs, available at https://trac.edgewall.org/log/.
#
# Author: Christopher Lenz <cmlenz@gmx.de>
from fnmatch import fnmatchcase
from itertools import groupby
import fnmatch
import inspect
import io
import os
import re
from trac.core import *
from trac.resource import (
Resource, ResourceNotFound, get_resource_name, get_resource_summary,
get_resource_url
)
from trac.util import as_int
from trac.util.datefmt import format_date, from_utimestamp, user_time
from trac.util.html import Markup, escape, find_element, tag
from trac.util.presentation import separated
from trac.util.text import unicode_quote, to_unicode, stripws
from trac.util.translation import _, dgettext, cleandoc_, tag_
from trac.web.chrome import chrome_resource_path
from trac.wiki.api import IWikiMacroProvider, WikiSystem, parse_args
from trac.wiki.formatter import (
MacroError, OutlineFormatter, ProcessorError, extract_link, format_to_html,
format_to_oneliner, system_message
) # ProcessorError unused, but imported for plugin use.
from trac.wiki.interwiki import InterWikiMap
# TODO: should be moved in .api
class WikiMacroBase(Component):
"""Abstract base class for wiki macros and processors.
On usage error, the `MacroError` or `ProcessorError` exception should be
raised, to ensure proper display of the error message in the rendered
wiki content.
The `_description` attribute contains the macro help.
Set the `hide_from_macro_index` attribute to `True` to prevent
displaying help in the macro index (`[[MacroList]]`). If the
default value of `False` and the `_description` is empty,
"No documentation found" will be displayed.
"""
implements(IWikiMacroProvider)
abstract = True
#: A gettext domain to translate the macro description
_domain = None
#: A macro description
_description = None
#: Hide from macro index
hide_from_macro_index = False
def get_macros(self):
"""Yield the name of the macro based on the class name."""
name = self.__class__.__name__
if name.endswith('Macro'):
name = name[:-5]
yield name
def get_macro_description(self, name):
"""Return the subclass's gettext domain and macro description"""
if self.hide_from_macro_index:
return None
domain, description = self._domain, self._description
if description:
return (domain, description) if domain else description
# For pre-0.12 compatibility
# don't use inspect.getdoc() to avoid retrieving from the inheritance
# hierarchy
return inspect.cleandoc(self.__doc__) if self.__doc__ else ''
def parse_macro(self, parser, name, content):
raise NotImplementedError
def expand_macro(self, formatter, name, content, args=None):
raise NotImplementedError(
"pre-0.11 Wiki macro %s by provider %s no longer supported" %
(name, self.__class__))
class TitleIndexMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Insert an alphabetic list of all wiki pages into the output.
Accepts a prefix string as parameter: if provided, only pages with names
that start with the prefix are included in the resulting list. If this
parameter is omitted, all pages are listed. If the prefix is specified,
a second argument of value `hideprefix` can be given as well, in order
to remove that prefix from the output.
The prefix string supports the standard relative-path notation ''when
using the macro in a wiki page''. A prefix string starting with `./`
will be relative to the current page, and parent pages can be
specified using `../`.
Several named parameters can be specified:
- `format=compact`: The pages are displayed as comma-separated links.
- `format=group`: The list of pages will be structured in groups
according to common prefix. This format also supports a `min=n`
argument, where `n` is the minimal number of pages for a group.
- `format=hierarchy`: The list of pages will be structured according
to the page name path hierarchy. This format also supports a `min=n`
argument, where higher `n` flatten the display hierarchy
- `depth=n`: limit the depth of the pages to list. If set to 0,
only toplevel pages will be shown, if set to 1, only immediate
children pages will be shown, etc. If not set, or set to -1,
all pages in the hierarchy will be shown.
- `include=page1:page*2`: include only pages that match an item in the
colon-separated list of pages. If the list is empty, or if no `include`
argument is given, include all pages.
- `exclude=page1:page*2`: exclude pages that match an item in the colon-
separated list of pages.
The `include` and `exclude` lists accept shell-style patterns.
""")
SPLIT_RE = re.compile(r"(/| )")
NUM_SPLIT_RE = re.compile(r"([0-9.]+)")
def expand_macro(self, formatter, name, content):
args, kw = parse_args(content)
prefix = args[0].strip() if args else None
hideprefix = args and len(args) > 1 and args[1].strip() == 'hideprefix'
minsize = _arg_as_int(kw.get('min', 1), 'min', min=1)
minsize_group = max(minsize, 2)
depth = _arg_as_int(kw.get('depth', -1), 'depth', min=-1)
format = kw.get('format', '')
def parse_list(name):
return [inc.strip() for inc in kw.get(name, '').split(':')
if inc.strip()]
includes = parse_list('include') or ['*']
excludes = parse_list('exclude')
wiki = formatter.wiki
resource = formatter.resource
if prefix and resource and resource.realm == 'wiki':
prefix = wiki.resolve_relative_name(prefix, resource.id)
start = prefix.count('/') if prefix else 0
if hideprefix:
omitprefix = lambda page: page[len(prefix):]
else:
omitprefix = lambda page: page
pages = sorted(page for page in wiki.get_pages(prefix)
if (depth < 0 or depth >= page.count('/') - start)
and 'WIKI_VIEW' in formatter.perm('wiki', page)
and any(fnmatchcase(page, inc) for inc in includes)
and not any(fnmatchcase(page, exc) for exc in excludes))
if format == 'compact':
return tag(
separated((tag.a(wiki.format_page_name(omitprefix(p)),
href=formatter.href.wiki(p)) for p in pages),
', '))
# the function definitions for the different format styles
# the different page split formats, each corresponding to its rendering
def split_pages_group(pages):
"""Return a list of (path elements, page_name) pairs,
where path elements correspond to the page name (without prefix)
split at Camel Case word boundaries, numbers and '/'.
"""
page_paths = []
for page in pages:
path = [elt.strip() for elt in self.SPLIT_RE.split(
self.NUM_SPLIT_RE.sub(r" \1 ",
wiki.format_page_name(omitprefix(page), split=True)))]
page_paths.append(([elt for elt in path if elt], page))
return page_paths
def split_pages_hierarchy(pages):
"""Return a list of (path elements, page_name) pairs,
where path elements correspond to the page name (without prefix)
split according to the '/' hierarchy.
"""
return [(wiki.format_page_name(omitprefix(page)).split("/"), page)
for page in pages]
# the different tree structures, each corresponding to its rendering
def tree_group(entries):
"""Transform a flat list of entries into a tree structure.
`entries` is a list of `(path_elements, page_name)` pairs
Return a list organized in a tree structure, in which:
- a leaf is a page name
- a node is a `(key, nodes)` pairs, where:
- `key` is the leftmost of the path elements, common to the
grouped (path element, page_name) entries
- `nodes` is a list of nodes or leaves
"""
def keyfn(args):
elts, name = args
return elts[0] if elts else ''
groups = []
for key, grouper in groupby(entries, keyfn):
# remove key from path_elements in grouped entries for further
# grouping
grouped_entries = [(path_elements[1:], page_name)
for path_elements, page_name in grouper]
if key and len(grouped_entries) >= minsize_group:
subnodes = tree_group(sorted(grouped_entries))
if len(subnodes) == 1:
subkey, subnodes = subnodes[0]
node = (key + subkey, subnodes)
groups.append(node)
elif self.SPLIT_RE.match(key):
for elt in subnodes:
if isinstance(elt, tuple):
subkey, subnodes = elt
elt = (key + subkey, subnodes)
groups.append(elt)
else:
node = (key, subnodes)
groups.append(node)
else:
for path_elements, page_name in grouped_entries:
groups.append(page_name)
return groups
def tree_hierarchy(entries):
"""Transform a flat list of entries into a tree structure.
`entries` is a list of `(path_elements, page_name)` pairs
Return a list organized in a tree structure, in which:
- a leaf is a `(rest, page)` pair, where:
- `rest` is the rest of the path to be shown
- `page` is a page name
- a node is a `(key, nodes, page)` pair, where:
- `key` is the leftmost of the path elements, common to the
grouped (path element, page_name) entries
- `page` is a page name (if one exists for that node)
- `nodes` is a list of nodes or leaves
"""
def keyfn(args):
elts, name = args
return elts[0] if elts else ''
groups = []
for key, grouper in groupby(entries, keyfn):
grouped_entries = [e for e in grouper]
sub_entries = [e for e in grouped_entries if len(e[0]) > 1]
key_entries = [e for e in grouped_entries if len(e[0]) == 1]
key_entry = key_entries[0] if key_entries else None
key_page = key_entry[1] if key_entries else None
if key and len(sub_entries) >= minsize:
# remove key from path_elements in grouped entries for
# further grouping
sub_entries = [(path_elements[1:], page)
for path_elements, page in sub_entries]
subnodes = tree_hierarchy(sorted(sub_entries))
node = (key, key_page, subnodes)
groups.append(node)
else:
if key_entry:
groups.append(key_entry)
groups.extend(sub_entries)
return groups
# the different rendering formats
def render_group(group):
return tag.ul(
tag.li(tag(tag.strong(elt[0].strip('/')), render_group(elt[1]))
if isinstance(elt, tuple) else
tag.a(wiki.format_page_name(omitprefix(elt)),
href=formatter.href.wiki(elt)))
for elt in group)
def render_hierarchy(group):
return tag.ul(
tag.li(tag(tag.a(elt[0], href=formatter.href.wiki(elt[1]))
if elt[1] else tag(elt[0]),
render_hierarchy(elt[2]))
if len(elt) == 3 else
tag.a('/'.join(elt[0]),
href=formatter.href.wiki(elt[1])))
for elt in group)
transform = {
'group': lambda p: render_group(tree_group(split_pages_group(p))),
'hierarchy': lambda p: render_hierarchy(
tree_hierarchy(split_pages_hierarchy(p))),
}.get(format)
if transform:
titleindex = transform(pages)
else:
titleindex = tag.ul(
tag.li(tag.a(wiki.format_page_name(omitprefix(page)),
href=formatter.href.wiki(page)))
for page in pages)
return tag.div(titleindex, class_='titleindex')
class RecentChangesMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""List all pages that have recently been modified, ordered by the
time they were last modified.
This macro accepts two ordered arguments and a named argument. The named
argument can be placed in any position within the argument list.
The first parameter is a prefix string: if provided, only pages with names
that start with the prefix are included in the resulting list. If this
parameter is omitted, all pages are included in the list.
The second parameter is the maximum number of pages to include in the
list.
The `group` parameter determines how the list is presented:
`group=date` :: The pages are presented in bulleted lists that are
grouped by date (default).
`group=none` :: The pages are presented in a single bulleted list.
Tip: if you only want to specify a maximum number of entries and
don't want to filter by prefix, specify an empty first parameter,
e.g. `[[RecentChanges(,10,group=none)]]`.
""")
def expand_macro(self, formatter, name, content):
args, kw = parse_args(content)
prefix = args[0].strip() if args else None
limit = _arg_as_int(args[1].strip(), min=1) if len(args) > 1 else None
group = kw.get('group', 'date')
sql = """SELECT name, max(version) AS max_version,
max(time) AS max_time FROM wiki"""
args = []
if prefix:
with self.env.db_query as db:
sql += " WHERE name %s" % db.prefix_match()
args.append(db.prefix_match_value(prefix))
sql += " GROUP BY name ORDER BY max_time DESC"
if limit:
sql += " LIMIT %s"
args.append(limit)
entries_per_date = []
prevdate = None
for name, version, ts in self.env.db_query(sql, args):
if 'WIKI_VIEW' not in formatter.perm('wiki', name, version):
continue
req = formatter.req
date = user_time(req, format_date, from_utimestamp(ts))
if date != prevdate:
prevdate = date
entries_per_date.append((date, []))
version = int(version)
diff_href = None
if version > 1:
diff_href = formatter.href.wiki(name, action='diff',
version=version)
page_name = formatter.wiki.format_page_name(name)
entries_per_date[-1][1].append((page_name, name, version,
diff_href))
items_per_date = (
(date, (tag.li(tag.a(page, href=formatter.href.wiki(name)),
tag.small(' (', tag.a(_("diff"), href=diff_href),
')') if diff_href else None,
'\n')
for page, name, version, diff_href in entries))
for date, entries in entries_per_date)
if group == 'date':
out = ((tag.h3(date, class_='section'), tag.ul(entries))
for date, entries in items_per_date)
else:
out = tag.ul(entries for date, entries in items_per_date)
return tag.div(out, class_="wikipage")
class PageOutlineMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Display a structural outline of the current wiki page, each item in the
outline being a link to the corresponding heading.
This macro accepts four optional parameters:
* The first is a number or range that allows configuring the minimum and
maximum level of headings that should be included in the outline. For
example, specifying "1" here will result in only the top-level headings
being included in the outline. Specifying "2-3" will make the outline
include all headings of level 2 and 3, as a nested list. The default is
to include all heading levels.
* The second parameter can be used to specify a custom title (the default
is no title).
* The third parameter selects the style of the outline. This can be
either `inline` or `pullout` (the latter being the default). The
`inline` style renders the outline as normal part of the content, while
`pullout` causes the outline to be rendered in a box that is by default
floated to the right side of the other content.
* The fourth parameter specifies whether the outline is numbered or not.
It can be either `numbered` or `unnumbered` (the former being the
default). This parameter only has an effect in `inline` style.
""")
def expand_macro(self, formatter, name, content):
min_depth, max_depth = 1, 6
title = None
inline = False
numbered = True
if content:
argv = [arg.strip() for arg in content.split(',')]
if len(argv) > 0:
depth = argv[0]
if '-' in depth:
min_depth, max_depth = \
[_arg_as_int(d, min=min_depth, max=max_depth)
for d in depth.split('-', 1)]
else:
min_depth = max_depth = \
_arg_as_int(depth, min=min_depth, max=max_depth)
if len(argv) > 1:
title = argv[1].strip()
for arg in argv[2:]:
arg = arg.strip().lower()
if arg == 'inline':
inline = True
elif arg == 'unnumbered':
numbered = False
# TODO: - integrate the rest of the OutlineFormatter directly here
# - use formatter.wikidom instead of formatter.source
out = io.StringIO()
oformatter = OutlineFormatter(self.env, formatter.context)
oformatter.format(formatter.source, out, max_depth, min_depth,
shorten=not inline)
outline = Markup(out.getvalue())
if title:
outline = tag.h4(title, class_='section') + outline
if not inline:
outline = tag.div(outline, class_='wiki-toc')
elif not numbered:
outline = tag.div(outline, class_='wiki-toc-un')
return outline
class ImageMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Embed an image in wiki-formatted text.
The first argument is the file specification. The file specification may
reference attachments in three ways:
* `module:id:file`, where module can be either '''wiki''' or '''ticket''',
to refer to the attachment named ''file'' of the specified wiki page or
ticket.
* `id:file`: same as above, but id is either a ticket shorthand or a Wiki
page name.
* `file` to refer to a local attachment named 'file'. This only works from
within that wiki page or a ticket.
The file specification may also refer to:
* repository files, using the `source:file` syntax
(`source:file@rev` works also).
* files, using direct URLs: `/file` for a project-relative,
`//file` for a server-relative, or `http://server/file` for
absolute location. An InterWiki prefix may be used.
* embedded data using the
[http://tools.ietf.org/html/rfc2397 rfc2397] `data` URL scheme,
provided the URL is enclosed in quotes.
The remaining arguments are optional and allow configuring the attributes
and style of the rendered `<img>` element:
* digits and unit are interpreted as the size (ex. 120px, 25%)
for the image
* `right`, `left`, `center`, `top`, `bottom` and `middle` are interpreted
as the alignment for the image (alternatively, the first three can be
specified using `align=...` and the last three using `valign=...`)
* `link=some TracLinks...` replaces the link to the image source by the
one specified using a TracLinks. If no value is specified, the link is
simply removed.
* `inline` specifies that the content generated be an inline XHTML
element. By default, inline content is not generated, therefore images
won't be rendered in section headings and other one-line content.
* `nolink` means without link to image source (deprecated, use `link=`)
* `key=value` style are interpreted as HTML attributes or CSS style
indications for the image. Valid keys are:
* align, valign, border, width, height, alt, title, longdesc, class,
margin, margin-(left,right,top,bottom), id and usemap
* `border`, `margin`, and `margin-`* can only be a single number
(units are pixels).
* `margin` is superseded by `center` which uses auto margins
Examples:
{{{
[[Image(photo.jpg)]] # simplest
[[Image(photo.jpg, 120px)]] # with image width size
[[Image(photo.jpg, right)]] # aligned by keyword
[[Image(photo.jpg, nolink)]] # without link to source
[[Image(photo.jpg, align=right)]] # aligned by attribute
}}}
You can use an image from a wiki page, ticket or other module.
{{{
[[Image(OtherPage:foo.bmp)]] # from a wiki page
[[Image(base/sub:bar.bmp)]] # from hierarchical wiki page
[[Image(#3:baz.bmp)]] # from another ticket
[[Image(ticket:36:boo.jpg)]] # from another ticket (long form)
[[Image(source:/img/bee.jpg)]] # from the repository
[[Image(htdocs:foo/bar.png)]] # from project htdocs dir
[[Image(shared:foo/bar.png)]] # from shared htdocs dir (since 1.0.2)
}}}
''Adapted from the Image.py macro created by Shun-ichi Goto
<gotoh@taiyo.co.jp>''
""")
def is_inline(self, content):
args = [stripws(arg) for arg
in self._split_args_re.split(content or '')[1::2]]
return 'inline' in args
_split_re = r'''((?:[^%s"']|"[^"]*"|'[^']*')+)'''
_split_args_re = re.compile(_split_re % ',')
_split_filespec_re = re.compile(_split_re % ':')
# https://developer.mozilla.org/en-US/docs/Web/CSS/length
_size_re = re.compile(r'(?:[0-9]+|'
r'(?:[0-9]+(?:\.[0-9]+)?|\.[0-9]+)(?:%|ch|em|ex|ic|'
r'rem|vh|vw|vmax|vmin|vb|vi|px|cm|mm|Q|in|pc|pt))$')
_attr_re = re.compile('(align|valign|border|width|height|alt'
'|margin(?:-(?:left|right|top|bottom))?'
'|title|longdesc|class|id|usemap)=(.+)')
_quoted_re = re.compile("(?:[\"'])(.*)(?:[\"'])$")
def expand_macro(self, formatter, name, content):
args = None
if content:
content = stripws(content)
# parse arguments
# we expect the 1st argument to be a filename (filespec)
args = [stripws(arg) for arg
in self._split_args_re.split(content)[1::2]]
if not args:
return ''
# strip unicode white-spaces and ZWSPs are copied from attachments
# section (#10668)
filespec = args.pop(0)
# style information
attr = {}
style = {}
link = ''
# helper for the special case `source:`
#
from trac.versioncontrol.web_ui import BrowserModule
# FIXME: somehow use ResourceSystem.get_known_realms()
# ... or directly trac.wiki.extract_link
try:
browser_links = [res[0] for res in
BrowserModule(self.env).get_link_resolvers()]
except Exception:
browser_links = []
while args:
arg = args.pop(0)
if self._size_re.match(arg):
# 'width' keyword
attr['width'] = arg
elif arg == 'nolink':
link = None
elif arg.startswith('link='):
val = arg.split('=', 1)[1]
elt = extract_link(self.env, formatter.context, val.strip())
elt = find_element(elt, 'href')
link = None
if elt is not None:
link = elt.attrib.get('href')
elif arg in ('left', 'right'):
style['float'] = arg
elif arg == 'center':
style['margin-left'] = style['margin-right'] = 'auto'
style['display'] = 'block'
style.pop('margin', '')
elif arg in ('top', 'bottom', 'middle'):
style['vertical-align'] = arg
else:
match = self._attr_re.match(arg)
if match:
key, val = match.groups()
if (key == 'align' and
val in ('left', 'right', 'center')) or \
(key == 'valign' and
val in ('top', 'middle', 'bottom')):
args.append(val)
elif key in ('margin-top', 'margin-bottom'):
style[key] = self._arg_as_length(val, key)
elif key in ('margin', 'margin-left', 'margin-right') and \
'display' not in style:
style[key] = self._arg_as_length(val, key)
elif key == 'border':
style['border'] = '%s solid' % \
self._arg_as_length(val, key)
else:
m = self._quoted_re.search(val) # unquote "..." and '...'
if m:
val = m.group(1)
attr[str(key)] = val # will be used as a __call__ kwd
if self._quoted_re.match(filespec):
filespec = filespec.strip('\'"')
# parse filespec argument to get realm and id if contained.
parts = [i.strip('\'"')
for i in self._split_filespec_re.split(filespec)[1::2]]
realm = parts[0] if parts else None
url = raw_url = desc = None
attachment = None
interwikimap = InterWikiMap(self.env)
if realm in ('http', 'https', 'ftp', 'data'): # absolute
raw_url = url = filespec
desc = url.rsplit('?')[0]
elif realm in interwikimap:
url, desc = interwikimap.url(realm, ':'.join(parts[1:]))
raw_url = url
elif filespec.startswith('//'): # server-relative
raw_url = url = filespec[1:]
desc = url.rsplit('?')[0]
elif filespec.startswith('/'): # project-relative
params = ''
if '?' in filespec:
filespec, params = filespec.rsplit('?', 1)
url = formatter.href(filespec)
if params:
url += '?' + params
raw_url, desc = url, filespec
elif len(parts) == 3: # realm:id:attachment-filename
# # or intertrac:realm:id
realm, id, filename = parts
intertrac_target = "%s:%s" % (id, filename)
it = formatter.get_intertrac_url(realm, intertrac_target)
if it:
url, desc = it
raw_url = url + unicode_quote('?format=raw')
else:
attachment = Resource(realm, id).child('attachment', filename)
elif len(parts) == 2:
realm, filename = parts
if realm in browser_links: # source:path
# TODO: use context here as well
rev = None
if '@' in filename:
filename, rev = filename.rsplit('@', 1)
url = formatter.href.browser(filename, rev=rev)
raw_url = formatter.href.browser(filename, rev=rev,
format='raw')
desc = filespec
else: # #ticket:attachment or WikiPage:attachment
# FIXME: do something generic about shorthand forms...
realm = None
id, filename = parts
if id and id[0] == '#':
realm = 'ticket'
id = id[1:]
elif id == 'htdocs':
raw_url = url = formatter.href.chrome('site', filename)
desc = os.path.basename(filename)
elif id == 'shared':
raw_url = url = formatter.href.chrome('shared', filename)
desc = os.path.basename(filename)
else:
realm = 'wiki'
if realm:
attachment = Resource(realm, id).child('attachment',
filename)
elif len(parts) == 1: # it's an attachment of the current resource
attachment = formatter.resource.child('attachment', filespec)
else:
return system_message(_("No filespec given"))
if attachment:
try:
desc = get_resource_summary(self.env, attachment)
except ResourceNotFound:
link = None
raw_url = chrome_resource_path(formatter.context.req,
'common/attachment.png')
desc = _('No image "%(id)s" attached to %(parent)s',
id=attachment.id,
parent=get_resource_name(self.env, attachment.parent))
else:
if 'ATTACHMENT_VIEW' in formatter.perm(attachment):
url = get_resource_url(self.env, attachment,
formatter.href)
raw_url = get_resource_url(self.env, attachment,
formatter.href, format='raw')
if desc:
for key in ('title', 'alt'):
if key not in attr:
attr[key] = desc
for key in ('width', 'height'):
if key not in attr:
continue
val = attr[key]
if not self._size_re.match(val):
del attr[key]
continue
if not val.endswith('%') and not val.isdigit():
style[key] = val
del attr[key]
if style:
attr['style'] = '; '.join('%s:%s' % (k, escape(style[k]))
for k in sorted(style))
if not WikiSystem(self.env).is_safe_origin(raw_url,
formatter.context.req):
attr['crossorigin'] = 'anonymous' # avoid password prompt
result = tag.img(src=raw_url, **attr)
if link is not None:
result = tag.a(result, href=link or url,
style='padding:0; border:none')
return result
def _arg_as_length(self, val, key):
if val.isdigit():
return val + 'px'
if self._size_re.match(val):
return val
int_val = as_int(val, None, min=1)
if int_val is not None:
return '%dpx' % int_val
raise _invalid_macro_arg(val, key)
class MacroListMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Display a list of all installed Wiki macros, including documentation if
available.
Optionally, the name of a specific macro can be provided as an argument. In
that case, only the documentation for that macro will be rendered.
Note that this macro will not be able to display the documentation of
macros if the `PythonOptimize` option is enabled for mod_python!
""")
def expand_macro(self, formatter, name, content):
content = content.strip() if content else ''
name_filter = content.strip('*')
def get_macro_descr():
for macro_provider in formatter.wiki.macro_providers:
names = list(macro_provider.get_macros() or [])
if name_filter and not any(name.startswith(name_filter)
for name in names):
continue
try:
name_descriptions = [
(name, macro_provider.get_macro_description(name))
for name in names]
except Exception as e:
yield system_message(
_("Error: Can't get description for macro %(name)s",
name=names[0]), e), names
else:
for descr, pairs in groupby(name_descriptions,
key=lambda p: p[1]):
if descr:
if isinstance(descr, (tuple, list)):
descr = dgettext(descr[0],
to_unicode(descr[1])) \
if descr[1] else ''
else:
descr = to_unicode(descr) or ''
if content == '*':
descr = format_to_oneliner(
self.env, formatter.context, descr,
shorten=True)
else:
descr = format_to_html(
self.env, formatter.context, descr)
elif descr is None:
continue
yield descr, [name for name, descr in pairs]
return tag.div(class_='trac-macrolist')(
(tag.h3(tag.code('[[', names[0], ']]'), id='%s-macro' % names[0]),
len(names) > 1 and tag.p(tag.strong(_("Aliases:")),
[tag.code(' [[', alias, ']]')
for alias in names[1:]]) or None,
description or tag.em(_("No documentation found")))
for description, names in sorted(get_macro_descr(),
key=lambda item: item[1][0]))
class TracIniMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Produce documentation for the Trac configuration file.
Typically, this will be used in the TracIni page. The macro accepts
two ordered arguments and two named arguments.
The ordered arguments are a configuration section filter,
and a configuration option name filter: only the configuration
options whose section and name start with the filters are output.
The named arguments can be specified:
section :: a glob-style filtering on the section names
option :: a glob-style filtering on the option names
""")
def expand_macro(self, formatter, name, content):
from trac.config import ConfigSection, Option
args, kw = parse_args(content)
filters = {}
for name, index in (('section', 0), ('option', 1)):
pattern = kw.get(name, '').strip()
if pattern:
filters[name] = fnmatch.translate(pattern)
continue
prefix = args[index].strip() if index < len(args) else ''
if prefix:
filters[name] = re.escape(prefix)
has_option_filter = 'option' in filters
for name in ('section', 'option'):
filters[name] = re.compile(filters[name], re.IGNORECASE).match \
if name in filters \
else lambda v: True
section_filter = filters['section']
option_filter = filters['option']
section_registry = ConfigSection.get_registry(self.compmgr)
option_registry = Option.get_registry(self.compmgr)
options = {}
for (section, key), option in option_registry.items():
if section_filter(section) and option_filter(key):
options.setdefault(section, {})[key] = option
if not has_option_filter:
for section in section_registry:
if section_filter(section):
options.setdefault(section, {})
for section in options:
options[section] = sorted(iter(options[section].values()),
key=lambda option: option.name)
sections = [(section, section_registry[section].doc
if section in section_registry else '')
for section in sorted(options)]
def default_cell(option):
default = option.default
if default is not None and default != '':
return tag.td(tag.code(option.dumps(default)),
class_='default')
else:
return tag.td(_("(no default)"), class_='nodefault')
def options_table(section, options):
if options:
return tag.table(class_='wiki')(
tag.tbody(
tag.tr(
tag.td(tag.a(tag.code(option.name),
class_='tracini-option',
href='#%s-%s-option' %
(section, option.name))),
tag.td(format_to_html(self.env, formatter.context,
option.doc)),
default_cell(option),
id='%s-%s-option' % (section, option.name),
class_='odd' if idx % 2 else 'even')
for idx, option in enumerate(options)))
return tag.div(class_='tracini')(
(tag.h3(tag.code('[%s]' % section), id='%s-section' % section),
format_to_html(self.env, formatter.context, section_doc),
options_table(section, options.get(section)))
for section, section_doc in sections)
class KnownMimeTypesMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""List all known mime-types which can be used as WikiProcessors.
Can be given an optional argument which is interpreted as mime-type filter.
""")
def expand_macro(self, formatter, name, content):
from trac.mimeview.api import Mimeview
mime_map = Mimeview(self.env).mime_map
mime_type_filter = ''
args, kw = parse_args(content)
if args:
mime_type_filter = args.pop(0).strip().rstrip('*')
mime_types = {}
for key, mime_type in mime_map.items():
if (not mime_type_filter or
mime_type.startswith(mime_type_filter)) and key != mime_type:
mime_types.setdefault(mime_type, []).append(key)
return tag.div(class_='mimetypes')(
tag.table(class_='wiki')(
tag.thead(tag.tr(
tag.th(_("MIME Types")), # always use plural
tag.th(tag.a("WikiProcessors",
href=formatter.context.href.wiki(
'WikiProcessors'))))),
tag.tbody(
tag.tr(tag.th(tag.code(mime_type),
style="text-align: left"),
tag.td(tag.code(
' '.join(sorted(mime_types[mime_type])))))
for mime_type in sorted(mime_types))))
class TracGuideTocMacro(WikiMacroBase):
_domain = 'messages'
_description = cleandoc_(
"""Display a table of content for the Trac guide.
This macro shows a quick and dirty way to make a table-of-contents
for the !Help/Guide. The table of contents will contain the Trac* and
WikiFormatting pages, and can't be customized. See the
[https://trac-hacks.org/wiki/TocMacro TocMacro] for a more customizable
table of contents.
""")
TOC = [('TracGuide', 'Index'),
('TracInstall', 'Installation'),
('TracInterfaceCustomization', 'Customization'),
('TracPlugins', 'Plugins'),
('TracUpgrade', 'Upgrading'),
('TracIni', 'Configuration'),
('TracAdmin', 'Administration'),
('TracBackup', 'Backup'),
('TracLogging', 'Logging'),
('TracPermissions', 'Permissions'),
('TracWiki', 'The Wiki'),
('WikiFormatting', 'Wiki Formatting'),
('TracTimeline', 'Timeline'),
('TracBrowser', 'Repository Browser'),
('TracRevisionLog', 'Revision Log'),
('TracChangeset', 'Changesets'),
('TracTickets', 'Tickets'),
('TracWorkflow', 'Workflow'),
('TracRoadmap', 'Roadmap'),
('TracQuery', 'Ticket Queries'),
('TracBatchModify', 'Batch Modify'),
('TracReports', 'Reports'),
('TracRss', 'RSS Support'),
('TracNotification', 'Notification'),
]
def expand_macro(self, formatter, name, content):
curpage = formatter.resource.id
# scoped TOC (e.g. TranslateRu/TracGuide or 0.11/TracGuide ...)
prefix = ''
idx = curpage.find('/')
if idx > 0:
prefix = curpage[:idx+1]
ws = WikiSystem(self.env)
return tag.div(
tag.h4(_('Table of Contents')),
tag.ul([tag.li(tag.a(title, href=formatter.href.wiki(prefix+ref),
class_=(not ws.has_page(prefix+ref) and
'missing')),
class_=(prefix+ref == curpage and 'active'))
for ref, title in self.TOC]),
class_='wiki-toc')
def _arg_as_int(val, key=None, min=None, max=None):
int_val = as_int(val, None, min=min, max=max)
if int_val is None:
raise _invalid_macro_arg(val, key)
return int_val
def _invalid_macro_arg(val, key=None):
expr = tag.code("%s=%s" % (key, val)) if key else tag.code(val)
return MacroError(tag_("Invalid macro argument %(expr)s", expr=expr))