def run(self, view, wholefile=False, preview=False):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings',
view.file_name())
self.preview = preview
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get('html_template')
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += '<html><head><meta name="viewport" content="width=device-width,initial-scale=1.0"/>'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_uml()
html += self.get_title()
html += '</head><body>'
html += '<article class="markdown-body">'
html += body
html += '</article>'
html += '</body>'
html += '</html>'
return html, body
def run(self):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings',
self.filename)
contents = self.get_contents(self.filename)
body = self.convert_markdown(contents)
# add self.html_template, it can be assigned.
if self.html_template is None:
html_template = self.settings.get('html_template')
else:
html_template = self.html_template
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_title()
html += '</head><body>'
html += body
html += '</body>'
html += '</html>'
return html, body
def run(self, view, wholefile=False, preview=False):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings', view.file_name())
self.preview = preview
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get('html_template')
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += '<html><head><meta name="viewport" content="width=device-width,initial-scale=1.0"/>'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_uml()
html += self.get_title()
html += '</head><body>'
html += '<article class="markdown-body">'
html += body
html += '</article>'
html += '</body>'
html += '</html>'
return html, body
def run(self, view, wholefile=False, preview=False):
""" return full html and body html for view. """
self.settings = Settings("MarkdownPreview.sublime-settings", view.file_name())
self.preview = preview
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get("html_template")
# use customized html template if given
if self.settings.get("html_simple", False):
html = body
elif html_template and os.path.exists(html_template):
head = u""
head += self.get_meta()
if not self.settings.get("skip_default_stylesheet"):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace("{{ HEAD }}", head, 1)
html = html.replace("{{ BODY }}", body, 1)
else:
html = u"<!DOCTYPE html>"
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_uml()
html += self.get_title()
html += "</head><body>"
html += '<article class="markdown-body">'
html += body
html += "</article>"
html += "</body>"
html += "</html>"
return html, body
def run(self):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings', self.filename)
contents = self.get_contents(self.filename)
body = self.convert_markdown(contents)
# add self.html_template, it can be assigned.
if self.html_template is None:
html_template = self.settings.get('html_template')
else:
html_template = self.html_template
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_title()
html += '</head><body>'
html += body
html += '</body>'
html += '</html>'
return html, body
class Compiler(object):
""" Do the markdown converting """
default_css = "markdown.css"
def isurl(self, css_name):
match = re.match(r"https?://", css_name)
if match:
return True
return False
def get_font_css(self):
fonts = ""
css_names = ["fonts/migu-1c-regular.css"]
for css_name in css_names:
fonts = u"%s%s" % (fonts, load_resource(css_name))
return u"<style>%s</style>" % fonts
def get_default_css(self):
""" locate the correct CSS with the 'css' setting """
css_name = self.settings.get("css", "default")
if self.isurl(css_name):
# link to remote URL
return u"<link href='%s' rel='stylesheet' type='text/css'>" % css_name
elif os.path.isfile(os.path.expanduser(css_name)):
# use custom CSS file
return u"<style>%s</style>" % load_utf8(os.path.expanduser(css_name))
elif css_name == "default":
# use parser CSS file
return u"<style>%s</style>" % load_resource(self.default_css)
return ""
def get_override_css(self):
""" handls allow_css_overrides setting. """
if self.settings.get("allow_css_overrides"):
filename = self.view.file_name()
filetypes = self.settings.get("markdown_filetypes")
if filename and filetypes:
for filetype in filetypes:
if filename.endswith(filetype):
css_filename = filename.rpartition(filetype)[0] + ".css"
if os.path.isfile(css_filename):
return u"<style>%s</style>" % load_utf8(css_filename)
return ""
def get_stylesheet(self):
""" return the correct CSS file based on parser and settings """
return self.get_font_css() + self.get_default_css() + self.get_override_css()
def get_javascript(self):
js_files = self.settings.get("js")
scripts = ""
if js_files is not None:
# Ensure string values become a list.
if isinstance(js_files, str) or isinstance(js_files, unicode_str):
js_files = [js_files]
# Only load scripts if we have a list.
if isinstance(js_files, list):
for js_file in js_files:
if os.path.isabs(js_file):
# Load the script inline to avoid cross-origin.
scripts += u"<script>%s</script>" % load_utf8(js_file)
else:
scripts += u"<script type='text/javascript' src='%s'></script>" % js_file
return scripts
def get_mathjax(self):
""" return the MathJax script if enabled """
if self.settings.get("enable_mathjax") is True:
return load_resource("mathjax.html")
return ""
def get_uml(self):
""" return the uml scripts if enabled """
if self.settings.get("enable_uml") is True:
flow = load_resource("flowchart-min.js")
return load_resource("uml.html").replace("{{ flowchart }}", flow, 1)
return ""
def get_highlight(self):
return ""
def get_contents(self, wholefile=False):
""" Get contents or selection from view and optionally strip the YAML front matter """
region = sublime.Region(0, self.view.size())
contents = self.view.substr(region)
if not wholefile:
# use selection if any
selection = self.view.substr(self.view.sel()[0])
if selection.strip() != "":
contents = selection
# Remove yaml front matter
if self.settings.get("strip_yaml_front_matter") and contents.startswith("---"):
frontmatter, contents = self.preprocessor_yaml_frontmatter(contents)
self.settings.apply_frontmatter(frontmatter)
references = self.settings.get("builtin").get("references", [])
for ref in references:
contents += get_references(ref)
contents = self.parser_specific_preprocess(contents)
return contents
def parser_specific_preprocess(self, text):
return text
def preprocessor_yaml_frontmatter(self, text):
""" Get frontmatter from string """
frontmatter = {}
if text.startswith("---"):
m = re.search(r"^(---(.*?)---[ \t]*\r?\n)", text, re.DOTALL)
if m:
try:
frontmatter = yaml.load(m.group(2))
except:
print(traceback.format_exc())
text = text[m.end(1) :]
return frontmatter, text
def parser_specific_postprocess(self, text):
return text
def postprocessor_pathconverter(self, html, image_convert, file_convert, absolute=False):
RE_TAG_HTML = r"""(?xus)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><(?P<tag>(?:%s)))
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
"""
RE_TAG_LINK_ATTR = re.compile(
r"""(?xus)
(?P<attr>
(?:
(?P<name>\s+(?:href|src)\s*=\s*)
(?P<path>"[^"]*"|'[^']*')
)
)
"""
)
RE_SOURCES = re.compile(
RE_TAG_HTML
% (
(r"img" if image_convert else "")
+ (r"|" if image_convert and file_convert else "")
+ (r"script|a|link" if file_convert else "")
)
)
def repl(m, base_path, rel_path=None):
if m.group("comments"):
tag = m.group("comments")
else:
tag = m.group("open")
if rel_path is None:
tag += RE_TAG_LINK_ATTR.sub(lambda m2: repl_absolute(m2, base_path), m.group("attr"))
else:
tag += RE_TAG_LINK_ATTR.sub(lambda m2: repl_relative(m2, base_path, rel_path), m.group("attr"))
tag += m.group("close")
return tag
basepath = self.settings.get("builtin").get("basepath")
if basepath is None:
basepath = ""
if absolute:
if basepath:
return RE_SOURCES.sub(lambda m: repl(m, basepath), html)
else:
if self.preview:
relativepath = getTempMarkdownPreviewPath(self.view)
else:
relativepath = self.settings.get("builtin").get("destination")
if not relativepath:
mdfile = self.view.file_name()
if mdfile is not None and os.path.exists(mdfile):
relativepath = os.path.splitext(mdfile)[0] + ".html"
if relativepath:
relativepath = os.path.dirname(relativepath)
if basepath and relativepath:
return RE_SOURCES.sub(lambda m: repl(m, basepath, relativepath), html)
return html
def postprocessor_base64(self, html):
""" convert resources (currently images only) to base64 """
file_types = {(".png",): "image/png", (".jpg", ".jpeg"): "image/jpeg", (".gif",): "image/gif"}
exclusion_list = tuple(["https://", "http://", "#"] + ["data:%s;base64," % ft for ft in file_types.values()])
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_TAG_HTML = re.compile(
r"""(?xus)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><(?P<tag>img))
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
"""
)
RE_TAG_LINK_ATTR = re.compile(
r"""(?xus)
(?P<attr>
(?:
(?P<name>\s+src\s*=\s*)
(?P<path>"[^"]*"|'[^']*')
)
)
"""
)
def b64(m):
import base64
src = url2pathname(m.group("path")[1:-1])
data = m.group(0)
base_path = self.settings.get("builtin").get("basepath")
if base_path is None:
base_path = ""
# Format the link
absolute = False
if src.startswith("file://"):
src = src.replace("file://", "", 1)
if sublime.platform() == "windows" and not src.startswith("//"):
src = src.lstrip("/")
absolute = True
elif sublime.platform() == "windows" and RE_WIN_DRIVE.match(src) is not None:
absolute = True
# Make sure we are working with an absolute path
if not src.startswith(exclusion_list):
if absolute:
src = os.path.normpath(src)
else:
src = os.path.normpath(os.path.join(base_path, src))
if os.path.exists(src):
ext = os.path.splitext(src)[1].lower()
for b64_ext in file_types:
if ext in b64_ext:
try:
with open(src, "rb") as f:
data = ' src="data:%s;base64,%s"' % (
file_types[b64_ext],
base64.b64encode(f.read()).decode("ascii"),
)
except Exception:
pass
break
return data
def repl(m):
if m.group("comments"):
tag = m.group("comments")
else:
tag = m.group("open")
tag += RE_TAG_LINK_ATTR.sub(lambda m2: b64(m2), m.group("attr"))
tag += m.group("close")
return tag
return RE_TAG_HTML.sub(repl, html)
def postprocessor_simple(self, html):
""" Strip out ids and classes for a simplified HTML output """
def repl(m):
if m.group("comments"):
tag = ""
else:
tag = m.group("open")
tag += RE_TAG_BAD_ATTR.sub("", m.group("attr"))
tag += m.group("close")
return tag
# Strip out id, class, on<word>, and style attributes for a simple html output
RE_TAG_HTML = re.compile(
r"""(?x)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><[\w\:\.\-]+)
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
""",
re.DOTALL | re.UNICODE,
)
RE_TAG_BAD_ATTR = re.compile(
r"""(?x)
(?P<attr>
(?:
\s+(?:id|class|style|on[\w]+)
(?:\s*=\s*(?:"[^"]*"|'[^']*'))
)*
)
""",
re.DOTALL | re.UNICODE,
)
return RE_TAG_HTML.sub(repl, html)
def convert_markdown(self, markdown_text):
""" convert input markdown to HTML, with github or builtin parser """
markdown_html = self.parser_specific_convert(markdown_text)
image_convert = self.settings.get("image_path_conversion", "absolute")
file_convert = self.settings.get("file_path_conversions", "absolute")
markdown_html = self.parser_specific_postprocess(markdown_html)
if "absolute" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(markdown_html, image_convert, file_convert, True)
if "relative" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(markdown_html, image_convert, file_convert, False)
if image_convert == "base64":
markdown_html = self.postprocessor_base64(markdown_html)
if self.settings.get("html_simple", False):
markdown_html = self.postprocessor_simple(markdown_html)
return markdown_html
def get_title(self):
if self.meta_title is not None:
title = self.meta_title
else:
title = self.view.name()
if not title:
fn = self.view.file_name()
title = "untitled" if not fn else os.path.splitext(os.path.basename(fn))[0]
return "<title>%s</title>" % cgi.escape(title)
def get_meta(self):
self.meta_title = None
meta = []
for k, v in self.settings.get("meta", {}).items():
if k == "title":
if isinstance(v, list):
if len(v) == 0:
v = ""
else:
v = v[0]
self.meta_title = unicode_str(v)
continue
if isinstance(v, list):
v = ",".join(v)
if v is not None:
meta.append('<meta name="%s" content="%s">' % (cgi.escape(k, True), cgi.escape(v, True)))
return "\n".join(meta)
def run(self, view, wholefile=False, preview=False):
""" return full html and body html for view. """
self.settings = Settings("MarkdownPreview.sublime-settings", view.file_name())
self.preview = preview
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get("html_template")
# use customized html template if given
if self.settings.get("html_simple", False):
html = body
elif html_template and os.path.exists(html_template):
head = u""
head += self.get_meta()
if not self.settings.get("skip_default_stylesheet"):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace("{{ HEAD }}", head, 1)
html = html.replace("{{ BODY }}", body, 1)
else:
html = u"<!DOCTYPE html>"
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_uml()
html += self.get_title()
html += "</head><body>"
html += '<article class="markdown-body">'
html += body
html += "</article>"
html += "</body>"
html += "</html>"
return html, body
class Compiler(object):
''' Do the markdown converting '''
default_css = "markdown.css"
filename = "../first.md"
html_template = None
def __init__(self, file_name, html_template=None):
self.filename = file_name
def isurl(self, css_name):
match = re.match(r'https?://', css_name)
if match:
return True
return False
def get_default_css(self):
''' locate the correct CSS with the 'css' setting '''
css_name = self.settings.get('css', 'default')
if self.isurl(css_name):
# link to remote URL
return u"<link href='%s' rel='stylesheet' type='text/css'>" % css_name
elif os.path.isfile(os.path.expanduser(css_name)):
# use custom CSS file
return u"<style>%s</style>" % load_utf8(
os.path.expanduser(css_name))
elif css_name == 'default':
# use parser CSS file
return u"<style>%s</style>" % load_resource(self.default_css)
return ''
def get_override_css(self):
''' handls allow_css_overrides setting. '''
if self.settings.get('allow_css_overrides'):
filetypes = self.settings.get('markdown_filetypes')
if self.filename and filetypes:
for filetype in filetypes:
if self.filename.endswith(filetype):
css_filename = self.filename.rpartition(
filetype)[0] + '.css'
if (os.path.isfile(css_filename)):
return u"<style>%s</style>" % load_utf8(
css_filename)
return ''
def get_stylesheet(self):
''' return the correct CSS file based on parser and settings '''
return self.get_default_css() + self.get_override_css()
def get_javascript(self):
js_files = self.settings.get('js')
scripts = ''
if js_files is not None:
# Ensure string values become a list.
if isinstance(js_files, str) or isinstance(js_files, unicode_str):
js_files = [js_files]
# Only load scripts if we have a list.
if isinstance(js_files, list):
for js_file in js_files:
if os.path.isabs(js_file):
# Load the script inline to avoid cross-origin.
scripts += u"<script>%s</script>" % load_utf8(js_file)
else:
scripts += u"<script type='text/javascript' src='%s'></script>" % js_file
return scripts
def get_mathjax(self):
''' return the MathJax script if enabled '''
if self.settings.get('enable_mathjax') is True:
return load_resource('mathjax.html')
return ''
def get_uml(self):
''' return the uml scripts if enabled '''
if self.settings.get('enable_uml') is True:
flow = load_resource('flowchart-min.js')
return load_resource('uml.html').replace('{{ flowchart }}', flow,
1)
return ''
def get_highlight(self):
return ''
def get_contents(self, filename):
''' Get contents or selection from view and optionally strip the YAML front matter '''
# TODO: 获取文件内容，从这里开始
contents = load_resource(filename)
# Remove yaml front matter
if self.settings.get(
'strip_yaml_front_matter') and contents.startswith('---'):
frontmatter, contents = self.preprocessor_yaml_frontmatter(
contents)
self.settings.apply_frontmatter(frontmatter)
references = self.settings.get('builtin').get('references', [])
for ref in references:
contents += get_references(ref)
contents = self.parser_specific_preprocess(contents)
return contents
def parser_specific_preprocess(self, text):
return text
def preprocessor_yaml_frontmatter(self, text):
""" Get frontmatter from string """
frontmatter = {}
if text.startswith("---"):
m = re.search(r'^(---(.*?)---\r?\n)', text, re.DOTALL)
if m:
try:
frontmatter = yaml.load(m.group(2))
except:
print(traceback.format_exc())
text = text[m.end(1):]
return frontmatter, text
def parser_specific_postprocess(self, text):
return text
def postprocessor_absolute(self, html, image_convert, file_convert):
''' fix relative paths in images, scripts, and links for the internal parser '''
def tag_fix(m):
tag = m.group('tag')
src = m.group('src')
if self.filename:
# TODO: delete paltform = windows
if (not src.startswith(ABS_EXCLUDE)
and not (RE_WIN_DRIVE.match(src) is not None)):
# Don't explicitly add file:// prefix,
# But don't remove them either
abs_path = u'%s/%s' % (os.path.dirname(self.filename), src)
# Don't replace just the first instance,
# but explicitly place it where it was before
# to ensure a file name 'img' dosen't replace
# the tag name etc.
if os.path.exists(abs_path):
tag = m.group('begin') + abs_path + m.group('end')
return tag
# Compile the appropriate regex to find images and/or files
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_SOURCES = re.compile(
r"""(?P<tag>(?P<begin><(?:%s%s%s)[^>]+(?:src%s)=["'])(?P<src>[^"']+)(?P<end>[^>]*>))"""
% (r"img" if image_convert else "",
r"|" if image_convert and file_convert else "", r"script|a"
if file_convert else "", r"|href" if file_convert else ""))
return RE_SOURCES.sub(tag_fix, html)
def postprocessor_pathconverter(self,
html,
image_convert,
file_convert,
absolute=False):
RE_TAG_HTML = r'''(?xus)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><(?P<tag>(?:%s)))
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
'''
RE_TAG_LINK_ATTR = re.compile(r'''(?xus)
(?P<attr>
(?:
(?P<name>\s+(?:href|src)\s*=\s*)
(?P<path>"[^"]*"|'[^']*')
)
)
''')
RE_SOURCES = re.compile(
RE_TAG_HTML % ((r"img" if image_convert else "") +
(r"|" if image_convert and file_convert else "") +
(r"script|a|link" if file_convert else "")))
def repl(m, base_path, rel_path=None):
if m.group('comments'):
tag = m.group('comments')
else:
tag = m.group('open')
if rel_path is None:
tag += RE_TAG_LINK_ATTR.sub(
lambda m2: repl_absolute(m2, base_path),
m.group('attr'))
else:
tag += RE_TAG_LINK_ATTR.sub(
lambda m2: repl_relative(m2, base_path, rel_path),
m.group('attr'))
tag += m.group('close')
return tag
basepath = self.settings.get('builtin').get("basepath")
if basepath is None:
basepath = ""
if absolute:
if basepath:
return RE_SOURCES.sub(lambda m: repl(m, basepath), html)
else:
if self.preview:
relativepath = getTempMarkdownPreviewPath(self.view)
else:
relativepath = self.settings.get('builtin').get("destination")
if not relativepath:
mdfile = self.view.file_name()
if mdfile is not None and os.path.exists(mdfile):
relativepath = os.path.splitext(mdfile)[0] + '.html'
if relativepath:
relativepath = os.path.dirname(relativepath)
if basepath and relativepath:
return RE_SOURCES.sub(
lambda m: repl(m, basepath, relativepath), html)
return html
def postprocessor_base64(self, html):
''' convert resources (currently images only) to base64 '''
file_types = {
".png": "image/png",
".jpg": "image/jpeg",
".jpeg": "image/jpeg",
".gif": "image/gif"
}
exclusion_list = tuple(
['https://', 'http://', '#'] +
["data:%s;base64," % ft for ft in file_types.values()])
def b64(m):
import base64
src = m.group('src')
data = m.group('tag')
base_path = self.settings.get("basepath")
if base_path is None:
base_path = ""
# Format the link
absolute = False
if src.startswith('file://'):
src = src.replace('file://', '', 1)
# if sublime.platform() == "windows" and not src.startswith('//'):
# src = src.lstrip("/")
absolute = True
# elif sublime.platform() == "windows" and RE_WIN_DRIVE.match(src) is not None:
# absolute = True
# Make sure we are working with an absolute path
if not src.startswith(exclusion_list):
if absolute:
src = os.path.normpath(src)
else:
src = os.path.normpath(os.path.join(base_path, src))
if os.path.exists(src):
ext = os.path.splitext(src)[1].lower()
if ext in file_types:
try:
with open(src, "rb") as f:
data = m.group(
'begin') + "data:%s;base64,%s" % (
file_types[ext],
base64.b64encode(f.read()).decode(
'ascii')) + m.group('end')
except Exception:
pass
return data
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_SOURCES = re.compile(
r"""(?P<tag>(?P<begin><(?:img)[^>]+(?:src)=["'])(?P<src>[^"']+)(?P<end>[^>]*>))"""
)
return RE_SOURCES.sub(b64, html)
def postprocessor_simple(self, html):
''' Strip out ids and classes for a simplified HTML output '''
def strip_html(m):
tag = m.group('open')
if m.group('attr1'):
tag += m.group('attr1')
if m.group('attr2'):
tag += m.group('attr2')
if m.group('attr3'):
tag += m.group('attr3')
if m.group('attr4'):
tag += m.group('attr4')
if m.group('attr5'):
tag += m.group('attr5')
if m.group('attr6'):
tag += m.group('attr6')
tag += m.group('close')
return tag
# Strip out id, class and style attributes for a simple html output
# Since we are stripping out two attributes, we need to set up the groups in such
# a way so we can retrieve the data we don't want to throw away
# up to these worst case scenarios:
#
# <tag attr=""... (id|class|style)=""... attr=""... (id|class|style)=""... attr=""... (id|class|style)=""...>
# <tag (id|class|style)=""... attr=""... (id|class|style)=""... attr=""... (id|class|style)=""... attr=""...>
STRIP_HTML = re.compile(
r'''
(?P<open><[\w\:\.\-]+) # Tag open
(?:
(?P<attr1>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target1>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)
(?:
(?P<attr2>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target2>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr3>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target3>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr4>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target4>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr5>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target5>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr6>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target6>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?P<close>\s*(?:\/?)>) # Tag end
''', re.MULTILINE | re.DOTALL | re.VERBOSE)
return STRIP_HTML.sub(strip_html, html)
def convert_markdown(self, markdown_text):
''' convert input markdown to HTML, with github or builtin parser '''
markdown_html = self.parser_specific_convert(markdown_text)
image_convert = self.settings.get("image_path_conversion", "absolute")
file_convert = self.settings.get("file_path_conversions", "absolute")
markdown_html = self.parser_specific_postprocess(markdown_html)
if "absolute" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(
markdown_html, image_convert, file_convert, True)
if "relative" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(
markdown_html, image_convert, file_convert, False)
if image_convert == "base64":
markdown_html = self.postprocessor_base64(markdown_html)
if self.settings.get("html_simple", False):
markdown_html = self.postprocessor_simple(markdown_html)
return markdown_html
def get_title(self):
if self.meta_title is not None:
title = self.meta_title
else:
title = self.filename
# TODO: I changed here
if not title:
fn = self.filename
title = 'untitled' if not fn else os.path.splitext(
os.path.basename(fn))[0]
return '<title>%s</title>' % cgi.escape(title)
def get_meta(self):
self.meta_title = None
meta = []
for k, v in self.settings.get("meta", {}).items():
if k == "title":
if isinstance(v, list):
if len(v) == 0:
v = ""
else:
v = v[0]
self.meta_title = unicode_str(v)
continue
if isinstance(v, list):
v = ','.join(v)
if v is not None:
meta.append('<meta name="%s" content="%s">' %
(cgi.escape(k, True), cgi.escape(v, True)))
return '\n'.join(meta)
# 主程序入口
def run(self):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings',
self.filename)
contents = self.get_contents(self.filename)
body = self.convert_markdown(contents)
# add self.html_template, it can be assigned.
if self.html_template is None:
html_template = self.settings.get('html_template')
else:
html_template = self.html_template
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_title()
html += '</head><body>'
html += body
html += '</body>'
html += '</html>'
return html, body
class Compiler(object):
''' Do the markdown converting '''
default_css = "markdown.css"
def isurl(self, css_name):
match = re.match(r'https?://', css_name)
if match:
return True
return False
def get_default_css(self):
''' locate the correct CSS with the 'css' setting '''
css_name = self.settings.get('css', 'default')
if self.isurl(css_name):
# link to remote URL
return u"<link href='%s' rel='stylesheet' type='text/css'>" % css_name
elif os.path.isfile(os.path.expanduser(css_name)):
# use custom CSS file
return u"<style>%s</style>" % load_utf8(
os.path.expanduser(css_name))
elif css_name == 'default':
# use parser CSS file
return u"<style>%s</style>" % load_resource(self.default_css)
return ''
def get_override_css(self):
''' handls allow_css_overrides setting. '''
if self.settings.get('allow_css_overrides'):
filename = self.view.file_name()
filetypes = self.settings.get('markdown_filetypes')
if filename and filetypes:
for filetype in filetypes:
if filename.endswith(filetype):
css_filename = filename.rpartition(
filetype)[0] + '.css'
if (os.path.isfile(css_filename)):
return u"<style>%s</style>" % load_utf8(
css_filename)
return ''
def get_stylesheet(self):
''' return the correct CSS file based on parser and settings '''
return self.get_default_css() + self.get_override_css()
def get_javascript(self):
js_files = self.settings.get('js')
scripts = ''
if js_files is not None:
# Ensure string values become a list.
if isinstance(js_files, str) or isinstance(js_files, unicode_str):
js_files = [js_files]
# Only load scripts if we have a list.
if isinstance(js_files, list):
for js_file in js_files:
if os.path.isabs(js_file):
# Load the script inline to avoid cross-origin.
scripts += u"<script>%s</script>" % load_utf8(js_file)
else:
scripts += u"<script type='text/javascript' src='%s'></script>" % js_file
return scripts
def get_mathjax(self):
''' return the MathJax script if enabled '''
if self.settings.get('enable_mathjax') is True:
return load_resource('mathjax.html')
return ''
def get_uml(self):
''' return the uml scripts if enabled '''
if self.settings.get('enable_uml') is True:
flow = load_resource('flowchart-min.js')
return load_resource('uml.html').replace('{{ flowchart }}', flow,
1)
return ''
def get_highlight(self):
return ''
def get_contents(self, wholefile=False):
''' Get contents or selection from view and optionally strip the YAML front matter '''
region = sublime.Region(0, self.view.size())
contents = self.view.substr(region)
if not wholefile:
# use selection if any
selection = self.view.substr(self.view.sel()[0])
if selection.strip() != '':
contents = selection
# Remove yaml front matter
if self.settings.get(
'strip_yaml_front_matter') and contents.startswith('---'):
frontmatter, contents = self.preprocessor_yaml_frontmatter(
contents)
self.settings.apply_frontmatter(frontmatter)
references = self.settings.get('builtin').get('references', [])
for ref in references:
contents += get_references(ref)
contents = self.parser_specific_preprocess(contents)
return contents
def parser_specific_preprocess(self, text):
return text
def preprocessor_yaml_frontmatter(self, text):
""" Get frontmatter from string """
frontmatter = {}
if text.startswith("---"):
m = re.search(r'^(---(.*?)---[ \t]*\r?\n)', text, re.DOTALL)
if m:
try:
frontmatter = yaml.load(m.group(2))
except:
print(traceback.format_exc())
text = text[m.end(1):]
return frontmatter, text
def parser_specific_postprocess(self, text):
return text
def postprocessor_pathconverter(self,
html,
image_convert,
file_convert,
absolute=False):
RE_TAG_HTML = r'''(?xus)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><(?P<tag>(?:%s)))
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
'''
RE_TAG_LINK_ATTR = re.compile(r'''(?xus)
(?P<attr>
(?:
(?P<name>\s+(?:href|src)\s*=\s*)
(?P<path>"[^"]*"|'[^']*')
)
)
''')
RE_SOURCES = re.compile(
RE_TAG_HTML % ((r"img" if image_convert else "") +
(r"|" if image_convert and file_convert else "") +
(r"script|a|link" if file_convert else "")))
def repl(m, base_path, rel_path=None):
if m.group('comments'):
tag = m.group('comments')
else:
tag = m.group('open')
if rel_path is None:
tag += RE_TAG_LINK_ATTR.sub(
lambda m2: repl_absolute(m2, base_path),
m.group('attr'))
else:
tag += RE_TAG_LINK_ATTR.sub(
lambda m2: repl_relative(m2, base_path, rel_path),
m.group('attr'))
tag += m.group('close')
return tag
basepath = self.settings.get('builtin').get("basepath")
if basepath is None:
basepath = ""
if absolute:
if basepath:
return RE_SOURCES.sub(lambda m: repl(m, basepath), html)
else:
if self.preview:
relativepath = getTempMarkdownPreviewPath(self.view)
else:
relativepath = self.settings.get('builtin').get("destination")
if not relativepath:
mdfile = self.view.file_name()
if mdfile is not None and os.path.exists(mdfile):
relativepath = os.path.splitext(mdfile)[0] + '.html'
if relativepath:
relativepath = os.path.dirname(relativepath)
if basepath and relativepath:
return RE_SOURCES.sub(
lambda m: repl(m, basepath, relativepath), html)
return html
def postprocessor_base64(self, html):
''' convert resources (currently images only) to base64 '''
file_types = {
(".png", ): "image/png",
(".jpg", ".jpeg"): "image/jpeg",
(".gif", ): "image/gif"
}
exclusion_list = tuple(
['https://', 'http://', '#'] +
["data:%s;base64," % ft for ft in file_types.values()])
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_TAG_HTML = re.compile(r'''(?xus)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><(?P<tag>img))
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
''')
RE_TAG_LINK_ATTR = re.compile(r'''(?xus)
(?P<attr>
(?:
(?P<name>\s+src\s*=\s*)
(?P<path>"[^"]*"|'[^']*')
)
)
''')
def b64(m):
import base64
src = url2pathname(m.group('path')[1:-1])
data = m.group(0)
base_path = self.settings.get('builtin').get("basepath")
if base_path is None:
base_path = ""
# Format the link
absolute = False
if src.startswith('file://'):
src = src.replace('file://', '', 1)
if sublime.platform(
) == "windows" and not src.startswith('//'):
src = src.lstrip("/")
absolute = True
elif sublime.platform() == "windows" and RE_WIN_DRIVE.match(
src) is not None:
absolute = True
# Make sure we are working with an absolute path
if not src.startswith(exclusion_list):
if absolute:
src = os.path.normpath(src)
else:
src = os.path.normpath(os.path.join(base_path, src))
if os.path.exists(src):
ext = os.path.splitext(src)[1].lower()
for b64_ext in file_types:
if ext in b64_ext:
try:
with open(src, "rb") as f:
data = " src=\"data:%s;base64,%s\"" % (
file_types[b64_ext],
base64.b64encode(
f.read()).decode('ascii'))
except Exception:
pass
break
return data
def repl(m):
if m.group('comments'):
tag = m.group('comments')
else:
tag = m.group('open')
tag += RE_TAG_LINK_ATTR.sub(lambda m2: b64(m2),
m.group('attr'))
tag += m.group('close')
return tag
return RE_TAG_HTML.sub(repl, html)
def postprocessor_simple(self, html):
''' Strip out ids and classes for a simplified HTML output '''
def repl(m):
if m.group('comments'):
tag = ''
else:
tag = m.group('open')
tag += RE_TAG_BAD_ATTR.sub('', m.group('attr'))
tag += m.group('close')
return tag
# Strip out id, class, on<word>, and style attributes for a simple html output
RE_TAG_HTML = re.compile(
r'''(?x)
(?:
(?P<comments>(\r?\n?\s*)<!--[\s\S]*?-->(\s*)(?=\r?\n)|<!--[\s\S]*?-->)|
(?P<open><[\w\:\.\-]+)
(?P<attr>(?:\s+[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*)
(?P<close>\s*(?:\/?)>)
)
''', re.DOTALL | re.UNICODE)
RE_TAG_BAD_ATTR = re.compile(
r'''(?x)
(?P<attr>
(?:
\s+(?:id|class|style|on[\w]+)
(?:\s*=\s*(?:"[^"]*"|'[^']*'))
)*
)
''', re.DOTALL | re.UNICODE)
return RE_TAG_HTML.sub(repl, html)
def convert_markdown(self, markdown_text):
''' convert input markdown to HTML, with github or builtin parser '''
markdown_html = self.parser_specific_convert(markdown_text)
image_convert = self.settings.get("image_path_conversion", "absolute")
file_convert = self.settings.get("file_path_conversions", "absolute")
markdown_html = self.parser_specific_postprocess(markdown_html)
if "absolute" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(
markdown_html, image_convert, file_convert, True)
if "relative" in (image_convert, file_convert):
markdown_html = self.postprocessor_pathconverter(
markdown_html, image_convert, file_convert, False)
if image_convert == "base64":
markdown_html = self.postprocessor_base64(markdown_html)
if self.settings.get("html_simple", False):
markdown_html = self.postprocessor_simple(markdown_html)
return markdown_html
def get_title(self):
if self.meta_title is not None:
title = self.meta_title
else:
title = self.view.name()
if not title:
fn = self.view.file_name()
title = 'untitled' if not fn else os.path.splitext(
os.path.basename(fn))[0]
return '<title>%s</title>' % cgi.escape(title)
def get_meta(self):
self.meta_title = None
meta = []
for k, v in self.settings.get("meta", {}).items():
if k == "title":
if isinstance(v, list):
if len(v) == 0:
v = ""
else:
v = v[0]
self.meta_title = unicode_str(v)
continue
if isinstance(v, list):
v = ','.join(v)
if v is not None:
meta.append('<meta name="%s" content="%s">' %
(cgi.escape(k, True), cgi.escape(v, True)))
return '\n'.join(meta)
def run(self, view, wholefile=False, preview=False):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings',
view.file_name())
self.preview = preview
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get('html_template')
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_uml()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_uml()
html += self.get_title()
html += '</head><body>'
html += '<article class="markdown-body">'
html += body
html += '</article>'
html += '</body>'
html += '</html>'
return html, body
class Compiler(object):
''' Do the markdown converting '''
default_css = "markdown.css"
def isurl(self, css_name):
match = re.match(r'https?://', css_name)
if match:
return True
return False
def get_default_css(self):
''' locate the correct CSS with the 'css' setting '''
css_name = self.settings.get('css', 'default')
if self.isurl(css_name):
# link to remote URL
return u"<link href='%s' rel='stylesheet' type='text/css'>" % css_name
elif os.path.isfile(os.path.expanduser(css_name)):
# use custom CSS file
return u"<style>%s</style>" % load_utf8(os.path.expanduser(css_name))
elif css_name == 'default':
# use parser CSS file
return u"<style>%s</style>" % load_resource(self.default_css)
return ''
def get_override_css(self):
''' handls allow_css_overrides setting. '''
if self.settings.get('allow_css_overrides'):
filename = self.view.file_name()
filetypes = self.settings.get('markdown_filetypes')
if filename and filetypes:
for filetype in filetypes:
if filename.endswith(filetype):
css_filename = filename.rpartition(filetype)[0] + '.css'
if (os.path.isfile(css_filename)):
return u"<style>%s</style>" % load_utf8(css_filename)
return ''
def get_stylesheet(self):
''' return the correct CSS file based on parser and settings '''
return self.get_default_css() + self.get_override_css()
def get_javascript(self):
js_files = self.settings.get('js')
scripts = ''
if js_files is not None:
# Ensure string values become a list.
if isinstance(js_files, str) or isinstance(js_files, unicode_str):
js_files = [js_files]
# Only load scripts if we have a list.
if isinstance(js_files, list):
for js_file in js_files:
if os.path.isabs(js_file):
# Load the script inline to avoid cross-origin.
scripts += u"<script>%s</script>" % load_utf8(js_file)
else:
scripts += u"<script type='text/javascript' src='%s'></script>" % js_file
return scripts
def get_mathjax(self):
''' return the MathJax script if enabled '''
if self.settings.get('enable_mathjax') is True:
return load_resource('mathjax.html')
return ''
def get_highlight(self):
return ''
def get_contents(self, wholefile=False):
''' Get contents or selection from view and optionally strip the YAML front matter '''
region = sublime.Region(0, self.view.size())
contents = self.view.substr(region)
if not wholefile:
# use selection if any
selection = self.view.substr(self.view.sel()[0])
if selection.strip() != '':
contents = selection
# Remove yaml front matter
if self.settings.get('strip_yaml_front_matter') and contents.startswith('---'):
frontmatter, contents = self.preprocessor_yaml_frontmatter(contents)
self.settings.apply_frontmatter(frontmatter)
references = self.settings.get('builtin').get('references', [])
for ref in references:
contents += get_references(ref)
contents = self.parser_specific_preprocess(contents)
return contents
def parser_specific_preprocess(self, text):
return text
def preprocessor_yaml_frontmatter(self, text):
""" Get frontmatter from string """
frontmatter = {}
if text.startswith("---"):
m = re.search(r'^(---(.*?)---\r?\n)', text, re.DOTALL)
if m:
try:
frontmatter = yaml.load(m.group(2))
except:
print(traceback.format_exc())
text = text[m.end(1):]
return frontmatter, text
def parser_specific_postprocess(self, text):
return text
def postprocessor_absolute(self, html, image_convert, file_convert):
''' fix relative paths in images, scripts, and links for the internal parser '''
def tag_fix(m):
tag = m.group('tag')
src = m.group('src')
filename = self.view.file_name()
if filename:
if (
not src.startswith(ABS_EXCLUDE) and
not (sublime.platform() == "windows" and RE_WIN_DRIVE.match(src) is not None)
):
# Don't explicitly add file:// prefix,
# But don't remove them either
abs_path = u'%s/%s' % (os.path.dirname(filename), src)
# Don't replace just the first instance,
# but explicitly place it where it was before
# to ensure a file name 'img' dosen't replace
# the tag name etc.
if os.path.exists(abs_path):
tag = m.group('begin') + abs_path + m.group('end')
return tag
# Compile the appropriate regex to find images and/or files
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_SOURCES = re.compile(
r"""(?P<tag>(?P<begin><(?:%s%s%s)[^>]+(?:src%s)=["'])(?P<src>[^"']+)(?P<end>[^>]*>))""" % (
r"img" if image_convert else "",
r"|" if image_convert and file_convert else "",
r"script|a" if file_convert else "",
r"|href" if file_convert else ""
)
)
return RE_SOURCES.sub(tag_fix, html)
def postprocessor_base64(self, html):
''' convert resources (currently images only) to base64 '''
file_types = {
".png": "image/png",
".jpg": "image/jpeg",
".jpeg": "image/jpeg",
".gif": "image/gif"
}
exclusion_list = tuple(
['https://', 'http://', '#'] +
["data:%s;base64," % ft for ft in file_types.values()]
)
def b64(m):
import base64
src = m.group('src')
data = m.group('tag')
base_path = self.settings.get("basepath")
if base_path is None:
base_path = ""
# Format the link
absolute = False
if src.startswith('file://'):
src = src.replace('file://', '', 1)
if sublime.platform() == "windows" and not src.startswith('//'):
src = src.lstrip("/")
absolute = True
elif sublime.platform() == "windows" and RE_WIN_DRIVE.match(src) is not None:
absolute = True
# Make sure we are working with an absolute path
if not src.startswith(exclusion_list):
if absolute:
src = os.path.normpath(src)
else:
src = os.path.normpath(os.path.join(base_path, src))
if os.path.exists(src):
ext = os.path.splitext(src)[1].lower()
if ext in file_types:
try:
with open(src, "rb") as f:
data = m.group('begin') + "data:%s;base64,%s" % (
file_types[ext],
base64.b64encode(f.read()).decode('ascii')
) + m.group('end')
except Exception:
pass
return data
RE_WIN_DRIVE = re.compile(r"(^[A-Za-z]{1}:(?:\\|/))")
RE_SOURCES = re.compile(r"""(?P<tag>(?P<begin><(?:img)[^>]+(?:src)=["'])(?P<src>[^"']+)(?P<end>[^>]*>))""")
return RE_SOURCES.sub(b64, html)
def postprocessor_simple(self, html):
''' Strip out ids and classes for a simplified HTML output '''
def strip_html(m):
tag = m.group('open')
if m.group('attr1'):
tag += m.group('attr1')
if m.group('attr2'):
tag += m.group('attr2')
if m.group('attr3'):
tag += m.group('attr3')
if m.group('attr4'):
tag += m.group('attr4')
if m.group('attr5'):
tag += m.group('attr5')
if m.group('attr6'):
tag += m.group('attr6')
tag += m.group('close')
return tag
# Strip out id, class and style attributes for a simple html output
# Since we are stripping out two attributes, we need to set up the groups in such
# a way so we can retrieve the data we don't want to throw away
# up to these worst case scenarios:
#
# <tag attr=""... (id|class|style)=""... attr=""... (id|class|style)=""... attr=""... (id|class|style)=""...>
# <tag (id|class|style)=""... attr=""... (id|class|style)=""... attr=""... (id|class|style)=""... attr=""...>
STRIP_HTML = re.compile(
r'''
(?P<open><[\w\:\.\-]+) # Tag open
(?:
(?P<attr1>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target1>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)
(?:
(?P<attr2>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target2>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr3>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target3>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr4>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target4>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr5>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target5>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?:
(?P<attr6>(?:\s+(?!id|class|style)[\w\-:]+(?:\s*=\s*(?:"[^"]*"|'[^']*'))?)*) # Attributes to keep
| (?P<target6>\s+(?:id|class|style)(?:\s*=\s*(?:"[^"]*"|'[^']*'))*) # Attributes to delte
)?
(?P<close>\s*(?:\/?)>) # Tag end
''',
re.MULTILINE | re.DOTALL | re.VERBOSE
)
return STRIP_HTML.sub(strip_html, html)
def convert_markdown(self, markdown_text):
''' convert input markdown to HTML, with github or builtin parser '''
markdown_html = self.parser_specific_convert(markdown_text)
image_convert = self.settings.get("image_path_conversion", "absolute")
file_convert = self.settings.get("file_path_conversions", "absolute")
markdown_html = self.parser_specific_postprocess(markdown_html)
if "absolute" in (image_convert, file_convert):
markdown_html = self.postprocessor_absolute(markdown_html, image_convert, file_convert)
if image_convert == "base64":
markdown_html = self.postprocessor_base64(markdown_html)
if self.settings.get("html_simple", False):
markdown_html = self.postprocessor_simple(markdown_html)
return markdown_html
def get_title(self):
if self.meta_title is not None:
title = self.meta_title
else:
title = self.view.name()
if not title:
fn = self.view.file_name()
title = 'untitled' if not fn else os.path.splitext(os.path.basename(fn))[0]
return '<title>%s</title>' % cgi.escape(title)
def get_meta(self):
self.meta_title = None
meta = []
for k, v in self.settings.get("meta", {}).items():
if k == "title":
if isinstance(v, list):
if len(v) == 0:
v = ""
else:
v = v[0]
self.meta_title = unicode_str(v)
continue
if isinstance(v, list):
v = ','.join(v)
if v is not None:
meta.append(
'<meta name="%s" content="%s">' % (cgi.escape(k, True), cgi.escape(v, True))
)
return '\n'.join(meta)
def run(self, view, wholefile=False):
''' return full html and body html for view. '''
self.settings = Settings('MarkdownPreview.sublime-settings', view.file_name())
self.view = view
contents = self.get_contents(wholefile)
body = self.convert_markdown(contents)
html_template = self.settings.get('html_template')
# use customized html template if given
if self.settings.get('html_simple', False):
html = body
elif html_template and os.path.exists(html_template):
head = u''
head += self.get_meta()
if not self.settings.get('skip_default_stylesheet'):
head += self.get_stylesheet()
head += self.get_javascript()
head += self.get_highlight()
head += self.get_mathjax()
head += self.get_title()
html = load_utf8(html_template)
html = html.replace('{{ HEAD }}', head, 1)
html = html.replace('{{ BODY }}', body, 1)
else:
html = u'<!DOCTYPE html>'
html += '<html><head><meta charset="utf-8">'
html += self.get_meta()
html += self.get_stylesheet()
html += self.get_javascript()
html += self.get_highlight()
html += self.get_mathjax()
html += self.get_title()
html += '</head><body>'
html += '<article class="markdown-body">'
html += body
html += '</article>'
html += '</body>'
html += '</html>'
return html, body
