` Element containing the matching text. """
def __init__ (self, pattern, contentGroup=2):
Pattern.__init__(self, pattern, contentGroup)
self.tag = "code"
def handleMatch(self, m):
el = etree.Element(self.tag)
el.text = m.group(self.contentGroup).strip()
return el
class DoubleTagPattern (SimpleTagPattern):
"""
Return a TextNode nested in tag2 nested in tag1.
Usefull for strong emphasis etc.
"""
def handleMatch(self, m):
tag1, tag2 = self.tag.split(",")
el1 = etree.Element(tag1)
el2 = etree.SubElement(el1, tag2)
el2.text = m.group(2)
return el1
class HtmlPattern (Pattern):
""" Store raw inline html and return a placeholder. """
def handleMatch (self, m):
rawhtml = m.group(2)
inline = True
place_holder = self.stash.store(rawhtml)
return place_holder
class LinkPattern (Pattern):
""" Return a NanoDom link Element from the given match. """
def handleMatch(self, m):
el = etree.Element("a")
el.text = m.group(2)
parts = m.group(9).split('"')
# We should now have [], [href], or [href, title]
if parts:
el.set("href", self.sanatize_url(parts[0].strip()))
else:
el.set("href", "")
if len(parts) > 1:
# we also got a title
title = ('"' + '"'.join(parts[1:]).strip())[1:-1]
#title = dequote(title) #.replace('"', """)
el.set("title", title)
return el
def sanatize_url(self, url):
"""
Sanitize a url against xss attacks in "safe_mode".
Rather than specifically blacklisting `javascript:alert("XSS")` and all
its aliases (see ), we whitelist known
safe url formats. Most urls contain a network location, however some
are known not to (i.e.: mailto links). Script urls do not contain a
location. Additionally, for `javascript:...`, the scheme would be
"javascript" but some aliases will appear to `urlparse()` to have no
scheme. On top of that relative links (i.e.: "foo/bar.html") have no
scheme. Therefore we must check "path", "parameters", "query" and
"fragment" for any literal colons. We don't check "scheme" for colons
because it *should* never have any and "netloc" must allow the form:
`username:password@host:port`.
"""
locless_schemes = ['', 'mailto', 'news']
scheme, netloc, path, params, query, fragment = url = urlparse(url)
safe_url = False
if netloc != '' or scheme in locless_schemes:
safe_url = True
for part in url[2:]:
if ":" in part:
safe_url = False
if self.safe_mode and not safe_url:
return ''
else:
return urlunparse(url)
class ImagePattern(LinkPattern):
""" Return a NanoDom img Element from the given match. """
def handleMatch(self, m):
el = etree.Element("img")
src_parts = m.group(9).split()
if src_parts:
el.set('src', self.sanatize_url(src_parts[0]))
else:
el.set('src', "")
if len(src_parts) > 1:
el.set('title', dequote(" ".join(src_parts[1:])))
if ENABLE_ATTRIBUTES:
truealt = handleAttributes(m.group(2), el)
else:
truealt = m.group(2)
el.set('alt', truealt)
return el
class ReferencePattern(LinkPattern):
""" Match to a stored reference and return a NanoDom link Element. """
def handleMatch(self, m):
if m.group(9):
id = m.group(9).lower()
else:
# if we got something like "[Google][]"
# we'll use "google" as the id
id = m.group(2).lower()
if not self.references.has_key(id): # ignore undefined refs
return None
href, title = self.references[id]
text = m.group(2)
return self.makeTag(href, title, text)
def makeTag(self, href, title, text):
el = etree.Element('a')
el.set('href', self.sanatize_url(href))
if title:
el.set('title', title)
el.text = text
return el
class ImageReferencePattern (ReferencePattern):
""" Match to a stored reference and return a NanoDom img Element. """
def makeTag(self, href, title, text):
el = etree.Element("img")
el.set("src", self.sanatize_url(href))
if title:
el.set("title", title)
el.set("alt", text)
return el
class AutolinkPattern (Pattern):
""" Return a link Element given an autolink (``). """
def handleMatch(self, m):
el = etree.Element("a")
el.set('href', m.group(2))
el.text = m.group(2)
return el
class AutomailPattern (Pattern):
"""
Return a mailto link Element given an automail link (``).
"""
def handleMatch(self, m):
el = etree.Element('a')
email = m.group(2)
if email.startswith("mailto:"):
email = email[len("mailto:"):]
el.text = ""
for letter in email:
el.text += codepoint2name(ord(letter))
mailto = "mailto:" + email
mailto = "".join([AND_SUBSTITUTE + '#%d;' %
ord(letter) for letter in mailto])
el.set('href', mailto)
return el
ESCAPE_PATTERN = SimpleTextPattern(ESCAPE_RE)
NOT_STRONG_PATTERN = SimpleTextPattern(NOT_STRONG_RE)
BACKTICK_PATTERN = BacktickPattern(BACKTICK_RE)
DOUBLE_BACKTICK_PATTERN = BacktickPattern(DOUBLE_BACKTICK_RE)
STRONG_PATTERN = SimpleTagPattern(STRONG_RE, 'strong')
STRONG_PATTERN_2 = SimpleTagPattern(STRONG_2_RE, 'strong')
EMPHASIS_PATTERN = SimpleTagPattern(EMPHASIS_RE, 'em')
EMPHASIS_PATTERN_2 = SimpleTagPattern(EMPHASIS_2_RE, 'em')
STRONG_EM_PATTERN = DoubleTagPattern(STRONG_EM_RE, 'strong,em')
STRONG_EM_PATTERN_2 = DoubleTagPattern(STRONG_EM_2_RE, 'strong,em')
LINE_BREAK_PATTERN = SubstituteTagPattern(LINE_BREAK_RE, 'br ')
LINE_BREAK_PATTERN_2 = SubstituteTagPattern(LINE_BREAK_2_RE, 'br ')
LINK_PATTERN = LinkPattern(LINK_RE)
LINK_ANGLED_PATTERN = LinkPattern(LINK_ANGLED_RE)
IMAGE_LINK_PATTERN = ImagePattern(IMAGE_LINK_RE)
IMAGE_REFERENCE_PATTERN = ImageReferencePattern(IMAGE_REFERENCE_RE)
REFERENCE_PATTERN = ReferencePattern(REFERENCE_RE)
HTML_PATTERN = HtmlPattern(HTML_RE)
ENTITY_PATTERN = HtmlPattern(ENTITY_RE)
AUTOLINK_PATTERN = AutolinkPattern(AUTOLINK_RE)
AUTOMAIL_PATTERN = AutomailPattern(AUTOMAIL_RE)
"""
======================================================================
========================== POST-PROCESSORS ===========================
======================================================================
Markdown also allows post-processors, which are similar to
preprocessors in that they need to implement a "run" method. However,
they are run after core processing.
There are two types of post-processors: Postprocessor and TextPostprocessor
"""
class Postprocessor:
"""
Postprocessors are run before the dom it converted back into text.
Each Postprocessor implements a "run" method that takes a pointer to a
NanoDom document, modifies it as necessary and returns a NanoDom
document.
Postprocessors must extend markdown.Postprocessor.
There are currently no standard post-processors, but the footnote
extension uses one.
"""
def run(self, dom):
"""
Subclasses of Postprocessor should implement a `run` method, which
takes a NanoDOm document and returns a (possably modified) NanoDom
document.
"""
pass
class TextPostprocessor:
"""
TextPostprocessors are run after the dom it converted back into text.
Each TextPostprocessor implements a "run" method that takes a pointer to a
text string, modifies it as necessary and returns a text string.
TextPostprocessors must extend markdown.TextPostprocessor.
"""
def run(self, text):
"""
Subclasses of TextPostprocessor should implement a `run` method, which
takes the html document as a single text string and returns a
(possably modified) string.
"""
pass
class RawHtmlTextPostprocessor(TextPostprocessor):
""" Restore raw html to the document. """
def __init__(self):
pass
def run(self, text):
""" Iterate over html stash and restore "safe" html. """
for i in range(self.stash.html_counter):
html, safe = self.stash.rawHtmlBlocks[i]
if self.safeMode and not safe:
if str(self.safeMode).lower() == 'escape':
html = self.escape(html)
elif str(self.safeMode).lower() == 'remove':
html = ''
else:
html = HTML_REMOVED_TEXT
text = text.replace("%s\n
" % (HTML_PLACEHOLDER % i),
html + "\n")
text = text.replace(HTML_PLACEHOLDER % i, html)
return text
def escape(self, html):
""" Basic html escaping """
html = html.replace('&', '&')
html = html.replace('<', '<')
html = html.replace('>', '>')
return html.replace('"', '"')
RAWHTMLTEXTPOSTPROCESSOR = RawHtmlTextPostprocessor()
class AndSubstitutePostprocessor(TextPostprocessor):
""" Restore valid entities """
def __init__(self):
pass
def run(self, text):
text = text.replace(AND_SUBSTITUTE, "&")
return text
ANDSUBSTITUTETEXTPOSTPROCESSOR = AndSubstitutePostprocessor()
"""
======================================================================
========================== MISC AUXILIARY CLASSES ====================
======================================================================
"""
class HtmlStash:
"""
This class is used for stashing HTML objects that we extract
in the beginning and replace with place-holders.
"""
def __init__ (self):
""" Create a HtmlStash. """
self.html_counter = 0 # for counting inline html segments
self.rawHtmlBlocks=[]
def store(self, html, safe=False):
"""
Saves an HTML segment for later reinsertion. Returns a
placeholder string that needs to be inserted into the
document.
Keyword arguments:
* html: an html segment
* safe: label an html segment as safe for safemode
Returns : a placeholder string
"""
self.rawHtmlBlocks.append((html, safe))
placeholder = HTML_PLACEHOLDER % self.html_counter
self.html_counter += 1
return placeholder
class BlockGuru:
""" Parse document for block level constructs (paragraphs, lists, etc.)."""
def _findHead(self, lines, fn, allowBlank=0):
"""
Functional magic to help determine boundaries of indented
blocks.
Keyword arguments:
* lines: an array of strings
* fn: a function that returns a substring of a string
if the string matches the necessary criteria
* allowBlank: specifies whether it's ok to have blank
lines between matching functions
Returns: a list of post processes items and the unused
remainder of the original list
"""
items = []
item = -1
i = 0 # to keep track of where we are
for line in lines:
if not line.strip() and not allowBlank:
return items, lines[i:]
if not line.strip() and allowBlank:
# If we see a blank line, this _might_ be the end
i += 1
# Find the next non-blank line
for j in range(i, len(lines)):
if lines[j].strip():
next = lines[j]
break
else:
# There is no more text => this is the end
break
# Check if the next non-blank line is still a part of the list
part = fn(next)
if part:
items.append("")
continue
else:
break # found end of the list
part = fn(line)
if part:
items.append(part)
i += 1
continue
else:
return items, lines[i:]
else:
i += 1
return items, lines[i:]
def detabbed_fn(self, line):
""" An auxiliary method to be passed to _findHead """
m = RE.regExp['tabbed'].match(line)
if m:
return m.group(4)
else:
return None
def detectTabbed(self, lines):
""" Find indented text and remove indent before further proccesing. """
return self._findHead(lines, self.detabbed_fn,
allowBlank = 1)
def dequote(string):
""" Removes quotes from around a string """
if ( ( string.startswith('"') and string.endswith('"'))
or (string.startswith("'") and string.endswith("'")) ):
return string[1:-1]
else:
return string
class InlineStash:
def __init__(self):
self.prefix = INLINE_PLACEHOLDER_PREFIX
self.suffix = INLINE_PLACEHOLDER_SUFFIX
self._nodes = {}
self.phLength = 4 + len(self.prefix) + len(self.suffix)
def _genPlaceholder(self, type):
""" Generates placeholder """
id = "%04d" % len(self._nodes)
hash = "%s%s:%s%s" % (self.prefix, type, id,
self.suffix)
return hash, id
def extractId(self, data, index):
""" Extracting id from data string, starting from index """
endIndex = data.find(self.suffix, index+1)
if endIndex == -1:
return None, index + 1
else:
pair = data[index + len(self.prefix): endIndex].split(":")
if len(pair) == 2:
return pair[1], endIndex + len(self.suffix)
else:
return None, index + 1
def isin(self, id):
return self._nodes.has_key(id)
def get(self, id):
""" Returns node by id """
return self._nodes.get(id)
def add(self, node, type):
pholder, id = self._genPlaceholder(type)
self._nodes[id] = node
return pholder
"""
======================================================================
========================== CORE MARKDOWN =============================
======================================================================
This stuff is hard, so if you are thinking of extending the syntax,
see first if you can do it via pre-processors, post-processors,
inline patterns or a combination of the three.
"""
class CorePatterns:
"""
This class is scheduled for removal as part of a refactoring effort.
"""
patterns = {
'header': r'(#{1,6})[ \t]*(.*?)[ \t]*(#*)', # # A title
'reference-def': r'(\ ?\ ?\ ?)\[([^\]]*)\]:\s*([^ ]*)(.*)',
# [Google]: http://www.google.com/
'containsline': r'([-]*)$|^([=]*)', # -----, =====, etc.
'ol': r'[ ]{0,3}[\d]*\.\s+(.*)', # 1. text
'ul': r'[ ]{0,3}[*+-]\s+(.*)', # "* text"
'isline1': r'(\**)', # ***
'isline2': r'(\-*)', # ---
'isline3': r'(\_*)', # ___
'tabbed': r'((\t)|( ))(.*)', # an indented line
'quoted': r'[ ]{0,2}> ?(.*)', # a quoted block ("> ...")
}
def __init__ (self):
self.regExp = {}
for key in self.patterns.keys():
self.regExp[key] = re.compile("^%s$" % self.patterns[key],
re.DOTALL)
self.regExp['containsline'] = re.compile(r'^([-]*)$|^([=]*)$', re.M)
self.regExp['attr'] = re.compile("\{@([^\}]*)=([^\}]*)}") # {@id=123}
RE = CorePatterns()
class Markdown:
"""
Markdown formatter class for creating an html document from Markdown text.
"""
def __init__(self, source=None, # depreciated
extensions=[],
extension_configs={},
safe_mode = False):
"""
Creates a new Markdown instance.
Keyword arguments:
* source: The text in Markdown format. Depreciated!
* extensions: A list of extensions.
If they are of type string, the module mdx_name.py will be loaded.
If they are a subclass of markdown.Extension, they will be used
as-is.
* extension-configs: Configuration setting for extensions.
* safe_mode: Disallow raw html. One of "remove", "replace" or "escape".
"""
self.source = source
if source is not None:
message(WARN, "The `source` arg of Markdown.__init__() is depreciated and will be removed in the future. Use `instance.convert(source)` instead.")
self.safeMode = safe_mode
self.blockGuru = BlockGuru()
self.registeredExtensions = []
self.docType = ""
self.stripTopLevelTags = True
self.textPreprocessors = [HTML_BLOCK_PREPROCESSOR]
self.preprocessors = [HEADER_PREPROCESSOR,
LINE_PREPROCESSOR,
# A footnote preprocessor will
# get inserted here
REFERENCE_PREPROCESSOR]
self.postprocessors = [] # a footnote postprocessor will get
# inserted later
self.textPostprocessors = [# a footnote postprocessor will get
# inserted here
RAWHTMLTEXTPOSTPROCESSOR,
ANDSUBSTITUTETEXTPOSTPROCESSOR]
self.prePatterns = []
self.inlinePatterns = [
DOUBLE_BACKTICK_PATTERN,
BACKTICK_PATTERN,
ESCAPE_PATTERN,
REFERENCE_PATTERN,
LINK_ANGLED_PATTERN,
LINK_PATTERN,
IMAGE_LINK_PATTERN,
IMAGE_REFERENCE_PATTERN,
AUTOLINK_PATTERN,
AUTOMAIL_PATTERN,
LINE_BREAK_PATTERN_2,
LINE_BREAK_PATTERN,
HTML_PATTERN,
ENTITY_PATTERN,
NOT_STRONG_PATTERN,
STRONG_EM_PATTERN,
STRONG_EM_PATTERN_2,
STRONG_PATTERN,
STRONG_PATTERN_2,
EMPHASIS_PATTERN,
EMPHASIS_PATTERN_2
# The order of the handlers matters!!!
]
self.inlineStash = InlineStash()
self.registerExtensions(extensions = extensions,
configs = extension_configs)
self.reset()
def registerExtensions(self, extensions, configs):
"""
Register extensions with this instance of Markdown.
Keyword aurguments:
* extensions: A list of extensions, which can either
be strings or objects. See the docstring on Markdown.
* configs: A dictionary mapping module names to config options.
"""
for ext in extensions:
if isinstance(ext, basestring):
ext = load_extension(ext, configs.get(ext, []))
elif hasattr(ext, 'extendMarkdown'):
# Looks like an Extension.
# Nothing to do here.
pass
else:
message(ERROR, "Incorrect type! Extension '%s' is "
"neither a string or an Extension." %(repr(ext)))
continue
ext.extendMarkdown(self, globals())
def registerExtension(self, extension):
""" This gets called by the extension """
self.registeredExtensions.append(extension)
def reset(self):
"""
Resets all state variables so that we can start with a new text.
"""
self.references={}
self.htmlStash = HtmlStash()
self.inlineStash = InlineStash()
HTML_BLOCK_PREPROCESSOR.stash = self.htmlStash
LINE_PREPROCESSOR.stash = self.htmlStash
REFERENCE_PREPROCESSOR.references = self.references
HTML_PATTERN.stash = self.htmlStash
ENTITY_PATTERN.stash = self.htmlStash
REFERENCE_PATTERN.references = self.references
IMAGE_REFERENCE_PATTERN.references = self.references
RAWHTMLTEXTPOSTPROCESSOR.stash = self.htmlStash
RAWHTMLTEXTPOSTPROCESSOR.safeMode = self.safeMode
for extension in self.registeredExtensions:
extension.reset()
for pattern in self.inlinePatterns:
pattern.safe_mode = self.safeMode
def _transform(self):
"""Transform the Markdown text into a XHTML body document.
Returns: ElementTree object
"""
# Setup the document
self.root = etree.Element("span")
# Split into lines and run the preprocessors that will work with
# self.lines
self.lines = self.source.split("\n")
# Run the pre-processors on the lines
for prep in self.preprocessors :
self.lines = prep.run(self.lines)
# Create a NanoDom tree from the lines and attach it to Document
buffer = []
for line in self.lines:
if line.startswith("#"):
self._processSection(self.root, buffer)
buffer = [line]
else:
buffer.append(line)
self._processSection(self.root, buffer)
#self._processSection(self.top_element, self.lines)
# Not sure why I put this in but let's leave it for now.
#self.top_element.appendChild(self.doc.createTextNode('\n'))
# Run the post-processors
for postprocessor in self.postprocessors:
postprocessor.run(self.doc)
return etree.ElementTree(self.root)
def _processSection(self, parent_elem, lines,
inList=0, looseList=0):
"""
Process a section of a source document, looking for high
level structural elements like lists, block quotes, code
segments, html blocks, etc. Some those then get stripped
of their high level markup (e.g. get unindented) and the
lower-level markup is processed recursively.
Keyword arguments:
* parent_elem: A ElementTree element to which the content will be added.
* lines: a list of lines
* inList: a level
Returns: None
"""
# Loop through lines until none left.
while lines:
# Check if this section starts with a list, a blockquote or
# a code block
processFn = { 'ul': self._processUList,
'ol': self._processOList,
'quoted': self._processQuote,
'tabbed': self._processCodeBlock}
for regexp in ['ul', 'ol', 'quoted', 'tabbed']:
m = RE.regExp[regexp].match(lines[0])
if m:
processFn[regexp](parent_elem, lines, inList)
return
# We are NOT looking at one of the high-level structures like
# lists or blockquotes. So, it's just a regular paragraph
# (though perhaps nested inside a list or something else). If
# we are NOT inside a list, we just need to look for a blank
# line to find the end of the block. If we ARE inside a
# list, however, we need to consider that a sublist does not
# need to be separated by a blank line. Rather, the following
# markup is legal:
#
# * The top level list item
#
# Another paragraph of the list. This is where we are now.
# * Underneath we might have a sublist.
#
if inList:
start, lines = self._linesUntil(lines, (lambda line:
RE.regExp['ul'].match(line)
or RE.regExp['ol'].match(line)
or not line.strip()))
self._processSection(parent_elem, start,
inList - 1, looseList = looseList)
inList = inList-1
else: # Ok, so it's just a simple block
paragraph, lines = self._linesUntil(lines, lambda line:
not line.strip())
if len(paragraph) and paragraph[0].startswith('#'):
self._processHeader(parent_elem, paragraph)
elif len(paragraph) and \
RE.regExp["isline3"].match(paragraph[0]):
self._processHR(parent_elem)
lines = paragraph[1:] + lines
elif paragraph:
self._processParagraph(parent_elem, paragraph,
inList, looseList)
if lines and not lines[0].strip():
lines = lines[1:] # skip the first (blank) line
def _processHR(self, parent_elem):
hr = etree.SubElement(parent_elem, "hr")
def _processHeader(self, parent_elem, paragraph):
m = RE.regExp['header'].match(paragraph[0])
if m:
level = len(m.group(1))
h = etree.SubElement(parent_elem, "h%d" % level)
inline = etree.SubElement(h, "inline")
inline.text = m.group(2).strip()
else:
message(CRITICAL, "We've got a problem header!")
def _processParagraph(self, parent_elem, paragraph, inList, looseList):
if ( parent_elem.tag == 'li'
and not (looseList or parent_elem.getchildren())):
# If this is the first paragraph inside "li", don't
# put around it - append the paragraph bits directly
# onto parent_elem
el = parent_elem
else:
# Otherwise make a "p" element
el = etree.SubElement(parent_elem, "p")
dump = []
# Searching for hr
for line in paragraph:
if RE.regExp["isline3"].match(line):
inline = etree.SubElement(el, "inline")
inline.text = "\n".join(dump)
etree.SubElement(el, "hr")
dump.clear()
else:
dump.append(line)
if dump:
text = "\n".join(dump)
inline = etree.SubElement(el, "inline")
inline.text = text
def _processUList(self, parent_elem, lines, inList):
self._processList(parent_elem, lines, inList,
listexpr='ul', tag = 'ul')
def _processOList(self, parent_elem, lines, inList):
self._processList(parent_elem, lines, inList,
listexpr='ol', tag = 'ol')
def _processList(self, parent_elem, lines, inList, listexpr, tag):
"""
Given a list of document lines starting with a list item,
finds the end of the list, breaks it up, and recursively
processes each list item and the remainder of the text file.
Keyword arguments:
* parent_elem: A dom element to which the content will be added
* lines: a list of lines
* inList: a level
Returns: None
"""
ul = etree.SubElement(parent_elem, tag) # ul might actually be '
'
looseList = 0
# Make a list of list items
items = []
item = -1
i = 0 # a counter to keep track of where we are
for line in lines:
loose = 0
if not line.strip():
# If we see a blank line, this _might_ be the end of the list
i += 1
loose = 1
# Find the next non-blank line
for j in range(i, len(lines)):
if lines[j].strip():
next = lines[j]
break
else:
# There is no more text => end of the list
break
# Check if the next non-blank line is still a part of the list
if ( RE.regExp['ul'].match(next) or
RE.regExp['ol'].match(next) or
RE.regExp['tabbed'].match(next) ):
# get rid of any white space in the line
items[item].append(line.strip())
looseList = loose or looseList
continue
else:
break # found end of the list
# Now we need to detect list items (at the current level)
# while also detabing child elements if necessary
for expr in ['ul', 'ol', 'tabbed']:
m = RE.regExp[expr].match(line)
if m:
if expr in ['ul', 'ol']: # We are looking at a new item
#if m.group(1) :
# Removed the check to allow for a blank line
# at the beginning of the list item
items.append([m.group(1)])
item += 1
elif expr == 'tabbed': # This line needs to be detabbed
items[item].append(m.group(4)) #after the 'tab'
i += 1
break
else:
items[item].append(line) # Just regular continuation
i += 1 # added on 2006.02.25
else:
i += 1
# Add the dom elements
for item in items:
li = etree.SubElement(ul, "li")
self._processSection(li, item, inList + 1, looseList = looseList)
# Process the remaining part of the section
self._processSection(parent_elem, lines[i:], inList)
def _linesUntil(self, lines, condition):
"""
A utility function to break a list of lines upon the
first line that satisfied a condition. The condition
argument should be a predicate function.
"""
i = -1
for line in lines:
i += 1
if condition(line): break
else:
i += 1
return lines[:i], lines[i:]
def _processQuote(self, parent_elem, lines, inList):
"""
Given a list of document lines starting with a quote finds
the end of the quote, unindents it and recursively
processes the body of the quote and the remainder of the
text file.
Keyword arguments:
* parent_elem: DOM element to which the content will be added
* lines: a list of lines
* inList: a level
Returns: None
"""
dequoted = []
i = 0
blank_line = False # allow one blank line between paragraphs
for line in lines:
m = RE.regExp['quoted'].match(line)
if m:
dequoted.append(m.group(1))
i += 1
blank_line = False
elif not blank_line and line.strip() != '':
dequoted.append(line)
i += 1
elif not blank_line and line.strip() == '':
dequoted.append(line)
i += 1
blank_line = True
else:
break
blockquote = etree.SubElement(parent_elem, "blockquote")
self._processSection(blockquote, dequoted, inList)
self._processSection(parent_elem, lines[i:], inList)
def _processCodeBlock(self, parent_elem, lines, inList):
"""
Given a list of document lines starting with a code block
finds the end of the block, puts it into the dom verbatim
wrapped in ("") and recursively processes the
the remainder of the text file.
Keyword arguments:
* parent_elem: DOM element to which the content will be added
* lines: a list of lines
* inList: a level
Returns: None
"""
detabbed, theRest = self.blockGuru.detectTabbed(lines)
pre = etree.SubElement(parent_elem, "pre")
code = etree.SubElement(pre, "code")
text = "\n".join(detabbed).rstrip()+"\n"
#text = text.replace("&", "&")
code.text = text
self._processSection(parent_elem, theRest, inList)
def _handleInline(self, data, patternIndex=0):
"""
Processinf string with inline patterns and replasing it
with placeholders
Keyword arguments:
* data: A line of Markdown text
* patternIndex: The index of the inlinePattern to start with
Return: String with placeholders.
"""
while patternIndex < len(self.inlinePatterns):
data, matched = self._applyInline(self.inlinePatterns[patternIndex],
data, patternIndex)
if not matched:
patternIndex += 1
return data
def _applyInline(self, pattern, data, patternIndex):
"""
Given a pattern name, this function checks if the line
fits the pattern, creates the necessary elements, adds it
to InlineStash, and returns string with placeholders,
instead of ElementTree elements.
Keyword arguments:
* data: the text to be processed
* pattern: the pattern to be checked
* patternIndex: index of current pattern
Returns: String with placeholders.
"""
match = pattern.getCompiledRegExp().match(data)
if not match:
return data, False
node = pattern.handleMatch(match)
if node is None:
return data, False
if not isstr(node):
if not node.tag in ["code", "pre"]:
# We need to process current node too
for child in [node] + node.getchildren():
if not isstr(node):
if child.text:
child.text = self._handleInline(child.text,
patternIndex)
if child.tail:
child.tail = self._handleInline(child.tail,
patternIndex)
pholder = self.inlineStash.add(node, pattern.type())
left = "".join([match.group(i) for i in xrange(1,
pattern.contentGroup)])
return "%s%s%s" % (left, pholder, match.groups()[-1]), True
def _processElementText(self, node, subnode, isText=True):
if isText:
text = subnode.text
subnode.text = None
else:
text = subnode.tail
subnode.tail = None
childResult = self._processPlaceholders(text, subnode)
if not isText and node is not subnode:
pos = node.getchildren().index(subnode)
node.remove(subnode)
else:
pos = 0
childResult.reverse()
for newChild in childResult:
node.insert(pos, newChild)
def _processPlaceholders(self, data, parent):
"""
Processes string with placeholders and generates DOM tree.
* data: string with placeholders instead of DOM elements.
Returns: NanoDOM Document object with applied inline patterns.
"""
def linkText(text):
if text:
if result:
if result[-1].tail:
result[-1].tail += text
else:
result[-1].tail = text
else:
if parent.text:
parent.text += text
else:
parent.text = text
result = []
prefix = self.inlineStash.prefix
strartIndex = 0
while data:
index = data.find(prefix, strartIndex)
if index != -1:
id, phEndIndex = self.inlineStash.extractId(data, index)
if self.inlineStash.isin(id):
node = self.inlineStash.get(id)
if index > 0:
text = data[strartIndex:index]
linkText(text)
if not isstr(node): # it's Element
for child in [node] + node.getchildren():
if child.tail:
self._processElementText(node, child, False)
if child.text:
self._processElementText(node, child)
else: # it's just a string
linkText(node)
strartIndex = phEndIndex
continue
strartIndex = phEndIndex
result.append(node)
else: # wrong placeholder
end = index + len(prefix)
linkText(data[strartIndex:end])
strartIndex = end
else:
text = data[strartIndex:]
linkText(text)
data = ""
return result
def _processTree(self, el):
"""
Processing ElementTree, and applying inline patterns
Keyword arguments:
* el - parent element of Document.
Returns: ElementTree object with applied inline patterns.
"""
stack = [el]
while stack:
currElement = stack.pop()
insertQueue = []
for child in currElement.getchildren():
if child.tag == "inline":
lst = self._processPlaceholders(self._handleInline(
child.text), currElement)
pos = currElement.getchildren().index(child)
insertQueue.append((child, pos, lst))
else:
stack.append(child)
for element, pos, lst in insertQueue:
currElement.remove(element)
if currElement.text:
currElement.text = handleAttributes(currElement.text,
currElement)
for newChild in lst:
# Processing attributes
if newChild.tail:
newChild.tail = handleAttributes(newChild.tail,
currElement)
if newChild.text:
newChild.text = handleAttributes(newChild.text,
newChild)
currElement.insert(pos, newChild)
pos += 1
def applyInlinePatterns(self, markdownTree):
"""
Retrun ElementTree, with applied
inline paterns
Keyword arguments:
* markdownTree: ElementTree object, reppresenting Markdown tree.
Returns: ElementTree object.
"""
el = markdownTree.getroot()
self._processTree(el)
return markdownTree
def markdownToTree(self, source=None):
"""
Retrun ElementTree, without applying
inline paterns
Keyword arguments:
* source: An ascii or unicode string of Markdown formated text.
Returns: ElementTree object.
"""
if source is not None: #Allow blank string
self.source = source
if not self.source:
return u""
try:
self.source = unicode(self.source)
except UnicodeDecodeError:
message(CRITICAL, 'UnicodeDecodeError: Markdown only accepts unicode or ascii input.')
return u""
# Fixup the source text
self.source = self.source.replace("\r\n", "\n").replace("\r", "\n")
self.source += "\n\n"
self.source = self.source.expandtabs(TAB_LENGTH)
for pp in self.textPreprocessors:
self.source = pp.run(self.source)
markdownTree = self._transform()
return markdownTree
def convert (self, source=None):
"""
Return the document in XHTML format.
Keyword arguments:
* source: An ascii or unicode string of Markdown formated text.
Returns: A serialized XHTML body.
"""
tree = self.markdownToTree(source)
root = self.applyInlinePatterns(tree).getroot()
indentETree(root)
xml = codecs.decode(etree.tostring(root, encoding="utf8"), "utf8")
if self.stripTopLevelTags:
xml = xml.strip()[44:-7] + "\n"
for pp in self.textPostprocessors:
xml = pp.run(xml)
return xml.strip()
def __str__(self):
''' Report info about instance. Markdown always returns unicode. '''
if self.source is None:
status = 'in which no source text has been assinged.'
else:
status = 'which contains %d chars and %d line(s) of source.'%\
(len(self.source), self.source.count('\n')+1)
return 'An instance of "%s" %s'% (self.__class__, status)
__unicode__ = convert # markdown should always return a unicode string
# ====================================================================
def markdownFromFile(input = None,
output = None,
extensions = [],
encoding = None,
message_threshold = CRITICAL,
safe = False):
"""
Convenience wrapper function that takes a filename as input.
Used from the command-line, although may be useful in other situations.
Decodes the file using the provided encoding (defaults to utf-8), passes
the file content to markdown, and outputs the html to either the provided
filename or stdout in the same encoding as the source file.
**Note:** This is the only place that decoding and encoding takes place
in Python-Markdown.
Keyword arguments:
* input: Name of source text file.
* output: Name of output file. Writes to stdout if `None`.
* extensions: A list of extension names (may contain config args).
* encoding: Encoding of input and output files. Defaults to utf-8.
* message_threshold: Error reporting level.
* safe_mode: Disallow raw html. One of "remove", "replace" or "escape".
Returns: An HTML document as a string.
"""
global console_hndlr
console_hndlr.setLevel(message_threshold)
message(DEBUG, "input file: %s" % input)
if not encoding:
encoding = "utf-8"
input_file = codecs.open(input, mode="r", encoding=encoding)
text = input_file.read()
input_file.close()
text = removeBOM(text, encoding)
new_text = markdown(text, extensions, safe_mode = safe)
if output:
output_file = codecs.open(output, "w", encoding=encoding)
output_file.write(new_text)
output_file.close()
else:
sys.stdout.write(new_text.encode(encoding))
def markdown(text,
extensions = [],
safe_mode = False):
"""
Convenience wrapper function for `Markdown` class.
Useful in a typical use case. Initializes an instance of the `Markdown`
class, loads any extensions and runs the parser on the given text.
Keyword arguments:
* text: An ascii or Unicode string of Markdown formatted text.
* extensions: A list of extension names (may contain config args).
* safe_mode: Disallow raw html. One of "remove", "replace" or "escape".
Returns: An HTML document as a string.
"""
message(DEBUG, "in markdown.markdown(), received text:\n%s" % text)
extensions = [load_extension(e) for e in extensions]
md = Markdown(extensions=extensions,
safe_mode = safe_mode)
return md.convert(text)
class Extension:
""" Base class for extensions to subclass. """
def __init__(self, configs = {}):
"""
Create an instance of an Extention.
Keyword arguments:
* configs: A dict of configuration setting used by an Extension.
"""
self.config = configs
def getConfig(self, key):
""" Return a setting for the given key or an empty string. """
if self.config.has_key(key):
return self.config[key][0]
else:
return ""
def getConfigInfo(self):
""" Return all config settings as a list of tuples. """
return [(key, self.config[key][1]) for key in self.config.keys()]
def setConfig(self, key, value):
""" Set a config setting for `key` with the given `value`. """
self.config[key][0] = value
def load_extension(ext_name, configs = []):
"""
Load extension by name, then return the module.
The extension name may contain arguments as part of the string in the
following format:
"extname(key1=value1,key2=value2)"
Print an error message and exit on failure.
"""
# I am making the assumption that the order of config options
# does not matter.
configs = dict(configs)
pos = ext_name.find("(")
if pos > 0:
ext_args = ext_name[pos+1:-1]
ext_name = ext_name[:pos]
pairs = [x.split("=") for x in ext_args.split(",")]
configs.update([(x.strip(), y.strip()) for (x, y) in pairs])
extension_module_name = "mdx_" + ext_name
try:
module = __import__(extension_module_name)
except:
message(CRITICAL,
"couldn't load extension %s (looking for %s module)"
% (ext_name, extension_module_name) )
sys.exit(1)
return module.makeExtension(configs.items())
OPTPARSE_WARNING = """
Python 2.3 or higher required for advanced command line options.
For lower versions of Python use:
%s INPUT_FILE > OUTPUT_FILE
""" % EXECUTABLE_NAME_FOR_USAGE
def parse_options():
"""
Define and parse `optparse` options for command-line usage.
"""
try:
optparse = __import__("optparse")
except:
if len(sys.argv) == 2:
return {'input': sys.argv[1],
'output': None,
'message_threshold': CRITICAL,
'safe': False,
'extensions': [],
'encoding': None }
else:
print OPTPARSE_WARNING
return None
parser = optparse.OptionParser(usage="%prog INPUTFILE [options]")
parser.add_option("-f", "--file", dest="filename",
help="write output to OUTPUT_FILE",
metavar="OUTPUT_FILE")
parser.add_option("-e", "--encoding", dest="encoding",
help="encoding for input and output files",)
parser.add_option("-q", "--quiet", default = CRITICAL,
action="store_const", const=60, dest="verbose",
help="suppress all messages")
parser.add_option("-v", "--verbose",
action="store_const", const=INFO, dest="verbose",
help="print info messages")
parser.add_option("-s", "--safe", dest="safe", default=False,
metavar="SAFE_MODE",
help="safe mode ('replace', 'remove' or 'escape' user's HTML tag)")
parser.add_option("--noisy",
action="store_const", const=DEBUG, dest="verbose",
help="print debug messages")
parser.add_option("-x", "--extension", action="append", dest="extensions",
help = "load extension EXTENSION", metavar="EXTENSION")
(options, args) = parser.parse_args()
if not len(args) == 1:
parser.print_help()
return None
else:
input_file = args[0]
if not options.extensions:
options.extensions = []
return {'input': input_file,
'output': options.filename,
'message_threshold': options.verbose,
'safe': options.safe,
'extensions': options.extensions,
'encoding': options.encoding }
if __name__ == '__main__':
""" Run Markdown from the command line. """
options = parse_options()
#if os.access(inFile, os.R_OK):
if not options:
sys.exit(0)
markdownFromFile(**options)