def check_status_codes(soup, realpath, res, location):
allowed_statuses = MCDPManualConstants.allowed_statuses
for h in all_headers(soup):
if MCDPManualConstants.ATTR_NOTOC in h.attrs:
continue
if MCDPManualConstants.ATTR_STATUS in h.attrs:
s = h.attrs[MCDPManualConstants.ATTR_STATUS]
if not s in allowed_statuses:
msg = 'Invalid status code %r.\n I expected one of: %s' % (
s, ", ".join(allowed_statuses))
# msg += '\n' + indent(str(h), ' ')
res.note_error(msg, HTMLIDLocation.for_element(h, location))
else:
# Only warn for h1 that are not part:
if h.name == 'h1' and not 'part:' in h.attrs.get('id', ''):
if not 'catkin_ws' in realpath:
# let's not worry about the Software repo for now
h2 = h.__copy__()
h2.attrs.pop('github-blob-url', None)
h2.attrs.pop(MCDPManualConstants.ATTR_GITHUB_EDIT_URL,
None)
msg = 'Status not found for this header:\n\n %s' % str(h2)
msg += '\n\n in file %s' % realpath
msg += '\n\nPlease set the status for all the top-level headers.'
msg += '\n\nThe syntax is:\n\n # My section {#SECTIONID status=STATUS}'
msg += '\n\nThese are the possible choices for the status:\n'
for k, v in allowed_statuses.items():
if k != MCDPManualConstants.STATUS_UNKNOWN:
msg += '\n' + indent(v, '', '%23s ' %
('status=%s' % k))
res.note_error(msg,
HTMLIDLocation.for_element(h, location))
h.attrs[MCDPManualConstants.
ATTR_STATUS] = MCDPManualConstants.STATUS_UNKNOWN
def detect_duplicate_IDs(soup, res):
from mcdp_docs.manual_join_imp import can_ignore_duplicated_id
id2element = OrderedDict()
for element in soup.select('[id]'):
ID = element.attrs['id']
if ID in id2element:
if can_ignore_duplicated_id(element):
continue
else:
# ignore this because it will be triggered for the sister element
# e.g. fig:howto-mount-motors-video, fig:howto-mount-motors-video-wrap
if ID.endswith('-wrap'):
continue
msg = 'Repeated use of ID "%s"' % ID
element.attrs['id'] = ID + '-duplicate-%s' % id(element)
locations = OrderedDict()
locations['repeated-use'] = HTMLIDLocation.for_element(element)
locations['original-use'] = HTMLIDLocation.for_element(id2element[ID])
res.note_error(msg, locations)
else:
id2element[ID] = element
def embed_img_data(soup, resolve, raise_on_error, res, location, embed=True):
"""
resolve: ref -> str or None --- how to get the data
if embed = True, embeds the data. Expects resolve to return the data.
if embed = False, just resolves the links. Expects resolve to return the path.
"""
img_extensions = MCDPManualConstants.embed_img_data_extensions
for tag in soup.select('img[src]'):
href = tag['src']
if href.startswith('data:'):
continue
if href.startswith('http'):
msg = 'I will not embed remote files, such as\n %s' % href
res.note_warning(msg, HTMLIDLocation.for_element(tag, location))
continue
for ext in img_extensions:
if not href.endswith('.' + ext):
continue
data = resolve(href)
if data is None:
msg = 'embed_img_data: Could not find file:\n %s' % href
if raise_on_error:
raise Exception(msg) # XXX
else:
res.note_error(msg,
HTMLIDLocation.for_element(tag, location))
continue
file_data = data['data']
realpath = data['realpath']
check_isinstance(data, dict)
if embed:
tag['src'] = data_encoded_for_src(file_data, ext)
else:
if not os.path.exists(realpath):
msg = 'Expecting a path from %r, but does not exist %s' % (
href, realpath)
raise Exception(msg) # XXX
max_width = MCDPManualConstants.max_width_for_image
try:
tag['src'] = get_link_to_image_file(realpath, max_width)
except IOError as e:
msg = 'Could not resize %s:' % realpath
msg += '\n\n ' + str(e)
res.note_error(msg,
locations=HTMLIDLocation.for_element(
tag, location))
break
def fix_header_id(header, globally_unique_id_part, res, location):
ID = header.get('id', None)
prefix = None if (ID is None or ':' not in ID) else ID[:ID.index(':')]
if header.name in MCDPManualConstants.allowed_prefixes_h:
allowed_prefixes = MCDPManualConstants.allowed_prefixes_h[header.name]
default_prefix = allowed_prefixes[0]
if ID is None:
main = '%s-%s' % (globally_unique_id_part, GlobalCounter.header_id)
use = '%s:%s' % (default_prefix, main)
header.attrs['id'] = use
header.attrs['id-short'] = main
header.attrs['id-autogenerated'] = "true"
GlobalCounter.header_id += 1
else:
if prefix is None:
if ID != 'booktitle': # XXX
# msg = ('Adding prefix %r to current id %r for %s.' %
# (default_prefix, ID, header.name))
# header.insert_before(Comment('Warning: ' + msg))
header.attrs['id'] = default_prefix + ':' + ID
header.attrs['id-short'] = ID
else:
if prefix not in allowed_prefixes:
msg = ('The prefix %r is not allowed for %s (ID=%r)' %
(prefix, header.name, ID))
# logger.error(msg) # TODO: add warning
# header.insert_after(Comment('Error: ' + msg))
res.note_error(
msg, HTMLIDLocation.for_element(header, location))
else:
ID_short = ID.replace(prefix + ':', '')
header.attrs['id-short'] = ID_short
def create_notes_from_elements(soup, res, location, unique):
for klass, tag in MCDPManualConstants.classes_that_create_notes.items():
markers = list(soup.select('.%s' % klass))
# print('Found %d markers for class %s' % (len(markers), klass))
for p in markers:
div = Tag(name='div')
s = Tag(name='p')
s.append('The following was marked as "%s".' % klass)
div.append(s)
div2 = Tag(name='div')
div2.attrs[
'style'] = 'margin: 1em; font-size: 90%; background-color: #eee; border-radius: 5px; padding: 0.5em;'
# Copy:
p2 = get_sanitized_copy(p)
div2.append(p2)
div.append(div2)
tags = [tag]
from mcdp_docs.manual_join_imp import split_robustly
assignees = split_robustly(p.attrs.get("for", ""), ',')
for a in assignees:
tags.append('for:%s' % a)
note = Note(div, HTMLIDLocation.for_element(p, location, unique=unique),
stacklevel=0, tags=tuple(sorted(tags)))
res.add_note(note)
def substitute_todo(soup, res, location):
prefix = "TODO"
klass = 'todo'
for r in get_elements_starting_with_string(soup, prefix=prefix):
# r.ns.replaceWith(r.rest)
div = Tag(name='div')
add_class(div, klass + '-wrap')
add_class(r.element, klass)
parent = r.element.parent
i = parent.index(r.element)
r.element.extract()
div.append(r.element)
parent.insert(i, div)
T = 'for'
if r.rest.strip().startswith(T):
after = r.rest[r.rest.index(T) + len(T):]
if ':' in after:
i = after.index(':')
dest = after[:i]
r.element.attrs['for'] = dest.strip()
else:
msg = 'Could not find ":" in "%s"' % after
res.note_error(msg, HTMLIDLocation.for_element(div, location))
def check_lang_codes(soup, res, location):
for h in all_headers(soup):
if MCDPManualConstants.LANG_ATTR in h.attrs:
s = h.attrs[MCDPManualConstants.LANG_ATTR]
if not s in MCDPManualConstants.allowed_langs:
msg = ('Invalid lang code %r; expected one of %r' %
(s, MCDPManualConstants.allowed_langs))
msg += '\n' + indent(str(h), ' ')
# note_error2(h, 'syntax error', msg)
res.note_error(msg, HTMLIDLocation.for_element(h, location))
def go(s0, func):
selectors = s0.split(',')
for selector_ in selectors:
for tag in soup.select(selector_):
try:
r = func(tag)
tag.replaceWith(r)
except (DPSyntaxError, DPSemanticError) as e:
if raise_error_dp:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
continue
except Exception as e:
if raise_error_others:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
continue
def go(selector, plotter, load, parse):
for tag in soup.select(selector):
try:
# load value with units in vu
def parsing(source_code):
context = Context()
return parse(source_code,
realpath=realpath,
context=context)
from mcdp_docs.highlight import load_or_parse_from_tag
vu = load_or_parse_from_tag(tag, load, parsing)
rendered = plotter(tag, vu)
if tag.has_attr('style'):
style = tag['style']
else:
style = ''
if style:
rendered['style'] = style
tag.replaceWith(rendered)
except (DPSyntaxError, DPSemanticError) as e:
if raise_errors:
raise
else:
msg = str(e)
res.note_error(msg,
HTMLIDLocation.for_element(tag, location))
# note_error(tag, e)
except Exception as e:
if raise_errors:
raise
else:
msg = str(e)
res.note_error(msg,
HTMLIDLocation.for_element(tag, location))
def get_empty_links_to_fragment(element_to_modify, extra_refs, res):
"""
Find all empty links that have a reference to a fragment.
yield LinkElement
"""
# logger.debug('building index')
# first find all elements by id
id2element_local, duplicates = get_id2element(element_to_modify, 'id')
id2element_extra, duplicates2 = get_id2element(extra_refs, 'id')
for k in id2element_extra:
if k in id2element_local:
if 'ignore_if_conflict' in id2element_extra[k].attrs:
continue
msg = 'ID %s in cross references also contained locally.' % k
def cut(x):
if len(x) < 500:
return x
else:
return x[:500] + ' ... '
msg += '\n\n' + indent(cut(id2element_local[k]), '', 'local: ')
msg += '\n\n' + indent(cut(id2element_extra[k]), '', 'crossrefs: ')
res.note_error(msg,
HTMLIDLocation.for_element(id2element_local[k]))
logger.error(msg)
id2element = {}
id2element.update(id2element_extra)
id2element.update(id2element_local)
# logger.debug('building index done')
for element in get_empty_links(element_to_modify):
if not 'href' in element.attrs:
continue
href = element.attrs['href']
if not href.startswith('#'):
continue
rest = href[1:]
eid = rest
query = None
linked = id2element.get(eid, None)
# noinspection PyArgumentList
yield LinkElement(linker=element, eid=eid, linked=linked, query=query)
def process_assignment(soup, res, location):
sep = ','
for e in soup.select('.assignment'):
try:
parent = find_first_parent_section(e)
except ValueError:
msg = 'Could not find parent section for this annotation.'
res.note_error(msg, HTMLIDLocation.for_element(e, location))
continue
current = split_robustly(parent.attrs.get(ATTR_ASSIGNMENT, ''), sep)
more = split_robustly(e.string, sep)
now = current + more
parent.attrs[ATTR_ASSIGNMENT] = sep.join(now)
fix_notes_assignees(soup, res)
def substitute_github_ref(a, defaults, res, location):
href = a.attrs['href']
try:
ref = parse_github_file_ref(href)
except InvalidGithubRef as e:
msg = 'Could not parse a reference in %s.' % str(a)
msg += '\n\n' + indent(e, ' > ')
res.note_warning(msg, HTMLIDLocation.for_element(a, location))
return
# raise_wrapped(DPSyntaxError, e, msg, compact=True)
if ref.url in FailedRepos.failed_repos:
msg = 'Skipped because checkout of %s already failed.' % ref.url
res.note_warning(msg, HTMLIDLocation.for_element(a, location))
return
if ref.path is None:
msg = 'There is no path specified.'
res.note_warning(msg, HTMLIDLocation.for_element(a, location))
return
# raise_desc(DPSyntaxError, e, msg, ref=ref)
try:
ref = resolve_reference(ref, defaults)
except CouldNotResolveRef as e:
res.note_error(str(e), HTMLIDLocation.for_element(a, location))
FailedRepos.failed_repos[ref.url] = str(e)
# logger.debug(ref.url)
return
a.attrs['href'] = ref.url
if not list(a.children):
c = Tag(name='code')
add_class(c, 'github-resource-link')
c.append(os.path.basename(ref.path))
a.append(c)
def check_no_patently_wrong_links(soup, res, location):
for a in soup.select('a[href]'):
href = a.attrs['href']
if href.startswith('#http:') or href.startswith('#https:'):
msg = """
This link is invalid:
URL = %s
I think there is an extra "#" at the beginning.
Note that the Markdown syntax is:
[description](URL)
where URL can be:
1) using the fragment notation, such as
URL = '#SECTIONID'
for example:
Look at [the section](#section-name)
2) a regular URL, such as:
URL = 'http://google.com'
that is:
Look at [the website](http://google.com)
You have mixed the two syntaxes.
You probably meant to write the url
%s
but you added an extra "#" at the beginning that should have not been there.
Please remove the "#".
""" % (href, href[1:])
# note_error2(a, 'syntax error', )
res.note_error(msg.lstrip(),
HTMLIDLocation.for_element(a, location))
def add_person_links(soup, users, res):
if not MCDPManualConstants.add_person_links:
return
for span in soup.select('span.person-name'):
name = span.text
try:
k = find_user_by_name(users, name)
span.name = 'a'
span.attrs['href'] = users[k]['user_url']
except KeyError:
msg = u'Could not find user "%s" in DB.' % name
res.note_warning(msg.encode('utf8'),
HTMLIDLocation.for_element(span))
def display_files(soup, defaults, res, location, raise_errors):
n = 0
for element in soup.find_all('display-file'):
src = element.attrs.get('src', '').strip()
element.attrs['src'] = src
if src.startswith('github:'):
display_file(element, defaults, res, location, raise_errors)
n += 1
else:
msg = 'Unknown schema %r; I only know "github:".' % src
if raise_errors:
raise DPSemanticError(msg)
else:
res.note_error(msg,
HTMLIDLocation.for_element(element, location))
# note_error2(element, 'syntax error', msg)
return n
def display_file(element, defaults, res, location, raise_errors):
assert element.name == 'display-file'
assert 'src' in element.attrs
src = element.attrs['src']
assert src.startswith('github:')
ref = parse_github_file_ref(src)
try:
ref = resolve_reference(ref, defaults=defaults)
except CouldNotResolveRef as e:
msg = 'Could not resolve reference %r' % src
if raise_errors:
raise_wrapped(DPSemanticError, e, msg, compact=True)
else:
msg += '\n\n' + indent(str(e), '> ')
# note_error2(element, 'reference error', msg)
res.note_error(msg, HTMLIDLocation.for_element(element, location))
return
lines = ref.contents.split('\n')
a = ref.from_line if ref.from_line is not None else 0
b = ref.to_line if ref.to_line is not None else len(lines) - 1
portion = lines[a:b + 1]
contents = "\n".join(portion)
div = Tag(name='div')
base = os.path.basename(ref.path)
short = base + '-%d-%d' % (a, b)
div.attrs['figure-id'] = 'code:%s' % short
figcaption = Tag(name='figcaption')
t = Tag(name='code')
t.append(base)
a = Tag(name='a')
a.append(t)
a.attrs['href'] = ref.url
figcaption.append(a)
div.append(figcaption)
pre = Tag(name='pre')
code = Tag(name='code')
pre.append(code)
code.append(contents)
div.append(pre)
element.replace_with(div)
def move_things_around(soup, raise_if_errors=False, res=None):
"""
Looks for tags like:
<move-here src="#line_detector2-line_detector_node2-autogenerated"/>
"""
if res is None:
res = AugmentedResult()
from mcdp_docs.check_missing_links import get_id2element
with timeit_wall('getting all IDs'):
id2element, duplicates = get_id2element(soup, 'id')
for e in soup.find_all('move-here'):
if not 'src' in e.attrs:
msg = 'Expected attribute "src" for element %s' % str(e)
raise ValueError(msg)
src = e.attrs['src']
if not src.startswith('#'):
msg = 'Expected that attribute "src" started with "#" for element %s.' % str(e)
raise ValueError(msg)
nid = src[1:]
# O(n^2)
# el = soup.find(id=nid)
el = id2element.get(nid, None)
if not el:
msg = 'move-here: Could not find ID %r.' % nid
e.name = 'span'
# note_error2(e, "invalid move-here reference", msg)
res.note_error(msg, HTMLIDLocation.for_element(e))
if raise_if_errors:
raise ValueError(msg)
else:
continue
el.extract()
e.replace_with(el)
def make_videos_(o, res, location, raise_on_errors):
if 'src' not in o.attrs:
msg = 'The video does not have a "src" attribute.'
res.note_error(msg, HTMLIDLocation.for_element(o, location))
return
# raise_desc(ValueError, msg, element=str(o))
src = o.attrs['src']
prefix = 'vimeo:'
if not src.startswith(prefix):
msg = 'Invalid src attribute "%s": it does not start with %r.' % (
src, prefix)
res.note_error(msg, HTMLIDLocation.for_element(o, location))
return
# raise_desc(ValueError, msg, element=str(o))
vimeo_id = src[len(prefix):]
# <iframe src="https://player.vimeo.com/video/152233002"
# class="embed-responsive-item"
# frameborder="0" webkitallowfullscreen="" mozallowfullscreen="" allowfullscreen="">
try:
vimeo_info = get_vimeo_info(vimeo_id)
except VimeoInfoException as e:
if raise_on_errors:
raise
else:
msg = str(e)
# note_error2(o, 'Resource error', str(e))
res.note_error(msg, HTMLIDLocation.for_element(o, location))
return
d = Tag(name='div')
d.attrs['class'] = 'video'
ONLY_WEB = 'only-web'
ONLY_EBOOK = 'only-ebook'
ONLY_DEADTREE = 'only-deadtree'
d.append(Comment('This is the iframe, for online playing.'))
C = Tag(name='div')
C.attrs['class'] = ONLY_WEB
if True:
r = Tag(name='iframe')
r.attrs['class'] = 'video-vimeo-player'
r.attrs['src'] = 'https://player.vimeo.com/video/' + vimeo_id
r.attrs['frameborder'] = 0
r.attrs['webkitallowfullscreen'] = 1
r.attrs['mozallowfullscreen'] = 1
r.attrs['allowfullscreen'] = 1
C.append(r)
d.append(C)
d.append(Comment('This is the thumbnail, for ebook'))
C = Tag(name='div')
C.attrs['class'] = ONLY_EBOOK
if True:
a = Tag(name='a')
a.attrs['href'] = vimeo_info.url
img = Tag(name='img')
img.attrs['class'] = 'video-vimeo-thumbnail-ebook'
img.attrs['src'] = vimeo_info.thumbnail_large
img.attrs['title'] = vimeo_info.title
a.append(img)
C.append(a)
d.append(C)
d.append(Comment('This is the textual version for printing.'))
C = Tag(name='div')
C.attrs['class'] = ONLY_DEADTREE
if True:
img = Tag(name='img')
img.attrs['class'] = 'video-vimeo-thumbnail-deadtree'
img.attrs['src'] = vimeo_info.thumbnail_large
img.attrs['title'] = vimeo_info.title
C.append(img)
p = Tag(name='p')
p.append("The video is at %s." % vimeo_info.url)
C.append(p)
d.append(C)
for att in ['style']:
if att in o.attrs:
d.attrs[att] = o.attrs[att]
o.replace_with(d)
def make_figures(library, soup, res, location, raise_error_dp,
raise_error_others, realpath, generate_pdf):
""" Looks for codes like:
<pre><code class="mcdp_ndp_graph_templatized">mcdp {
# empty model
}
</code></pre>
and creates a link to the image
"""
def go(s0, func):
selectors = s0.split(',')
for selector_ in selectors:
for tag in soup.select(selector_):
try:
r = func(tag)
tag.replaceWith(r)
except (DPSyntaxError, DPSemanticError) as e:
if raise_error_dp:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
continue
except Exception as e:
if raise_error_others:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
continue
def make_tag(tag0, klass, data, ndp=None, template=None, poset=None):
svg = data['svg']
tag_svg = BeautifulSoup(svg, 'lxml', from_encoding='utf-8').svg
assert tag_svg.name == 'svg'
if tag_svg.has_attr('width'):
ws = tag_svg['width']
hs = tag_svg['height']
assert 'pt' in ws
w = float(ws.replace('pt', ''))
h = float(hs.replace('pt', ''))
scale = MCDPConstants.scale_svg
w2 = w * scale
h2 = h * scale
tag_svg['width'] = w2
tag_svg['height'] = h2
tag_svg['rescaled'] = 'Rescaled from %s %s, scale = %s' % (ws, hs,
scale)
else:
print('no width in SVG tag: %s' % tag_svg)
tag_svg['class'] = klass
if tag0.has_attr('style'):
tag_svg['style'] = tag0['style']
if tag0.has_attr('id'):
tag_svg['id'] = tag0['id']
if generate_pdf:
pdf0 = data['pdf']
pdf = crop_pdf(pdf0, margins=0)
div = Tag(name='div')
att = MCDPConstants.ATTR_LOAD_NAME
if tag0.has_attr('id'):
basename = tag0['id']
elif ndp is not None and hasattr(ndp, att):
basename = getattr(ndp, att)
elif template is not None and hasattr(template, att):
basename = getattr(template, att)
elif poset is not None and hasattr(poset, att):
basename = getattr(poset, att)
else:
hashcode = hashlib.sha224(tag0.string).hexdigest()[-8:]
basename = 'code-%s' % hashcode
docname = os.path.splitext(os.path.basename(realpath))[0]
download = docname + "." + basename + "." + klass + '.pdf'
a = create_a_to_data(download=download,
data_format='pdf',
data=pdf)
a['class'] = 'pdf_data'
a.append(NavigableString(download))
div.append(tag_svg)
div.append(a)
return div
else:
return tag_svg
image_source = ImagesFromPaths(library.get_images_paths())
mf0 = MakeFiguresNDP(None, None, None)
available_ndp = set(mf0.available()) | set(mf0.aliases)
for which in available_ndp:
def callback(tag0):
assert tag0.parent is not None
context = Context()
load = lambda x: library.load_ndp(x, context=context)
parse = lambda x: library.parse_ndp(
x, realpath=realpath, context=context)
ndp = load_or_parse_from_tag(tag0, load, parse)
mf = MakeFiguresNDP(ndp=ndp,
image_source=image_source,
yourname=None) # XXX
formats = ['svg']
if generate_pdf:
formats.append('pdf')
data = mf.get_figure(which, formats)
tag = make_tag(tag0, which, data, ndp=ndp, template=None)
return tag
selector = 'render.%s,pre.%s,img.%s' % (which, which, which)
go(selector, callback)
mf0 = MakeFiguresTemplate(None, None, None)
available_template = set(mf0.available()) | set(mf0.aliases)
for which in available_template:
def callback(tag0):
context = Context()
load = lambda x: library.load_spec(
SPEC_TEMPLATES, x, context=context)
parse = lambda x: library.parse_template(
x, realpath=realpath, context=context)
template = load_or_parse_from_tag(tag0, load, parse)
mf = MakeFiguresTemplate(template=template,
library=library,
yourname=None) # XXX
formats = ['svg']
if generate_pdf:
formats.append('pdf')
data = mf.get_figure(which, formats)
tag = make_tag(tag0, which, data, ndp=None, template=template)
return tag
selector = 'render.%s,pre.%s,img.%s' % (which, which, which)
go(selector, callback)
mf0 = MakeFiguresPoset(None, None)
available_poset = set(mf0.available()) | set(mf0.aliases)
for which in available_poset:
def callback(tag0):
context = Context()
load = lambda x: library.load_poset(x, context=context)
parse = lambda x: library.parse_poset(
x, realpath=realpath, context=context)
poset = load_or_parse_from_tag(tag0, load, parse)
mf = MakeFiguresPoset(poset=poset, image_source=image_source)
formats = ['svg']
if generate_pdf:
formats.append('pdf')
data = mf.get_figure(which, formats)
tag = make_tag(tag0,
which,
data,
ndp=None,
template=None,
poset=poset)
return tag
selector = 'render.%s,pre.%s,img.%s' % (which, which, which)
go(selector, callback)
unsure = list(soup.select('render'))
unsure = [_ for _ in unsure if 'errored' not in _.attrs.get('class', '')]
for _ in unsure:
msg = 'Invalid "render" element.'
# msg += '\n\n' + '\n\n'.join(str(_) for _ in unsure)
msg += '\n\n' + " Available for NDPs: %s." % ", ".join(
sorted(available_ndp))
msg += '\n\n' + " Available for templates: %s." % ", ".join(
sorted(available_template))
msg += '\n\n' + " Available for posets: %s." % ", ".join(
sorted(available_poset))
# raise ValueError(msg)
res.note_error(msg, HTMLIDLocation.for_element(_))
return to_html_stripping_fragment(soup)
def go(soup, towrap, ID, figure_class, res, location):
from mcdp_docs.highlight import add_class
parent = towrap.parent
fig = Tag(name='figure')
fig['id'] = ID
caption_below = True
if ID.startswith('fig:'):
add_class(fig, 'figure')
elif ID.startswith('subfig:'):
add_class(fig, 'subfloat')
elif ID.startswith('tab:'):
add_class(fig, 'table')
caption_below = False
elif ID.startswith('code:'):
add_class(fig, 'code')
pass
else:
msg = 'The ID %r should start with fig: or tab: or code:' % ID
res.note_error(msg,
locations=HTMLIDLocation.for_element(towrap, location))
return
if 'caption-left' in figure_class:
caption_below = False
external_caption_id = '%s:caption' % ID
external_caption = soup.find(id=external_caption_id)
if external_caption is None:
external_caption = towrap.find(name='figcaption')
if external_caption is not None:
# print('using external caption %s' % str(external_caption))
external_caption.extract()
if external_caption.name != 'figcaption':
logger.error('Element %s#%r should have name figcaption.' %
(external_caption.name, external_caption_id))
external_caption.name = 'figcaption'
figcaption = external_caption
if towrap.has_attr('figure-caption'):
msg = 'Already using external caption for %s' % ID
res.note_error(msg,
location=HTMLIDLocation.for_element(
towrap, location))
return
else:
# print('could not find external caption %s' % external_caption_id)
if towrap.has_attr('figure-caption'):
caption = towrap['figure-caption']
else:
caption = ''
figcaption = Tag(name='figcaption')
figcaption.append(NavigableString(caption))
outside = Tag(name='div')
outside['id'] = ID + '-wrap'
if towrap.has_attr('figure-style'):
outside['style'] = towrap['figure-style']
for k in figure_class:
# logger.debug('figure-class: %s' % k)
add_class(towrap, k)
## XXX but not to figure itself?
add_class(fig, k)
add_class(outside, k)
i = parent.index(towrap)
towrap.extract()
figcontent = Tag(name='div', attrs={'class': 'figcontent'})
if towrap.name == 'figure':
towrap.name = 'div'
add_class(towrap, 'figure-conv-to-div')
figcontent.append(towrap)
# <div style='clear: both;'></div> <!-- for floating stuff-->
# Not 100% where it should go
breaking_div = Tag(name='div')
breaking_div.attrs['style'] = 'clear: both'
figcontent.append(breaking_div)
fig.append(figcontent)
if caption_below:
fig.append(figcaption)
else:
fig.insert(0, figcaption)
add_class(outside, 'generated-figure-wrap')
add_class(fig, 'generated-figure')
outside.append(fig)
parent.insert(i, outside)
def embed_pdf_image(tag, resolve, density, raise_on_error, res, location):
assert tag.name == 'img'
assert tag.has_attr('src')
# print('!!embedding %s' % str(tag))
# raise Exception(str(tag))
# load pdf data
src = tag['src']
if src.startswith('http'):
msg = 'I will not embed remote files, such as %s: ' % src
logger.warning(msg)
found = resolve(src)
if found is None:
msg = 'Could not find PDF file %r.' % src
if raise_on_error:
raise Exception(msg) # xxx
else:
# note_error2(tag, 'Resource error', msg, ['missing-image'])
res.note_error(msg, HTMLIDLocation.for_element(tag, location))
return
data_pdf = found['data']
_realpath = found['realpath']
# convert PDF to PNG
# density = pixels per inch
try:
data_png = png_from_pdf(data_pdf, density=density)
except ConversionError as e:
msg = 'I was not able to convert the PDF "%s" to PNG.' % tag['src']
if raise_on_error:
raise_wrapped(ConversionError, e, msg, compact=True)
else:
# note_error2(tag, 'Conversion error', msg, [])
res.note_error(msg, HTMLIDLocation.for_element(tag, location))
return
# get PNG image size in pixels
width_px, height_px = get_pixel_width_height_of_png(data_png)
# compute what was the original width of PDF in points
width_in = width_px / float(density)
height_in = height_px / float(density)
latex_options = tag.get('latex-options', '')
props = parse_includegraphics_option_string(latex_options)
if 'height' in props:
msg = ('Cannot deal with "height" yet: latex_options = %s' %
latex_options)
res.note_warning(msg, HTMLIDLocation.for_element(tag, location))
if 'scale' in props:
scale = float(props['scale'])
use_width_in = width_in * scale
use_height_in = height_in * scale
elif 'width' in props:
try:
use_width_in = get_length_in_inches(props['width'])
except ValueError as e:
logger.error('Cannot interpret %s: %s' % (latex_options, e))
use_width_in = 5.0
ratio = height_in / width_in
use_height_in = use_width_in * ratio
else:
use_width_in = width_in
use_height_in = height_in
# Add it before so that we can override
add_style(tag,
after=False,
width='%sin' % use_width_in,
height='%sin' % use_height_in)
tag['size_in_pixels'] = '%s, %s' % (width_px, height_px)
# encode
tag['src'] = data_encoded_for_src(data_png, 'png')
def sub_link(a, element_id, element, res):
"""
a: the link with href= #element_id
element: the link to which we refer
"""
assert isinstance(element, Tag)
CLASS_ONLY_NUMBER = MCDPManualConstants.CLASS_ONLY_NUMBER
CLASS_NUMBER_NAME = MCDPManualConstants.CLASS_NUMBER_NAME
CLASS_ONLY_NAME = MCDPManualConstants.CLASS_ONLY_NAME
if MCDPManualConstants.ATTR_NONUMBER in element.attrs:
label_what_number = None
label_number = None
try:
label_what = element.attrs[LABEL_WHAT]
label_name = element.attrs[LABEL_NAME]
except KeyError as e:
msg = 'Cannot find %r in %s' % (e, element.attrs)
raise Exception(msg) # XXX
classes = [CLASS_ONLY_NAME]
else:
if (not LABEL_WHAT_NUMBER in element.attrs) or \
(not LABEL_NAME in element.attrs):
msg = (
'substituting_empty_links: Could not find attributes %s or %s in %s'
% (LABEL_NAME, LABEL_WHAT_NUMBER, compact_desc_tag(element)))
res.note_error(
msg, {
'original': HTMLIDLocation(element_id),
'reference': HTMLIDLocation.for_element(a)
})
return
label_what_number = element.attrs[LABEL_WHAT_NUMBER]
label_number = element.attrs[LABEL_NUMBER]
label_what = element.attrs[LABEL_WHAT]
label_name = element.attrs[LABEL_NAME]
classes = list(a.attrs.get('class', [])) # bug: I was modifying
if MCDPManualConstants.CLASS_TOC_LINK in classes:
if not CLASS_ONLY_NAME in classes:
s = Tag(name='span')
s.string = label_what
add_class(s, 'toc_what')
a.append(s)
a.append(' ')
s = Tag(name='span')
s.string = label_number
add_class(s, 'toc_number')
a.append(s)
s = Tag(name='span')
s.string = ' - '
add_class(s, 'toc_sep')
a.append(s)
if label_name is not None and '<' in label_name:
contents = bs(label_name)
# sanitize the label name
for br in contents.findAll('br'):
br.replaceWith(NavigableString(' '))
for _ in contents.findAll('a'):
_.extract()
contents.name = 'span'
add_class(contents, 'toc_name')
a.append(contents)
# logger.debug('From label_name = %r to a = %r' % (label_name, a))
else:
if label_name is None:
s = Tag(name='span')
s.string = '(unnamed)' # XXX
else:
s = bs(label_name)
assert s.name == 'fragment'
s.name = 'span'
# add_class(s, 'produced-here') # XXX
add_class(s, 'toc_name')
a.append(s)
else:
if CLASS_ONLY_NUMBER in classes:
label = label_number
elif CLASS_NUMBER_NAME in classes:
if label_name is None:
label = label_what_number + \
' - ' + '(unnamed)' # warning
else:
label = label_what_number + ' - ' + label_name
elif CLASS_ONLY_NAME in classes:
if label_name is None:
label = '(unnamed)' # warning
else:
label = label_name
else:
# default behavior
if string_starts_with(['fig:', 'tab:', 'bib:', 'code:'],
element_id):
label = label_what_number
elif label_name is None:
label = label_what_number
else:
label = label_what_number + ' - ' + label_name
frag = bs(label)
assert frag.name == 'fragment'
frag.name = 'span'
add_class(frag, 'reflabel')
a.append(frag)
if 'base_url' in element.attrs:
a['href'] = element.attrs['base_url'] + a['href']
def go(selector, parse_expr, extension, use_pre=True, refine=None):
for tag in soup.select(selector):
source_code = '<unset>' # XXX
try:
if tag.string is None: # or not tag.string.strip():
if not tag.has_attr('id'):
msg = "If <pre> is empty then it needs to have an id."
raise_desc(ValueError, msg, tag=describe_tag(tag))
# load it
tag_id = tag['id'].encode('utf-8')
if '.' in tag_id:
i = tag_id.index('.')
libname, name = tag_id[:i], tag_id[i + 1:]
use_library = library.load_library(libname)
else:
name = tag_id
use_library = library
basename = '%s.%s' % (name, extension)
data = use_library._get_file_data(basename)
source_code = data['data']
else:
source_code = get_source_code(tag)
# prettify.
# remove spurious indentation
source_code = source_code.strip()
do_apply_suggestions = (not tag.has_attr('noprettify')
and not tag.has_attr('np'))
# then apply suggestions
try:
if do_apply_suggestions:
x = parse_wrap(parse_expr, source_code)[0]
xr = parse_ndp_refine(x, Context())
suggestions = get_suggestions(xr)
source_code = apply_suggestions(
source_code, suggestions)
except DPSyntaxError as e:
if raise_errors:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
continue
# we don't want the browser to choose different tab size
# source_code = source_code.replace('\t', ' ' * 4)
# we are not using it
_realpath = realpath
context = Context()
def postprocess(x):
if refine is not None:
return refine(x, context=context)
else:
return x
# print('rendering source code %r' % source_code)
html = ast_to_html(source_code,
parse_expr=parse_expr,
add_line_gutter=False,
postprocess=postprocess)
for w in context.warnings:
if w.where is not None:
from mcdp_web.editor_fancy.app_editor_fancy_generic import html_mark
html = html_mark(html, w.where, "language_warning")
frag2 = BeautifulSoup(html, 'lxml', from_encoding='utf-8')
if use_pre:
rendered = Tag(name='div', attrs={'class': 'rendered'})
pre = frag2.pre
pre.extract()
rendered.append(pre)
if not rendered.has_attr('class'):
rendered['class'] = ""
if tag.has_attr('label'):
text = tag['label']
tag_label = Tag(name='span')
add_class(tag_label, 'label')
add_class(tag_label, 'label_inside')
tag_label.append(NavigableString(text))
pre.insert(0, tag_label)
tag_label_outside = Tag(name='span')
add_class(tag_label_outside, 'label')
add_class(tag_label_outside, 'label_outside')
tag_label_outside.append(NavigableString(text))
rendered.insert(0, tag_label_outside)
max_len = max_len_of_pre_html(html)
if tag.has_attr('label'):
add_class(rendered, 'has_label')
max_len = max(max_len, len(tag['label']) + 6)
style = ''
else:
# using <code>
rendered = frag2.pre.code
rendered.extract()
if not rendered.has_attr('class'):
rendered['class'] = ""
style = ''
if tag.has_attr('style'):
style = style + tag['style']
if style:
rendered['style'] = style
if tag.has_attr('class'):
add_class(rendered, tag['class'])
if tag.has_attr('id'):
rendered['id'] = tag['id']
if use_pre:
if generate_pdf:
pdf = get_ast_as_pdf(source_code, parse_expr)
if tag.has_attr('id'):
basename = tag['id']
else:
hashcode = hashlib.sha224(
source_code).hexdigest()[-8:]
basename = 'code-%s' % hashcode
docname = os.path.splitext(
os.path.basename(realpath))[0]
download = docname + '.' + basename + '.source_code.pdf'
a = create_a_to_data(download=download,
data_format='pdf',
data=pdf)
a['class'] = 'pdf_data'
a.append(NavigableString(download))
div = Tag(name='div')
div.append(rendered)
div.append(a)
tag.replaceWith(div)
else:
tag.replaceWith(rendered)
else:
tag.replaceWith(rendered)
except DPSyntaxError as e:
if raise_errors:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
# note_error(tag, e)
if tag.string is None:
tag.string = "`%s" % tag['id']
continue
except DPSemanticError as e:
if raise_errors:
raise
else:
res.note_error(str(e), HTMLIDLocation.for_element(tag))
# note_error(tag, e)
if tag.string is None:
tag.string = "`%s" % tag['id']
continue
except DPInternalError as ex:
msg = 'Error while interpreting the code:\n\n'
msg += indent(source_code, ' | ')
raise_wrapped(DPInternalError, ex, msg, exc=sys.exc_info())
def check_if_any_href_is_invalid(soup, res, location0, extra_refs=None,
ignore_ref_errors=False):
"""
Checks if references are invalid and tries to correct them.
also works the magic
"""
if extra_refs is None:
extra_refs = Tag(name='div')
else:
print('using extra cross refs')
# let's first find all the IDs
id2element_current, duplicates = get_id2element(soup, 'id')
id2element_extra, _ = get_id2element(extra_refs, 'id')
id2element = {}
id2element.update(id2element_extra)
id2element.update(id2element_current)
for a in soup.select('[href^="#"]'):
href = a['href']
assert href.startswith('#')
ID = href[1:]
if a.has_attr('class') and "mjx-svg-href" in a['class']:
msg = 'Invalid math reference (sorry, no details): href = %s .' % href
location = HTMLIDLocation.for_element(a, location0)
res.note_error(msg, location)
continue
if ID not in id2element:
# try to fix it
# if there is already a prefix, remove it
if ':' in href:
i = href.index(':')
core = href[i + 1:]
else:
core = ID
possible = MCDPManualConstants.all_possible_prefixes_that_can_be_implied
matches = []
others = []
for possible_prefix in possible:
why_not = possible_prefix + ':' + core
others.append(why_not)
if why_not in id2element:
matches.append(why_not)
if len(matches) > 1:
msg = '%s not found, and multiple matches for heuristics (%s)' % (href, matches)
location = HTMLIDLocation.for_element(a, location0)
res.note_error(msg, location)
elif len(matches) == 1:
# if 'base_url' in element.attrs:
# a['href'] = element.attrs['base_url'] + '#' + matches[0]
# else:
a.attrs['href'] = '#' + matches[0]
if matches[0] not in id2element_current:
element = id2element[matches[0]]
# msg = 'Using foreign resolve for %s -> %s' % (matches[0], a['href'])
# logger.info(msg)
a.attrs['href_external'] = element.attrs['base_url'] + '#' + matches[0]
if show_debug_message_for_corrected_links:
msg = '%s not found, but corrected in %s' % (href, matches[0])
location = HTMLIDLocation.for_element(a, location0)
res.note_warning(msg, location)
else:
if has_class(a, MCDPConstants.CLASS_IGNORE_IF_NOT_EXISTENT):
del a.attrs['href']
# logger.warning('ignoring link %s' % a)
else:
msg = 'I do not know what is indicated by the link %r.' % href
marker = Tag(name='span')
marker.attrs['class'] = 'inside-unknown-link'
marker.append(' (unknown ref %s)' % core)
a.append(marker)
location = HTMLIDLocation.for_element(a, location0)
if ignore_ref_errors:
msg2 = 'I will ignore this error because this is the first pass:'******'\n\n' + indent(msg, ' > ')
res.note_warning(msg2, location)
else:
res.note_error(msg, location)
if ID in duplicates:
msg = 'More than one element matching %r.' % href
location = HTMLIDLocation.for_element(a, location0)
res.note_error(msg, location)
def number_items2(root, res):
counters = set(MCDPManualConstants.counters)
# TODO: make configurable
# style = get_style_book()
style = get_style_duckietown()
resets = style.resets
labels = style.labels
for c in counters:
assert c in resets, c
assert c in labels, c
from collections import defaultdict
counter_parents = defaultdict(lambda: set())
for c, cc in resets.items():
for x in cc:
counter_parents[x].add(c)
counter_state = {}
for counter in counters:
counter_state[counter] = 0
for item in root.depth_first_descendants():
counter = item.id.split(":")[0]
nonumber = MCDPManualConstants.ATTR_NONUMBER in item.tag.attrs
# print('counter %s id %s %s' % (counter, item.id, counter_state))
if counter in counters:
if not nonumber:
counter_state[counter] += 1
for counter_to_reset in resets[counter]:
counter_state[counter_to_reset] = 0
label_spec = labels[counter]
what = label_spec.what
number = render(label_spec.number, counter_state)
if LABEL_NAME in item.tag.attrs:
pass
# msg = "Don't overwrite %s for element = %s" % (LABEL_NAME, item.tag.attrs[LABEL_NAME])
# logger.warn(msg)
else:
item.tag.attrs[LABEL_NAME] = item.name
if nonumber:
item.tag.attrs[LABEL_WHAT] = what
item.tag.attrs[
LABEL_SELF] = item.name # ??? render(label_spec.label_self, counter_state)
item.tag.attrs[LABEL_WHAT_NUMBER_NAME] = item.name
# item.tag.attrs[LABEL_WHAT_NUMBER] = None
# item.tag.attrs[LABEL_NUMBER] = None
else:
item.tag.attrs[LABEL_WHAT] = what
item.tag.attrs[LABEL_SELF] = render(label_spec.label_self,
counter_state)
if ZERO in item.tag.attrs[LABEL_SELF]:
msg = 'This has zero counter.'
res.note_error(msg, HTMLIDLocation.for_element(item.tag))
if item.name is None:
item.tag.attrs[
LABEL_WHAT_NUMBER_NAME] = what + ' ' + number
else:
item.tag.attrs[
LABEL_WHAT_NUMBER_NAME] = what + ' ' + number + ' - ' + item.name
item.tag.attrs[LABEL_WHAT_NUMBER] = what + ' ' + number
item.tag.attrs[LABEL_NUMBER] = number
allattrs = [
LABEL_NAME, LABEL_WHAT, LABEL_WHAT_NUMBER_NAME, LABEL_NUMBER,
LABEL_SELF
]
for c in counters:
if c in counter_parents[counter] or c == counter:
attname = 'counter-%s' % c
allattrs.append(attname)
item.tag.attrs[attname] = counter_state[c]
if item.tag.name == 'figure':
# also copy to the caption
for figcaption in item.tag.findAll(['figcaption']):
if figcaption.parent != item.tag:
continue
for x in allattrs:
figcaption.attrs[x] = item.tag.attrs[x]
def substituting_empty_links(soup,
raise_errors=False,
res=None,
extra_refs=None):
"""
soup: where to look for references
elemtn_to_modify: what to modify (if None, it is equal to soup)
default style is [](#sec:systems) "Chapter 10"
You can also use "class":
<a href='#sec:name' class='only_number'></a>
"""
if extra_refs is None:
extra_refs = Tag(name='div')
if res is None:
res = AugmentedResult()
for le in get_empty_links_to_fragment(soup, extra_refs=extra_refs,
res=res):
a = le.linker
element_id = le.eid
element = le.linked
if not element:
msg = ('Cannot find %s' % element_id)
res.note_error(msg, HTMLIDLocation.for_element(a))
if raise_errors:
raise ValueError(msg)
continue
sub_link(a, element_id, element, res)
for a in get_empty_links(soup):
href = a.attrs.get('href', '(not present)')
if not href:
href = '""'
if href.startswith('python:'):
continue
if href.startswith('http:') or href.startswith('https:'):
msg = """
This link text is empty:
ELEMENT
Note that the syntax for links in Markdown is
[link text](URL)
For the internal links (where URL starts with "#"), then the documentation
system can fill in the title automatically, leading to the format:
[](#other-section)
However, this does not work for external sites, such as:
[](MYURL)
So, you need to provide some text, such as:
[this useful website](MYURL)
"""
msg = msg.replace('ELEMENT', str(a))
msg = msg.replace('MYURL', href)
# note_error2(a, 'syntax error', msg.strip())
res.note_error(msg, HTMLIDLocation.for_element(a))
else:
msg = """
This link is empty:
ELEMENT
It might be that the writer intended for this
link to point to something, but they got the syntax wrong.
href = %s
As a reminder, to refer to other parts of the document, use
the syntax "#ID", such as:
See [](#fig:my-figure).
See [](#section-name).
""" % href
msg = msg.replace('ELEMENT', str(a))
# note_error2(a, 'syntax error', msg.strip())
res.note_error(msg, HTMLIDLocation.for_element(a))