git subrepo clone --branch=0.30.3 https://github.com/commonmark/cmark.git deps/cmark

subrepo: subdir: "deps/cmark" merged: "5ba25ff" upstream: origin: "https://github.com/commonmark/cmark.git" branch: "0.30.3" commit: "5ba25ff" git-subrepo: version: "0.4.6" commit: "d4444b563"
2023-09-03 20:42:10 -07:00
parent 17a4224cb8
commit 24810cbbbd
120 changed files with 48683 additions and 0 deletions
--- a/deps/cmark/test/CMakeLists.txt
+++ b/deps/cmark/test/CMakeLists.txt
@@ -0,0 +1,80 @@
+# To get verbose output: cmake --build build --target "test" -- ARGS='-V'
+
+# By default, we run the spec tests only if python3 is available.
+# To require the spec tests, compile with -DSPEC_TESTS=1
+
+if (SPEC_TESTS)
+  find_package(PythonInterp 3 REQUIRED)
+else(SPEC_TESTS)
+  find_package(PythonInterp 3)
+endif(SPEC_TESTS)
+
+if (CMARK_SHARED OR CMARK_STATIC)
+  add_test(NAME api_test COMMAND api_test)
+endif()
+
+if (WIN32)
+  file(TO_NATIVE_PATH ${CMAKE_BINARY_DIR}/src WIN_DLL_DIR)
+  set(NEWPATH "${WIN_DLL_DIR};$ENV{PATH}")
+  string(REPLACE ";" "\\;" NEWPATH "${NEWPATH}")
+  set_tests_properties(api_test PROPERTIES ENVIRONMENT "PATH=${NEWPATH}")
+  set(ROUNDTRIP "${CMAKE_CURRENT_SOURCE_DIR}/roundtrip.bat")
+else(WIN32)
+  set(ROUNDTRIP "${CMAKE_CURRENT_SOURCE_DIR}/roundtrip.sh")
+endif(WIN32)
+
+IF (PYTHONINTERP_FOUND)
+
+  add_test(html_normalization
+    ${PYTHON_EXECUTABLE} "-m" "doctest"
+    "${CMAKE_CURRENT_SOURCE_DIR}/normalize.py"
+    )
+
+  if (CMARK_SHARED)
+    add_test(spectest_library
+      ${PYTHON_EXECUTABLE} "${CMAKE_CURRENT_SOURCE_DIR}/spec_tests.py" "--no-normalize" "--spec"
+      "${CMAKE_CURRENT_SOURCE_DIR}/spec.txt" "--library-dir" "${CMAKE_CURRENT_BINARY_DIR}/../src"
+      )
+
+    add_test(pathological_tests_library
+      ${PYTHON_EXECUTABLE} "${CMAKE_CURRENT_SOURCE_DIR}/pathological_tests.py"
+      "--library-dir" "${CMAKE_CURRENT_BINARY_DIR}/../src"
+      )
+
+    add_test(roundtriptest_library
+      ${PYTHON_EXECUTABLE}
+      "${CMAKE_CURRENT_SOURCE_DIR}/roundtrip_tests.py"
+      "--spec" "${CMAKE_CURRENT_SOURCE_DIR}/spec.txt"
+      "--library-dir" "${CMAKE_CURRENT_BINARY_DIR}/../src"
+      )
+
+    add_test(entity_library
+      ${PYTHON_EXECUTABLE}
+      "${CMAKE_CURRENT_SOURCE_DIR}/entity_tests.py"
+      "--library-dir" "${CMAKE_CURRENT_BINARY_DIR}/../src"
+      )
+  endif()
+
+  add_test(spectest_executable
+    ${PYTHON_EXECUTABLE} "${CMAKE_CURRENT_SOURCE_DIR}/spec_tests.py" "--no-normalize" "--spec" "${CMAKE_CURRENT_SOURCE_DIR}/spec.txt" "--program" "${CMAKE_CURRENT_BINARY_DIR}/../src/cmark"
+    )
+
+  add_test(smartpuncttest_executable
+    ${PYTHON_EXECUTABLE} "${CMAKE_CURRENT_SOURCE_DIR}/spec_tests.py" "--no-normalize" "--spec" "${CMAKE_CURRENT_SOURCE_DIR}/smart_punct.txt" "--program" "${CMAKE_CURRENT_BINARY_DIR}/../src/cmark --smart"
+    )
+
+  add_test(regressiontest_executable
+    ${PYTHON_EXECUTABLE}
+    "${CMAKE_CURRENT_SOURCE_DIR}/spec_tests.py" "--no-normalize" "--spec"
+    "${CMAKE_CURRENT_SOURCE_DIR}/regression.txt" "--program"
+    "${CMAKE_CURRENT_BINARY_DIR}/../src/cmark"
+    )
+
+ELSE(PYTHONINTERP_FOUND)
+
+  message("\n*** A python 3 interpreter is required to run the spec tests.\n")
+  add_test(skipping_spectests
+    echo "Skipping spec tests, because no python 3 interpreter is available.")
+
+ENDIF(PYTHONINTERP_FOUND)
+
--- a/deps/cmark/test/afl_test_cases/test.md
+++ b/deps/cmark/test/afl_test_cases/test.md
@@ -0,0 +1,36 @@
+# H1
+
+H2
+--
+
+t ☺  
+*b* **em** `c`
+&ge;\&\
+\_e\_
+
+4) I1
+
+5) I2
+   > [l](/u "t")
+   >
+   > - [f]
+   > - ![a](/u "t")
+   >
+   >> <ftp://hh>
+   >> <u@hh>
+
+~~~ l☺
+cb
+~~~
+
+    c1
+    c2
+
+***
+
+<div>
+<b>x</b>
+</div>
+
+[f]: /u "t"
+
--- a/deps/cmark/test/cmark-fuzz.c
+++ b/deps/cmark/test/cmark-fuzz.c
@@ -0,0 +1,33 @@
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+#include "cmark.h"
+
+int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
+  struct __attribute__((packed)) {
+    int options;
+    int width;
+  } fuzz_config;
+
+  if (size >= sizeof(fuzz_config)) {
+    /* The beginning of `data` is treated as fuzzer configuration */
+    memcpy(&fuzz_config, data, sizeof(fuzz_config));
+
+    /* Mask off valid option bits */
+    fuzz_config.options &= (CMARK_OPT_SOURCEPOS | CMARK_OPT_HARDBREAKS | CMARK_OPT_UNSAFE | CMARK_OPT_NOBREAKS | CMARK_OPT_NORMALIZE | CMARK_OPT_VALIDATE_UTF8 | CMARK_OPT_SMART);
+
+    /* Remainder of input is the markdown */
+    const char *markdown = (const char *)(data + sizeof(fuzz_config));
+    const size_t markdown_size = size - sizeof(fuzz_config);
+    cmark_node *doc = cmark_parse_document(markdown, markdown_size, fuzz_config.options);
+
+    free(cmark_render_commonmark(doc, fuzz_config.options, fuzz_config.width));
+    free(cmark_render_html(doc, fuzz_config.options));
+    free(cmark_render_latex(doc, fuzz_config.options, fuzz_config.width));
+    free(cmark_render_man(doc, fuzz_config.options, fuzz_config.width));
+    free(cmark_render_xml(doc, fuzz_config.options));
+
+    cmark_node_free(doc);
+  }
+  return 0;
+}
--- a/deps/cmark/test/cmark.py
+++ b/deps/cmark/test/cmark.py
@@ -0,0 +1,62 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+from ctypes import CDLL, c_char_p, c_size_t, c_int, c_void_p
+from subprocess import *
+import platform
+import os
+
+def pipe_through_prog(prog, text):
+    p1 = Popen(prog.split(), stdout=PIPE, stdin=PIPE, stderr=PIPE)
+    [result, err] = p1.communicate(input=text.encode('utf-8'))
+    return [p1.returncode, result.decode('utf-8'), err]
+
+def to_html(lib, text):
+    markdown = lib.cmark_markdown_to_html
+    markdown.restype = c_char_p
+    markdown.argtypes = [c_char_p, c_size_t, c_int]
+    textbytes = text.encode('utf-8')
+    textlen = len(textbytes)
+    # 1 << 17 == CMARK_OPT_UNSAFE
+    result = markdown(textbytes, textlen, 1 << 17).decode('utf-8')
+    return [0, result, '']
+
+def to_commonmark(lib, text):
+    textbytes = text.encode('utf-8')
+    textlen = len(textbytes)
+    parse_document = lib.cmark_parse_document
+    parse_document.restype = c_void_p
+    parse_document.argtypes = [c_char_p, c_size_t, c_int]
+    render_commonmark = lib.cmark_render_commonmark
+    render_commonmark.restype = c_char_p
+    render_commonmark.argtypes = [c_void_p, c_int, c_int]
+    node = parse_document(textbytes, textlen, 0)
+    result = render_commonmark(node, 0, 0).decode('utf-8')
+    return [0, result, '']
+
+class CMark:
+    def __init__(self, prog=None, library_dir=None):
+        self.prog = prog
+        if prog:
+            prog += ' --unsafe'
+            self.to_html = lambda x: pipe_through_prog(prog, x)
+            self.to_commonmark = lambda x: pipe_through_prog(prog + ' -t commonmark', x)
+        else:
+            sysname = platform.system()
+            if sysname == 'Darwin':
+                libnames = [ "libcmark.dylib" ]
+            elif sysname == 'Windows':
+                libnames = [ "cmark.dll", "libcmark.dll" ]
+            else:
+                libnames = [ "libcmark.so" ]
+            if not library_dir:
+                library_dir = os.path.join("build", "src")
+            for libname in libnames:
+                candidate = os.path.join(library_dir, libname)
+                if os.path.isfile(candidate):
+                    libpath = candidate
+                    break
+            cmark = CDLL(libpath)
+            self.to_html = lambda x: to_html(cmark, x)
+            self.to_commonmark = lambda x: to_commonmark(cmark, x)
+
--- a/deps/cmark/test/entity_tests.py
+++ b/deps/cmark/test/entity_tests.py
@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import re
+import os
+import argparse
+import sys
+import platform
+import html
+from cmark import CMark
+
+def get_entities():
+    regex = r'^{\(unsigned char\*\)"([^"]+)", \{([^}]+)\}'
+    with open(os.path.join(os.path.dirname(__file__), '..', 'src', 'entities.inc')) as f:
+        code = f.read()
+    entities = []
+    for entity, utf8 in re.findall(regex, code, re.MULTILINE):
+        utf8 = bytes(map(int, utf8.split(", ")[:-1])).decode('utf-8')
+        entities.append((entity, utf8))
+    return entities
+
+parser = argparse.ArgumentParser(description='Run cmark tests.')
+parser.add_argument('--program', dest='program', nargs='?', default=None,
+        help='program to test')
+parser.add_argument('--library-dir', dest='library_dir', nargs='?',
+        default=None, help='directory containing dynamic library')
+args = parser.parse_args(sys.argv[1:])
+
+cmark = CMark(prog=args.program, library_dir=args.library_dir)
+
+entities = get_entities()
+
+passed = 0
+errored = 0
+failed = 0
+
+exceptions = {
+    'quot': '&quot;',
+    'QUOT': '&quot;',
+
+    # These are broken, but I'm not too worried about them.
+    'nvlt': '&lt;⃒',
+    'nvgt': '&gt;⃒',
+}
+
+print("Testing entities:")
+for entity, utf8 in entities:
+    [rc, actual, err] = cmark.to_html("&{};".format(entity))
+    check = exceptions.get(entity, utf8)
+
+    if rc != 0:
+        errored += 1
+        print(entity, '[ERRORED (return code {})]'.format(rc))
+        print(err)
+    elif check in actual:
+        # print(entity, '[PASSED]') # omit noisy success output
+        passed += 1
+    else:
+        print(entity, '[FAILED]')
+        print(repr(actual))
+        failed += 1
+
+print("{} passed, {} failed, {} errored".format(passed, failed, errored))
+if failed == 0 and errored == 0:
+    exit(0)
+else:
+    exit(1)
--- a/deps/cmark/test/fuzzing_dictionary
+++ b/deps/cmark/test/fuzzing_dictionary
@@ -0,0 +1,49 @@
+asterisk="*"
+attr_generic=" a=\"1\""
+attr_href=" href=\"1\""
+attr_xml_lang=" xml:lang=\"1\""
+attr_xmlns=" xmlns=\"1\""
+backslash="\\"
+backtick="`"
+colon=":"
+dashes="---"
+double_quote="\""
+entity_builtin="&lt;"
+entity_decimal="&#1;"
+entity_external="&a;"
+entity_hex="&#x1;"
+equals="==="
+exclamation="!"
+greater_than=">"
+hash="#"
+hyphen="-"
+indent="  "
+left_bracket="["
+left_paren="("
+less_than="<"
+plus="+"
+right_bracket="]"
+right_paren=")"
+single_quote="'"
+string_any="ANY"
+string_brackets="[]"
+string_cdata="CDATA"
+string_dashes="--"
+string_empty_dblquotes="\"\""
+string_empty_quotes="''"
+string_idrefs="IDREFS"
+string_parentheses="()"
+string_pcdata="#PCDATA"
+tag_cdata="<![CDATA["
+tag_close="</a>"
+tag_doctype="<!DOCTYPE"
+tag_element="<!ELEMENT"
+tag_entity="<!ENTITY"
+tag_notation="<!NOTATION"
+tag_open="<a>"
+tag_open_close="<a />"
+tag_open_exclamation="<!"
+tag_open_q="<?"
+tag_sq2_close="]]>"
+tag_xml_q="<?xml?>"
+underscore="_"
--- a/deps/cmark/test/normalize.py
+++ b/deps/cmark/test/normalize.py
@@ -0,0 +1,194 @@
+# -*- coding: utf-8 -*-
+from html.parser import HTMLParser
+import urllib
+
+try:
+    from html.parser import HTMLParseError
+except ImportError:
+    # HTMLParseError was removed in Python 3.5. It could never be
+    # thrown, so we define a placeholder instead.
+    class HTMLParseError(Exception):
+        pass
+
+from html.entities import name2codepoint
+import sys
+import re
+import html
+
+# Normalization code, adapted from
+# https://github.com/karlcow/markdown-testsuite/
+significant_attrs = ["alt", "href", "src", "title"]
+whitespace_re = re.compile('\s+')
+class MyHTMLParser(HTMLParser):
+    def __init__(self):
+        HTMLParser.__init__(self)
+        self.convert_charrefs = False
+        self.last = "starttag"
+        self.in_pre = False
+        self.output = ""
+        self.last_tag = ""
+    def handle_data(self, data):
+        after_tag = self.last == "endtag" or self.last == "starttag"
+        after_block_tag = after_tag and self.is_block_tag(self.last_tag)
+        if after_tag and self.last_tag == "br":
+            data = data.lstrip('\n')
+        if not self.in_pre:
+            data = whitespace_re.sub(' ', data)
+        if after_block_tag and not self.in_pre:
+            if self.last == "starttag":
+                data = data.lstrip()
+            elif self.last == "endtag":
+                data = data.strip()
+        self.output += data
+        self.last = "data"
+    def handle_endtag(self, tag):
+        if tag == "pre":
+            self.in_pre = False
+        elif self.is_block_tag(tag):
+            self.output = self.output.rstrip()
+        self.output += "</" + tag + ">"
+        self.last_tag = tag
+        self.last = "endtag"
+    def handle_starttag(self, tag, attrs):
+        if tag == "pre":
+            self.in_pre = True
+        if self.is_block_tag(tag):
+            self.output = self.output.rstrip()
+        self.output += "<" + tag
+        # For now we don't strip out 'extra' attributes, because of
+        # raw HTML test cases.
+        # attrs = filter(lambda attr: attr[0] in significant_attrs, attrs)
+        if attrs:
+            attrs.sort()
+            for (k,v) in attrs:
+                self.output += " " + k
+                if v in ['href','src']:
+                    self.output += ("=" + '"' +
+                            urllib.quote(urllib.unquote(v), safe='/') + '"')
+                elif v != None:
+                    self.output += ("=" + '"' + html.escape(v,quote=True) + '"')
+        self.output += ">"
+        self.last_tag = tag
+        self.last = "starttag"
+    def handle_startendtag(self, tag, attrs):
+        """Ignore closing tag for self-closing """
+        self.handle_starttag(tag, attrs)
+        self.last_tag = tag
+        self.last = "endtag"
+    def handle_comment(self, data):
+        self.output += '<!--' + data + '-->'
+        self.last = "comment"
+    def handle_decl(self, data):
+        self.output += '<!' + data + '>'
+        self.last = "decl"
+    def unknown_decl(self, data):
+        self.output += '<!' + data + '>'
+        self.last = "decl"
+    def handle_pi(self,data):
+        self.output += '<?' + data + '>'
+        self.last = "pi"
+    def handle_entityref(self, name):
+        try:
+            c = chr(name2codepoint[name])
+        except KeyError:
+            c = None
+        self.output_char(c, '&' + name + ';')
+        self.last = "ref"
+    def handle_charref(self, name):
+        try:
+            if name.startswith("x"):
+                c = chr(int(name[1:], 16))
+            else:
+                c = chr(int(name))
+        except ValueError:
+                c = None
+        self.output_char(c, '&' + name + ';')
+        self.last = "ref"
+    # Helpers.
+    def output_char(self, c, fallback):
+        if c == '<':
+            self.output += "&lt;"
+        elif c == '>':
+            self.output += "&gt;"
+        elif c == '&':
+            self.output += "&amp;"
+        elif c == '"':
+            self.output += "&quot;"
+        elif c == None:
+            self.output += fallback
+        else:
+            self.output += c
+
+    def is_block_tag(self,tag):
+        return (tag in ['article', 'header', 'aside', 'hgroup', 'blockquote',
+            'hr', 'iframe', 'body', 'li', 'map', 'button', 'object', 'canvas',
+            'ol', 'caption', 'output', 'col', 'p', 'colgroup', 'pre', 'dd',
+            'progress', 'div', 'section', 'dl', 'table', 'td', 'dt',
+            'tbody', 'embed', 'textarea', 'fieldset', 'tfoot', 'figcaption',
+            'th', 'figure', 'thead', 'footer', 'tr', 'form', 'ul',
+            'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'video', 'script', 'style'])
+
+def normalize_html(html):
+    r"""
+    Return normalized form of HTML which ignores insignificant output
+    differences:
+
+    Multiple inner whitespaces are collapsed to a single space (except
+    in pre tags):
+
+        >>> normalize_html("<p>a  \t b</p>")
+        '<p>a b</p>'
+
+        >>> normalize_html("<p>a  \t\nb</p>")
+        '<p>a b</p>'
+
+    * Whitespace surrounding block-level tags is removed.
+
+        >>> normalize_html("<p>a  b</p>")
+        '<p>a b</p>'
+
+        >>> normalize_html(" <p>a  b</p>")
+        '<p>a b</p>'
+
+        >>> normalize_html("<p>a  b</p> ")
+        '<p>a b</p>'
+
+        >>> normalize_html("\n\t<p>\n\t\ta  b\t\t</p>\n\t")
+        '<p>a b</p>'
+
+        >>> normalize_html("<i>a  b</i> ")
+        '<i>a b</i> '
+
+    * Self-closing tags are converted to open tags.
+
+        >>> normalize_html("<br />")
+        '<br>'
+
+    * Attributes are sorted and lowercased.
+
+        >>> normalize_html('<a title="bar" HREF="foo">x</a>')
+        '<a href="foo" title="bar">x</a>'
+
+    * References are converted to unicode, except that '<', '>', '&', and
+      '"' are rendered using entities.
+
+        >>> normalize_html("&forall;&amp;&gt;&lt;&quot;")
+        '\u2200&amp;&gt;&lt;&quot;'
+
+    """
+    html_chunk_re = re.compile("(\<!\[CDATA\[.*?\]\]\>|\<[^>]*\>|[^<]+)")
+    try:
+        parser = MyHTMLParser()
+        # We work around HTMLParser's limitations parsing CDATA
+        # by breaking the input into chunks and passing CDATA chunks
+        # through verbatim.
+        for chunk in re.finditer(html_chunk_re, html):
+            if chunk.group(0)[:8] == "<![CDATA":
+                parser.output += chunk.group(0)
+            else:
+                parser.feed(chunk.group(0))
+        parser.close()
+        return parser.output
+    except HTMLParseError as e:
+        sys.stderr.write("Normalization error: " + e.msg + "\n")
+        return html  # on error, return unnormalized HTML
--- a/deps/cmark/test/pathological_tests.py
+++ b/deps/cmark/test/pathological_tests.py
@@ -0,0 +1,193 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import re
+import argparse
+import sys
+import platform
+import itertools
+import multiprocessing
+import queue
+import time
+from cmark import CMark
+
+TIMEOUT = 5
+
+parser = argparse.ArgumentParser(description='Run cmark tests.')
+parser.add_argument('--program', dest='program', nargs='?', default=None,
+        help='program to test')
+parser.add_argument('--library-dir', dest='library_dir', nargs='?',
+        default=None, help='directory containing dynamic library')
+args = parser.parse_args(sys.argv[1:])
+
+allowed_failures = {"many references": True}
+
+cmark = CMark(prog=args.program, library_dir=args.library_dir)
+
+def hash_collisions():
+    REFMAP_SIZE = 16
+    COUNT = 25000
+
+    def badhash(ref):
+        h = 0
+        for c in ref:
+            a = (h << 6) & 0xFFFFFFFF
+            b = (h << 16) & 0xFFFFFFFF
+            h = ord(c) + a + b - h
+            h = h & 0xFFFFFFFF
+
+        return (h % REFMAP_SIZE) == 0
+
+    keys = ("x%d" % i for i in itertools.count())
+    collisions = itertools.islice((k for k in keys if badhash(k)), COUNT)
+    bad_key = next(collisions)
+
+    document = ''.join("[%s]: /url\n\n[%s]\n\n" % (key, bad_key) for key in collisions)
+
+    return document, re.compile("(<p>\[%s\]</p>\n){%d}" % (bad_key, COUNT-1))
+
+
+# list of pairs consisting of input and a regex that must match the output.
+pathological = {
+    # note - some pythons have limit of 65535 for {num-matches} in re.
+    "nested strong emph":
+                (("*a **a " * 32500) + "b" + (" a** a*" * 32500),
+                 re.compile("(<em>a <strong>a ){32500}b( a</strong> a</em>){32500}")),
+    "many emph closers with no openers":
+                 (("a_ " * 32500),
+                  re.compile("(a[_] ){32499}a_")),
+    "many emph openers with no closers":
+                 (("_a " * 32500),
+                  re.compile("(_a ){32499}_a")),
+    "many link closers with no openers":
+                 (("a]" * 32500),
+                  re.compile("(a\]){32500}")),
+    "many link openers with no closers":
+                 (("[a" * 32500),
+                  re.compile("(\[a){32500}")),
+    "mismatched openers and closers":
+                 (("*a_ " * 25000),
+                  re.compile("([*]a[_] ){24999}[*]a_")),
+    "issue #389":
+                 (("*a " * 20000 + "_a*_ " * 20000),
+                  re.compile("(<em>a ){20000}(_a<\/em>_ ?){20000}")),
+    "openers and closers multiple of 3":
+                 (("a**b" + ("c* " * 25000)),
+                  re.compile("a[*][*]b(c[*] ){24999}c[*]")),
+    "link openers and emph closers":
+                 (("[ a_" * 25000),
+                  re.compile("(\[ a_){25000}")),
+    "pattern [ (]( repeated":
+                 (("[ (](" * 40000),
+                  re.compile("(\[ \(\]\(){40000}")),
+    "pattern ![[]() repeated":
+                 ("![[]()" * 160000,
+                  re.compile("(!\[<a href=\"\"></a>){160000}")),
+    "hard link/emph case":
+                 ("**x [a*b**c*](d)",
+                  re.compile("\\*\\*x <a href=\"d\">a<em>b\\*\\*c</em></a>")),
+    "nested brackets":
+                 (("[" * 25000) + "a" + ("]" * 25000),
+                  re.compile("\[{25000}a\]{25000}")),
+    "nested block quotes":
+                 ((("> " * 25000) + "a"),
+                  re.compile("(<blockquote>\n){25000}")),
+    "deeply nested lists":
+                 ("".join(map(lambda x: ("  " * x + "* a\n"), range(0,500))),
+                  re.compile("<ul>\n(<li>a\n<ul>\n){499}<li>a</li>\n</ul>\n(</li>\n</ul>\n){499}")),
+    "U+0000 in input":
+                 ("abc\u0000de\u0000",
+                  re.compile("abc\ufffd?de\ufffd?")),
+    "backticks":
+                 ("".join(map(lambda x: ("e" + "`" * x), range(1,2500))),
+                  re.compile("^<p>[e`]*</p>\n$")),
+    "unclosed links A":
+                 ("[a](<b" * 30000,
+                  re.compile("(\[a\]\(&lt;b){30000}")),
+    "unclosed links B":
+                 ("[a](b" * 30000,
+                  re.compile("(\[a\]\(b){30000}")),
+    "unclosed <!--":
+                 ("</" + "<!--" * 300000,
+                  re.compile("\&lt;\/(\&lt;!--){300000}")),
+    "reference collisions": hash_collisions()
+#    "many references":
+#                 ("".join(map(lambda x: ("[" + str(x) + "]: u\n"), range(1,5000 * 16))) + "[0] " * 5000,
+#                  re.compile("(\[0\] ){4999}"))
+    }
+
+pathological_cmark = {
+    "nested inlines":
+                 ("*" * 20000 + "a" + "*" * 20000,
+                  re.compile("^\*+a\*+$")),
+    }
+
+whitespace_re = re.compile('/s+/')
+
+def run_pathological(q, inp):
+    q.put(cmark.to_html(inp))
+
+def run_pathological_cmark(q, inp):
+    q.put(cmark.to_commonmark(inp))
+
+def run_tests():
+    q = multiprocessing.Queue()
+    passed = []
+    errored = []
+    failed = []
+    ignored = []
+
+    print("Testing pathological cases:")
+    for description in (*pathological, *pathological_cmark):
+        if description in pathological:
+            (inp, regex) = pathological[description]
+            p = multiprocessing.Process(target=run_pathological,
+                      args=(q, inp))
+        else:
+            (inp, regex) = pathological_cmark[description]
+            p = multiprocessing.Process(target=run_pathological_cmark,
+                      args=(q, inp))
+        p.start()
+        try:
+            # wait TIMEOUT seconds or until it finishes
+            rc, actual, err = q.get(True, TIMEOUT)
+            p.join()
+            if rc != 0:
+                print(description, '[ERRORED (return code %d)]' %rc)
+                print(err)
+                if description in allowed_failures:
+                    ignored.append(description)
+                else:
+                    errored.append(description)
+            elif regex.search(actual):
+                print(description, '[PASSED]')
+                passed.append(description)
+            else:
+                print(description, '[FAILED]')
+                print(repr(actual[:60]))
+                if description in allowed_failures:
+                    ignored.append(description)
+                else:
+                    failed.append(description)
+        except queue.Empty:
+            p.terminate()
+            p.join()
+            print(description, '[TIMEOUT]')
+            if description in allowed_failures:
+                ignored.append(description)
+            else:
+                errored.append(description)
+
+    print("%d passed, %d failed, %d errored" %
+          (len(passed), len(failed), len(errored)))
+    if ignored:
+        print("Ignoring these allowed failures:")
+        for x in ignored:
+            print(x)
+    if failed or errored:
+        exit(1)
+    else:
+        exit(0)
+
+if __name__ == "__main__":
+    run_tests()
--- a/deps/cmark/test/regression.txt
+++ b/deps/cmark/test/regression.txt
@@ -0,0 +1,215 @@
+### Regression tests
+
+Issue #113: EOL character weirdness on Windows
+(Important: first line ends with CR + CR + LF)
+
+```````````````````````````````` example
+line1
+
+line2
+.
+<p>line1</p>
+<p>line2</p>
+````````````````````````````````
+
+Issue #114: cmark skipping first character in line
+(Important: the blank lines around "Repeatedly" contain a tab.)
+
+```````````````````````````````` example
+By taking it apart
+
+- alternative solutions
+→
+Repeatedly solving
+→
+- how techniques
+.
+<p>By taking it apart</p>
+<ul>
+<li>alternative solutions</li>
+</ul>
+<p>Repeatedly solving</p>
+<ul>
+<li>how techniques</li>
+</ul>
+````````````````````````````````
+
+Issue jgm/CommonMark#430:  h2..h6 not recognized as block tags.
+
+```````````````````````````````` example
+<h1>lorem</h1>
+
+<h2>lorem</h2>
+
+<h3>lorem</h3>
+
+<h4>lorem</h4>
+
+<h5>lorem</h5>
+
+<h6>lorem</h6>
+.
+<h1>lorem</h1>
+<h2>lorem</h2>
+<h3>lorem</h3>
+<h4>lorem</h4>
+<h5>lorem</h5>
+<h6>lorem</h6>
+````````````````````````````````
+
+Issue jgm/commonmark.js#109 - tabs after setext header line
+
+
+```````````````````````````````` example
+hi
+--→
+.
+<h2>hi</h2>
+````````````````````````````````
+
+Issue #177 - incorrect emphasis parsing
+
+```````````````````````````````` example
+a***b* c*
+.
+<p>a*<em><em>b</em> c</em></p>
+````````````````````````````````
+
+Issue #193 - unescaped left angle brackets in link destination
+
+```````````````````````````````` example
+[a]
+
+[a]: <te<st>
+.
+<p>[a]</p>
+<p>[a]: &lt;te<st></p>
+````````````````````````````````
+
+Issue #192 - escaped spaces in link destination
+
+
+```````````````````````````````` example
+[a](te\ st)
+.
+<p>[a](te\ st)</p>
+````````````````````````````````
+
+Issue #527 - meta tags in inline contexts
+
+```````````````````````````````` example
+City:
+<span itemprop="contentLocation" itemscope itemtype="https://schema.org/City">
+  <meta itemprop="name" content="Springfield">
+</span>
+.
+<p>City:
+<span itemprop="contentLocation" itemscope itemtype="https://schema.org/City">
+<meta itemprop="name" content="Springfield">
+</span></p>
+````````````````````````````````
+
+Issue #530 - link parsing corner cases
+
+```````````````````````````````` example
+[a](\ b)
+
+[a](<<b)
+
+[a](<b
+)
+.
+<p>[a](\ b)</p>
+<p>[a](&lt;&lt;b)</p>
+<p>[a](&lt;b
+)</p>
+````````````````````````````````
+
+Issue commonmark#526 - unescaped ( in link title
+
+```````````````````````````````` example
+[link](url ((title))
+.
+<p>[link](url ((title))</p>
+````````````````````````````````
+
+Issue commonamrk#517 - script, pre, style close tag without
+opener.
+
+```````````````````````````````` example
+</script>
+
+</pre>
+
+</style>
+.
+</script>
+</pre>
+</style>
+````````````````````````````````
+
+Issue #289.
+
+```````````````````````````````` example
+[a](<b) c>
+.
+<p>[a](&lt;b) c&gt;</p>
+````````````````````````````````
+
+Issue #334 - UTF-8 BOM
+
+```````````````````````````````` example
+# Hi
+.
+<h1>Hi</h1>
+````````````````````````````````
+
+Issue commonmark.js#213 - type 7 blocks can't interrupt
+paragraph
+
+```````````````````````````````` example
+- <script>
+- some text
+some other text
+</script>
+.
+<ul>
+<li>
+<script>
+</li>
+<li>some text
+some other text
+</script></li>
+</ul>
+````````````````````````````````
+
+Issue #383 - emphasis parsing.
+
+```````````````````````````````` example
+*****Hello*world****
+.
+<p>**<em><strong>Hello<em>world</em></strong></em></p>
+````````````````````````````````
+
+Issue #424 - emphasis before links
+
+```````````````````````````````` example
+*text* [link](#section)
+.
+<p><em>text</em> <a href="#section">link</a></p>
+````````````````````````````````
+
+`<!doctype` is case-insensitive
+```````````````````````````````` example
+<!docType html>
+.
+<!docType html>
+````````````````````````````````
+
+Declarations don't need spaces, according to the spec
+```````````````````````````````` example
+x <!A>
+.
+<p>x <!A></p>
+````````````````````````````````
+
--- a/deps/cmark/test/roundtrip_tests.py
+++ b/deps/cmark/test/roundtrip_tests.py
@@ -0,0 +1,46 @@
+import re
+import sys
+from spec_tests import get_tests, do_test
+from cmark import CMark
+import argparse
+
+parser = argparse.ArgumentParser(description='Run cmark roundtrip tests.')
+parser.add_argument('-p', '--program', dest='program', nargs='?', default=None,
+        help='program to test')
+parser.add_argument('-s', '--spec', dest='spec', nargs='?', default='spec.txt',
+        help='path to spec')
+parser.add_argument('-P', '--pattern', dest='pattern', nargs='?',
+        default=None, help='limit to sections matching regex pattern')
+parser.add_argument('--library-dir', dest='library_dir', nargs='?',
+        default=None, help='directory containing dynamic library')
+parser.add_argument('--no-normalize', dest='normalize',
+        action='store_const', const=False, default=True,
+        help='do not normalize HTML')
+parser.add_argument('-n', '--number', type=int, default=None,
+        help='only consider the test with the given number')
+args = parser.parse_args(sys.argv[1:])
+
+spec = sys.argv[1]
+
+def converter(md):
+  cmark = CMark(prog=args.program, library_dir=args.library_dir)
+  [ec, result, err] = cmark.to_commonmark(md)
+  if ec == 0:
+    [ec, html, err] = cmark.to_html(result)
+    if ec == 0:
+        # In the commonmark writer we insert dummy HTML
+        # comments between lists, and between lists and code
+        # blocks.  Strip these out, since the spec uses
+        # two blank lines instead:
+        return [ec, re.sub('<!-- end list -->\n', '', html), '']
+    else:
+        return [ec, html, err]
+  else:
+    return [ec, result, err]
+
+tests = get_tests(args.spec)
+result_counts = {'pass': 0, 'fail': 0, 'error': 0, 'skip': 0}
+for test in tests:
+    do_test(converter, test, args.normalize, result_counts)
+
+exit(result_counts['fail'] + result_counts['error'])
--- a/deps/cmark/test/run-cmark-fuzz
+++ b/deps/cmark/test/run-cmark-fuzz
@@ -0,0 +1,4 @@
+#!/bin/bash -eu
+CMARK_FUZZ="$1"
+shift
+ASAN_OPTIONS="quarantine_size_mb=10:detect_leaks=1" "${CMARK_FUZZ}" -max_len=256 -timeout=1 -dict=test/fuzzing_dictionary "$@"
--- a/deps/cmark/test/smart_punct.txt
+++ b/deps/cmark/test/smart_punct.txt
@@ -0,0 +1,177 @@
+## Smart punctuation
+
+Open quotes are matched with closed quotes.
+The same method is used for matching openers and closers
+as is used in emphasis parsing:
+
+```````````````````````````````` example
+"Hello," said the spider.
+"'Shelob' is my name."
+.
+<p>“Hello,” said the spider.
+“‘Shelob’ is my name.”</p>
+````````````````````````````````
+
+```````````````````````````````` example
+'A', 'B', and 'C' are letters.
+.
+<p>‘A’, ‘B’, and ‘C’ are letters.</p>
+````````````````````````````````
+
+```````````````````````````````` example
+'Oak,' 'elm,' and 'beech' are names of trees.
+So is 'pine.'
+.
+<p>‘Oak,’ ‘elm,’ and ‘beech’ are names of trees.
+So is ‘pine.’</p>
+````````````````````````````````
+
+```````````````````````````````` example
+'He said, "I want to go."'
+.
+<p>‘He said, “I want to go.”’</p>
+````````````````````````````````
+
+A single quote that isn't an open quote matched
+with a close quote will be treated as an
+apostrophe:
+
+```````````````````````````````` example
+Were you alive in the 70's?
+.
+<p>Were you alive in the 70’s?</p>
+````````````````````````````````
+
+```````````````````````````````` example
+Here is some quoted '`code`' and a "[quoted link](url)".
+.
+<p>Here is some quoted ‘<code>code</code>’ and a “<a href="url">quoted link</a>”.</p>
+````````````````````````````````
+
+Here the first `'` is treated as an apostrophe, not
+an open quote, because the final single quote is matched
+by the single quote before `jolly`:
+
+```````````````````````````````` example
+'tis the season to be 'jolly'
+.
+<p>’tis the season to be ‘jolly’</p>
+````````````````````````````````
+
+Multiple apostrophes should not be marked as open/closing quotes.
+
+```````````````````````````````` example
+'We'll use Jane's boat and John's truck,' Jenna said.
+.
+<p>‘We’ll use Jane’s boat and John’s truck,’ Jenna said.</p>
+````````````````````````````````
+
+An unmatched double quote will be interpreted as a
+left double quote, to facilitate this style:
+
+```````````````````````````````` example
+"A paragraph with no closing quote.
+
+"Second paragraph by same speaker, in fiction."
+.
+<p>“A paragraph with no closing quote.</p>
+<p>“Second paragraph by same speaker, in fiction.”</p>
+````````````````````````````````
+
+A quote following a `]` or `)` character cannot
+be an open quote:
+
+```````````````````````````````` example
+[a]'s b'
+.
+<p>[a]’s b’</p>
+````````````````````````````````
+
+Quotes that are escaped come out as literal straight
+quotes:
+
+```````````````````````````````` example
+\"This is not smart.\"
+This isn\'t either.
+5\'8\"
+.
+<p>&quot;This is not smart.&quot;
+This isn't either.
+5'8&quot;</p>
+````````````````````````````````
+
+Two hyphens form an en-dash, three an em-dash.
+
+```````````````````````````````` example
+Some dashes:  em---em
+en--en
+em --- em
+en -- en
+2--3
+.
+<p>Some dashes:  em—em
+en–en
+em — em
+en – en
+2–3</p>
+````````````````````````````````
+
+A sequence of more than three hyphens is
+parsed as a sequence of em and/or en dashes,
+with no hyphens. If possible, a homogeneous
+sequence of dashes is used (so, 10 hyphens
+= 5 en dashes, and 9 hyphens = 3 em dashes).
+When a heterogeneous sequence must be used,
+the em dashes come first, followed by the en
+dashes, and as few en dashes as possible are
+used (so, 7 hyphens = 2 em dashes an 1 en
+dash).
+
+```````````````````````````````` example
+one-
+two--
+three---
+four----
+five-----
+six------
+seven-------
+eight--------
+nine---------
+thirteen-------------.
+.
+<p>one-
+two–
+three—
+four––
+five—–
+six——
+seven—––
+eight––––
+nine———
+thirteen———––.</p>
+````````````````````````````````
+
+Hyphens can be escaped:
+
+```````````````````````````````` example
+Escaped hyphens: \-- \-\-\-.
+.
+<p>Escaped hyphens: -- ---.</p>
+````````````````````````````````
+
+Three periods form an ellipsis:
+
+```````````````````````````````` example
+Ellipses...and...and....
+.
+<p>Ellipses…and…and….</p>
+````````````````````````````````
+
+Periods can be escaped if ellipsis-formation
+is not wanted:
+
+```````````````````````````````` example
+No ellipses\.\.\.
+.
+<p>No ellipses...</p>
+````````````````````````````````
--- a/deps/cmark/test/spec.txt
+++ b/deps/cmark/test/spec.txt
--- a/deps/cmark/test/spec_tests.py
+++ b/deps/cmark/test/spec_tests.py
@@ -0,0 +1,142 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import sys
+from difflib import unified_diff
+import argparse
+import re
+import json
+from cmark import CMark
+from normalize import normalize_html
+
+parser = argparse.ArgumentParser(description='Run cmark tests.')
+parser.add_argument('-p', '--program', dest='program', nargs='?', default=None,
+        help='program to test')
+parser.add_argument('-s', '--spec', dest='spec', nargs='?', default='spec.txt',
+        help='path to spec')
+parser.add_argument('-P', '--pattern', dest='pattern', nargs='?',
+        default=None, help='limit to sections matching regex pattern')
+parser.add_argument('--library-dir', dest='library_dir', nargs='?',
+        default=None, help='directory containing dynamic library')
+parser.add_argument('--no-normalize', dest='normalize',
+        action='store_const', const=False, default=True,
+        help='do not normalize HTML')
+parser.add_argument('-d', '--dump-tests', dest='dump_tests',
+        action='store_const', const=True, default=False,
+        help='dump tests in JSON format')
+parser.add_argument('--debug-normalization', dest='debug_normalization',
+        action='store_const', const=True,
+        default=False, help='filter stdin through normalizer for testing')
+parser.add_argument('-n', '--number', type=int, default=None,
+        help='only consider the test with the given number')
+args = parser.parse_args(sys.argv[1:])
+
+def out(str):
+    sys.stdout.buffer.write(str.encode('utf-8')) 
+
+def print_test_header(headertext, example_number, start_line, end_line):
+    out("Example %d (lines %d-%d) %s\n" % (example_number,start_line,end_line,headertext))
+
+def do_test(converter, test, normalize, result_counts):
+    [retcode, actual_html, err] = converter(test['markdown'])
+    if retcode == 0:
+        expected_html = test['html']
+        unicode_error = None
+        if normalize:
+            try:
+                passed = normalize_html(actual_html) == normalize_html(expected_html)
+            except UnicodeDecodeError as e:
+                unicode_error = e
+                passed = False
+        else:
+            passed = actual_html == expected_html
+        if passed:
+            result_counts['pass'] += 1
+        else:
+            print_test_header(test['section'], test['example'], test['start_line'], test['end_line'])
+            out(test['markdown'] + '\n')
+            if unicode_error:
+                out("Unicode error: " + str(unicode_error) + '\n')
+                out("Expected: " + repr(expected_html) + '\n')
+                out("Got:      " + repr(actual_html) + '\n')
+            else:
+                expected_html_lines = expected_html.splitlines(True)
+                actual_html_lines = actual_html.splitlines(True)
+                for diffline in unified_diff(expected_html_lines, actual_html_lines,
+                                "expected HTML", "actual HTML"):
+                    out(diffline)
+            out('\n')
+            result_counts['fail'] += 1
+    else:
+        print_test_header(test['section'], test['example'], test['start_line'], test['end_line'])
+        out("program returned error code %d\n" % retcode)
+        sys.stdout.buffer.write(err)
+        result_counts['error'] += 1
+
+def get_tests(specfile):
+    line_number = 0
+    start_line = 0
+    end_line = 0
+    example_number = 0
+    markdown_lines = []
+    html_lines = []
+    state = 0  # 0 regular text, 1 markdown example, 2 html output
+    headertext = ''
+    tests = []
+
+    header_re = re.compile('#+ ')
+
+    with open(specfile, 'r', encoding='utf-8', newline='\n') as specf:
+        for line in specf:
+            line_number = line_number + 1
+            l = line.strip()
+            if l == "`" * 32 + " example":
+                state = 1
+            elif l == "`" * 32:
+                state = 0
+                example_number = example_number + 1
+                end_line = line_number
+                tests.append({
+                    "markdown":''.join(markdown_lines).replace('→',"\t"),
+                    "html":''.join(html_lines).replace('→',"\t"),
+                    "example": example_number,
+                    "start_line": start_line,
+                    "end_line": end_line,
+                    "section": headertext})
+                start_line = 0
+                markdown_lines = []
+                html_lines = []
+            elif l == ".":
+                state = 2
+            elif state == 1:
+                if start_line == 0:
+                    start_line = line_number - 1
+                markdown_lines.append(line)
+            elif state == 2:
+                html_lines.append(line)
+            elif state == 0 and re.match(header_re, line):
+                headertext = header_re.sub('', line).strip()
+    return tests
+
+if __name__ == "__main__":
+    if args.debug_normalization:
+        out(normalize_html(sys.stdin.read()))
+        exit(0)
+
+    all_tests = get_tests(args.spec)
+    if args.pattern:
+        pattern_re = re.compile(args.pattern, re.IGNORECASE)
+    else:
+        pattern_re = re.compile('.')
+    tests = [ test for test in all_tests if re.search(pattern_re, test['section']) and (not args.number or test['example'] == args.number) ]
+    if args.dump_tests:
+        out(json.dumps(tests, ensure_ascii=False, indent=2))
+        exit(0)
+    else:
+        skipped = len(all_tests) - len(tests)
+        converter = CMark(prog=args.program, library_dir=args.library_dir).to_html
+        result_counts = {'pass': 0, 'fail': 0, 'error': 0, 'skip': skipped}
+        for test in tests:
+            do_test(converter, test, args.normalize, result_counts)
+        out("{pass} passed, {fail} failed, {error} errored, {skip} skipped\n".format(**result_counts))
+        exit(result_counts['fail'] + result_counts['error'])