cmark

Minor changes to the rendering engine

Updated the README

Changed the code to emmit code compatible with my CSS sheets

Speed up hierarchy check in tree manipulation API

Fix quadratic behavior when parsing inlines

Replace invalid characters in XML output

Fix issue with type-7 HTML blocks interrupting paragraphs.

Avoid quadratic output growth with reference links

docs: fix simple typo, chacters -> characters

Fixed installation on other than Ubuntu GNU/Linux distributions.

Increase timeout for pathological tests to avoid CI failure

Generate and install cmake-config file

Update the Racket wrapper with the safe -> unsafe flag change

Update .travis.yml

Increase timeout for pathological tests.

define whitespace per spec

Add MAX_INDENT for xml.

Treat textarea like script, style, pre (type 1 HTML block)...

Update spec.txt.

Fix handling of empty strings when creating XML/HTML output.

Re-add `--safe` command-line option as a no-op.

Adjust the MinGW paths for MinGW64.

Don't call memcpy with NULL as 1st parameter.

Add more NOLINTNEXTLINE directives.

Add NOLINTNEXTLINE to avoid a useless linter error.

Add needed include in blocks.c

Comment out unused function cmark_strbuf_cstr in buffer.h

CI: Run linter.

Add lint target to Makefile

Fix unnecessary variable assignment.

New lint target using clang-tidy.

CI: Another attempt to add clang-tidy.

Remove appveyor build.

CI: change shared to cmark_opts in matrix, add linter.

CI: test with different shared library configs.

CI: try setting CPP in matrix.

Fix syntax for matrix.

Linux CI: use both gcc and clang.

Add .gitattributes to ensure that line endings are normalized.

Windows CI: ensure UTF-8.

Revert "spec_tests.py: ignore line endings on diff."

spec_tests.py: ignore line endings on diff.

Revert "spec_tests.py: don't keep line endings (for windows CI)."

spec_tests.py: don't keep line endings (for windows CI).

Add CI badge.

Windows CI: try installing msvc tools.

Revert "Revert "CI: avoid using nmake.bat.""

Windows CI: specify cmd shell.

Revert "CI: avoid using nmake.bat."

CI: avoid using nmake.bat.

CI: install valgrind for linux.

Fix CI.

Diagnostic for CI.

Revert "Setup python."

Setup python.

Add GitHub actions CI.

Update to Unicode 13.0

Add uninstall target to Makefile.

Adding CIFuzz

Skip UTF-8 BOM if present at beginning of buffer.

Add casts for MSVC10.

Fix #220 (hash collisions for references).

Add pathological test for reference collisions (see #220).

Update date on cmark.1.

cmark.1 - Document --unsafe instead of --safe.

cmark.1: remove docs for `--normalize` which no longer exists.

Add cmark_get_default_mem_allocator().

Fix URL check in is_autolink

Fix null pointer deref in is_autolink

build: substitute the path into the generate files

build: use absolute path for cmarkTargets.cmake

Rearrange struct cmark_node

Improve packing of struct cmark_list

Use C string instead of chunk in renderer

Use C string instead of chunk for literal text

Use C string instead of chunk for custom block contents

Use C string instead of chunk for link URL and title

Use C string instead of chunk for code info and literal

Helper function to set C strings in nodes

Fix pathological_tests.py on Windows

Remove useless __name__ check in test scripts

Remove unused variable

Reintroduce version check for MSVC /TP flag

Fix CMake generator expression checking for MSVC

commonmark renderer: better escaping in smart mode.

Add options field to cmark_renderer.

Move C_VISIBILITY_PRESET back to src/CMakeLists.txt.

commonmark.c - use size_t instead of int.

Include string.h in cmark-fuzz.c.

fix -Wconst-qual warning

build: add exports targets for build tree usage

build: use target properties for include paths

build: chmod -x CMakeLists.txt (NFC)

build: reduce property computation in CMake

build: use `CMAKE_INCLUDE_CURRENT_DIRECTORY`

build: improve man page installation

build: only include GNUInstallDirs once

build: replace `add_compile_definitions` (#321)

build: cleanup CMake (#319)

Commonmark renderer: always use fences for code (#317).

Ensure that consecutive indented code blocks aren't merged...

Improve rendering of commonmark code spans with spaces.

normalize.py: use html.escape instead of cgi.escape.

Cleaner approach to max digits for numeric entities.

Fix entity parser (and api test) to respect length limit on numeric entities.

Code reformat

Don't allow link destinations with unbalanced unescaped parentheses.

Updaet spec.txt.

Create FUNDING.yml

print_usage(): Minor grammar fix, swap two words (#305)

Clarify which backtick strings (#301)

Remove "-rdynamic" flag for static builds (#300)

Merge pull request #294 from JerwuQu/cmark-issue-291-jerwuqu

Merge pull request #295 from nwellnhof/link-exe-with-lib

Revise changelog

Updaet changelog.

For afl target, don't build tests.

Update spec.

Bump version to 0.29.0.

Update spec.

Update spec.

Update spec.

Resolve link references before creating setext header.

commonmark renderer: improve escaping.

render: only emit actual newline when escape mode is LITERAL.

Update spec.

Update code span normalization...

Update spec.

Allow empty `<>` link destination in reference link.

Update spec

Remove leftover includes of memory.h.

Merge pull request #269 from foonathan/master

Fix #289.

Update spec.txt.

Update regression tests in light of recent spec changes...

Add regression test for #517.

Update test/spec.txt

Update spec; allow internal delimiter runs to match if...

Include references.h in parser.h

Update spec. Fix `[link](<foo\>)`.

Define CMARK_OPT_SAFE for API compatibility.

Make rendering safe by default.

Merge pull request #254 from github/empty-input

Avoid gcc warning for uninitialized variable.

Use hand-rolled scanner for thematic break.

Do cheaper test first.

Rename ends_with_blank_line with S_ prefix.

Add CMARK_NODE__LAST_LINE_CHECKED flag.

In ends_with_blank_line, call S_set_last_line_blank...

Add pathological test for #285.

Add regression test for commonmark#526.

Disallow unescaped ( in parenthesized link title.

Update test/spec.txt

Merge pull request #277 from elibarzilay/master

Merge pull request #278 from github/upstream-pledge

Merge pull request #275 from github/inline-sourcepos-off

render.c: reset last_breakable after cr.

Merge pull request #272 from felixonmars/patch-2

commonmark writer: use ~~~ fences if info string contains backtick.

Update scanners for new info string rules.

Update spec.txt.

Merge pull request #257 from pclouds/xml-space-preserve

Merge pull request #267 from nwellnhof/xml2md

Update spec.txt.

Check for whitespace before reference title.

Update spec.txt.

Merge pull request #265 from compnerd/upstream-windows-cross-compile

Fix appveyor links.

Simplify code normalization, in line with spec change.

Updated spec.txt.

Code span spec changes.

Update spec.txt.

entity_tests.py - omit noisy success output.

Merge pull request #259 from github/link-parsing-corner-cases

Fix spaces on regression test.

Removed meta from list of block tags.

pathological_tests.py: make tests run faster.

pathological_tests.py: added test for deeply nested lists.

Optimize S_find_first_nonspace.

Allow spaces in link destination delimited with pointy brackets.

Update spec.

Adjust max length of decimal/numeric entities.

Fix inline raw HTML parsing.

Don't allow list markers to be indented >= 4 spaces.

Update spec.txt from repository.

README - use commonmark/cmark for appveyor badge.

Merge pull request #250 from nwellnhof/test-static-only-build

Updated travis badges and issue tracker jgm -> commonmark.

Merge pull request #249 from nwellnhof/static-only-build

Fixed exit code for pathological tests.

Add allowed failures to pathological_tests.py.

Add timeout to pathological_tests.py.

Refactored pathological_tests.py

Merge pull request #230 from mity/more-pathological-tests

Merge pull request #217 from github/upstream/limit-nested-parens

Merge pull request #228 from github/upstream/inline-sourcepos

Updated changelog.

BUmp to 0.28.3

Merge pull request #240 from nwellnhof/add-cmake-include

Bump to 0.28.2, update changelog.

Fixed regression in install dest for static library.

Bump to 0.28.1, update changelog.

Don't use CMAKE_INSTALL_LIBDIR to create libcmark.pc.

Merge pull request #232 from github/upstream/refparse

--smart: open quote can never occur right after `]` or `)`.

Merge pull request #224 from pclouds/make-format-source-code

Merge pull request #222 from PavloKapyshin/master

Update spec.

Update changelog.

Merge branch 'master' of github.com:jgm/cmark

Update changelog.

Bump version to 0.28.0.

Update spec.

Merge pull request #219 from github/upstream/left-angle-link

Reverted mistaken fix to #193.

Merge pull request #213 from github/upstream/undefined-shift

Merge pull request #212 from github/upstream/memcpy-nonnull

Updated spec.txt.

DeMorgan simplification of some tests in emphasis parser.

Fixed undefined shift in commonmark writer.

latex writer: fix memory overflow.

Makefile: add libFuzzer to phony

Merge pull request #209 from philipturnbull/libFuzzer

Merge pull request #207 from github/upstream/oob-read

Merge pull request #205 from github/dont-scan-past-eol

Merge pull request #195 from github/revert-194-upstream/remove-normalize

Document cases where get_ functions return NULL.

Properly handle backslashes in link destinations.

Fixed bug with unescaped left angle bracket in link destination.

Fixed cmark_node_get_list_start to return 0 for bullet lists...

Use CMARK_NO_DELIM for bullet lists. Closes #201.

Merge pull request #198 from PavloKapyshin/master

README: Add link to cmark-scala (#196)

Remove normalize as an option per #190 (#194)

Fixed code for freeing delimiter stack.

Removed abort outside of conditional (typo).

Removed coercion in error message when aborting from buffer.

Print message to stderr when we abort due to memory demands.

libcmark.pc: use CMAKE_INSTALL_LIBDIR (#185)

Fixed buffer overflow error in S_parser_feed.

Update emphasis parsing for spec change.

Update spec.

Fixes for the LaTeX renderer (#182)

Merge pull request #179 from winterheart/cmake

Automatic code reformatting.

Added a test for NULL when freeing subj->last_delim.

Cleaned up setting of lower bounds for openers.

Fixes #178, quadratic parsing bug.

Added pathological test for #178 (quadratic).

Slight improvement of clarity of logic in emph matching.

Added regression test for #177.

Revert "More sourcepos! (#169)"

Revert "wrapper.py: remove unused (and faulty) libc instantiation (#175)"

Revert "Update cmark.3 man page."

Revert "Change types for source map offsets (#174)"

Fix "multiple of 3" determination in emph/strong parsing.

Update cmark.3 man page.

Change types for source map offsets (#174)

wrapper.py: remove unused (and faulty) libc instantiation (#175)

More sourcepos! (#169)

Correctly initialize chunk in S_process_line (#170)

Added 'make newbench'.

scanners.c generated with re2c 0.16 (68K smaller!).

scanners.re - fixed warnings; use * for fallback.

Makefile: use warnings with re2c.

Fixed some warnings in scanners.re.

Update CaseFolding to latest (#168)

Allow balanced nested parens in link destinations (#166)

Allocate enough bytes for backticks array.

Added pathological test for backticks.

Inlines: Ensure that the delimiter stack is freed in subject.

Fixed pathological cases with backtick code spans:

Remove redundant cmake_minimum_required.

Bump version to 0.27.1, update changelog.

Use VERSION_GREATER to clean up cmake version test.

Put set_policy under cmake version test.

Improve afl target. Use afl-clang by default. Set default for path.

Set policy for CMP0063 to avoid a warning.

Update changelog.

Bump to 0.27.0.

Update man 3 page.

Update spec.

Fix warnings building with MSVC on Windows (#165)

Fix CMAKE_C_VISIBILITY_PRESET for cmake versions greater than 1.8 (e.g. 3.6.2) (#162)

Fix for non-matching entities (#161)

Modified print_delimiters (commented out) so it compiles again.

Ran 'make format' to reformat code.

'make format': don't change order of includes.

Changed logic for null/eol checks.

Merge pull request #160 from kivikakk/kivikakk/fix-nul-lf-sequence

Merge pull request #158 from kivikakk/one-more-mem

Merge pull request #157 from kivikakk/list-parse-mem-leak

Merge pull request #156 from kivikakk/use-mem-to-free

Allow a shortcut link before a (.

Update spec.

Update spec.

Allow tabs after setext header line.

Regenerated cmark.3 man page.

Regenerated scanners.c.

Don't let URI schemes start with spaces.

Merge pull request #153 from gaborcsardi/patch-1

Fixed h2..h6 HTML blocks (jgm/CommonMark#430).

Update spec

Merge pull request #152 from nwellnhof/function-prototypes

Merge pull request #151 from petere/update-copying

Merge pull request #149 from PavloKapyshin/master

Update changelog.

Bump to 0.26.1.

Merge pull request #146 from t6/t6-patch-1

Removed size_t and ssize_t defs for WIN32.

Updated changelog.

Bump version to 0.26.0

Reformatted.

Updated spec.

Merge pull request #137 from foonathan/master

Fix sourcepos for blockquotes.

Merge pull request #143 from tst2005/patch-1

Replaced check for `\n` with `S_is_line_end_char`.

Empty list items cannot interrupt paragraphs (spec change).

Updated spec.

Fix mistaken sourcepos for atx headers.

Updated spec.

Modified api_test for change in "two blanks break out of a list."

Removed "two blanks breaks out of a list" feature.

Updated spec.

Don't allow ordered lists to interrupt paragraphs unless...

Makefile: have leakcheck stop on first error...

Merge pull request #140 from nwellnhof/set-cstr-fix

Fixed ATX headers and thematic breaks to allow tabs as well as spaces.

Updated spec.txt.

Reformatted.

Removed redundant check.

Updated spec and pathological tests for changes in strong/emph parsing.

Changed `process_emphasis` to get better results in corner cases.

Removed positon from delimiter struct.

Updated man page cmark.3.

Removed check for same mem allocator in S_can_contain.

Try to fix travis cmake.

Updated spec.

Travis: install more recent cmake on linux.

Ported robinst's changes to link parsing.

Revert "Better parsing of shortcut references."

Updated spec.txt.

Better parsing of shortcut references.

cmark_reference_lookup: Return NULL if reference is null string.

msvc: Fix warnings and errors

cmark: Remove old include

mem: Rename the new APIs

mem: Add a `realloc` pointer to the memory handler

Do not include `stdbool`

node: Memory diet

buffer: rever to using a 32-bit bufsize_t

node: Memory diet

cmark: Implement support for custom allocators

config: Add SSIZE_T compat for Win32

cmake: Global handler for OOM situations

test: Add tests for memory exhaustion

buffer: proper safety checks for unbounded memory

Merge pull request #135 from nwellnhof/fix-python-ctypes

Merge pull request #134 from nwellnhof/ctype-fixes

commonmark renderer: fixed code block as first in list item.

roundtrip tests: remove spurious failures.

Added new roundtrip_tests.py.

cmark.py: added to_commonmark (for round-trip tests).

spec_test.py - parameterize do_test with converter.

Updated spec.txt (no new or changed test cases).

spec_tests.py: exit code is sum of failures and errors.

renderer: no_linebreaks instead of no_wrap.

Fixed round trip tests.

Coerce realurllen to int.

Merge pull request #130 from MathieuDuponchelle/fix_unused_variable

Merge pull request #132 from BenedictC/master

Merge pull request #128 from kevinburke/copyright

Merge pull request #126 from nwellnhof/mingw-tests

Better documentation of memory-freeing responsibilities.

Clarify that it's the caller's responsibility to free the buffer...

Reformatted.

Fixed whitespace.

Use library functions to insert nodes in emphasis/link processing.

Correctly handle list marker followed only by spaces.

Updated spec test

Fixed a number of issues relating to line wrapping.

Merge pull request #111 from PavloKapyshin/master

Merge pull request #120 from nwellnhof/regression-tests

Merge pull request #118 from nwellnhof/win-eol-fix2

api_test: fix leak

Updated changelog.

Version bump to 0.25.2.

Handle buffer split across a CRLF line ending (closes #117).

Added api_test for #117.

Merge pull request #115 from nwellnhof/tab-fix

Merge pull request #116 from nwellnhof/win-eol-fix

Updated changelog.

Fixed typo in changelog.txt.

Updated changelog.

Updated date in man1 page.

Updated man 3 page.

Updated version to 0.25.0.

Updated spec.

Removed circle.yml.

Added circle.yml.

Regenerated cmark.3 man page.

Doc: clarify that cmark_node_free frees a node's children too.

Fixed documentation of --width in man page.

Merge pull request #109 from nwellnhof/msvc-c99

Merge pull request #108 from tinysun212/master

Merge pull request #107 from krytarowski/netbsd-support-1

Merge pull request #105 from chriseidhof/master

Merge pull request #104 from MathieuDuponchelle/more_refactoring

Removed unnecessary check for empty string_content.

Revert "Simplified condition for lazy line."

Factored out contains_inlines.

Simplified condition for lazy line.

Added code comments.

Added code comment.

Code cleanup: add function to test for space or tab.

Merge branch 'MathieuDuponchelle-refactor-S_processLine'

Fixed handling of tabs in lists.

Properly handle tabs with blockquotes and fenced blocks.

Clarify logic in S_advance_offset.

Updated spec.

Fixed typo in re2c version number.

Generated scanners.c with more recent re2c.

Require re2c >= 1.14.2.

S_advance_offset: Only set partially_consumed_tab in columns mode.

Updated spec.

Updated spec.

Updated spec test.

Simplified add_line (only need parser parameter).

Properly handle partially consumed tab.

Added partially_consumed_tab to parser.

Update test/spec.txt.

Fixed tabs in indentation.

Removed an unnecessary C99-ism in buffer.c.

Include starg.h when needed in config.h.

config.h: include stdio.h for _vscprintf etc.

Add include guards to config.h

Automatic code reformat.

Merge branch 'mbenelli-master'

config.h.in - added compatibility snprintf, vsnprintf for MSVC.

Bump to 0.24.1, updated changelog.

commonmark: is_autolink - handle case where link has no children.

iterator: handle NULL root.

Improved escaping in commonmark renderer.

render: initialize begin_content to true.

Commonmark renderer: use HTML comment to separate list from

Adjusted api test for last commonmark renderer change.

commonmark renderer: use 4-space indent for bullet lists.

Use 2 space + cr for line break in commonmark output.

Bump version to 0.24.0

Updated changelog.txt

Updated spec.txt

cmark_node_replace - unlink, but don't free, oldnode.

Fixed edge case with escaped parens in link destination.

Updated spec.txt.

Removed unnecessary call to cmark_node_unlink in cmark_node_replace.

Regenerated cmark.3 man page.

cmark.h: moved typedefs for iterator into iterator section.

Fixed make_man_page.py so it includes typedefs again.

Added cmark_node_replace(oldnode, newnode).

html renderer - removed unnecessary check for 'fenced'.

Updated format of test/smart_punct.txt.

Updated spec.txt and spec_tests.py to new format.

Fixed get_containing_block logic in src/commonmark.c.

Added an api_test that segfaults.

Fixed string representations of CUSTOM_BLOCK, CUSTOM_INLINE.

Use default opts in python sample wrapper.

Updated spec.txt.

Allow multiline setext header content, as per spec.

Updated spec.txt.

Don't allow spaces in link destinations, even with pointy brackets.

updated spec.txt

Updated `scheme` scanner according to spec change.

Updated spec.txt.

Allow NODE_ITEM inside NODE_CUSTOM_BLOCK.

Allow both block and inline nodes as children of CUSTOM_BLOCK.

Reorganized changelog.

Updated changelog.txt.

Defined aliases for old NODE_HTML and NODE_HTML_INLINE

Bump version to 0.23.

Don't allow space between link text and link label

Updated spec.txt and normalize.py.

Added xml renderer test to API tests.

Commonmark renderer: ensure that literal characters get escaped

api_test: updated commonmark renderer test.

render: added begin_content field.

Added to commonmark renderer test.

Added commonmark renderer test (currently failing).

Added simple latex renderer test to api tests.

Added a minimal man renderer test to api tests.

Reformat sources.

'make format' - reformat tests as well.

Replaced hard-coded 4 with TAB_STOP.

Rename NODE_HTML -> NODE_HTML_BLOCK, NODE_INLINE_HTML -> NODE_HTML_INLINE.

Another small optimization in _scan_at.

Small optimization in _scan_at.

Use input not parser->curline to determine last line length.

Removed out-of-date lua wrapper.

Make LIB_INSTALL_DIR configurable.

latex renderer - better handling of internal links.

Removed obsolete TIMER macro.

Added parens around expressions like 1 << 1.

Added version number to XML namespace.

Adjusted scanner for link url.

Regenerate scanners.c using re2c 0.15.3.

Separate parsing and rendering opts in cmark.h.

Rename hrule -> thematic_break.

CMARK_NODE_HEADER -> CMARK_NODE_HEADING.

Rename 'header' -> 'heading'.

xml renderer - Added xmlns attribute to document node.

Updated spec.txt.

Made roundtrip test part of the test suite run by cmake.

Commonmark renderer: ensure html blocks surrounded by blanks.

Added api_tests for custom nodes.

Fixed bug in html renderer for custom nodes.

Fixed S_can_contain to allow custom nodes to have children.

Updated API docs; getters return empty strings if not set.

Changed API for CUSTOM_BLOCK and CUSTOM_INLINE.

Rename RAW_BLOCK -> CUSTOM_BLOCK, RAW_INLINE -> CUSTOM_INLINE.

Removed bench.h and timing calls in main.c

Improved benchmark. Use longer test.

Added RAW_BLOCK and RAW_INLINE node types.

Removed need for sudo in make bench.

Removed outdated targets from Makefile.

Use fully qualified versions of constants.

Documented another strange hoedown behavior.

Merge pull request #95 from kainjow/master

Install static library.

Added why-cmark-and-not-x page.

Merge pull request #94 from kainjow/patch-1

Merge pull request #93 from Anderas/master

Merge pull request #89 from Eljay/fix-example

Merge pull request #86 from kainjow/patch-1

Fixed DOCTYPE in xml output.

Makefile.nmake: be more robust when cmake is missing.

Revert "Temporary change to see what is in the build directory."

Temporary change to see what is in the build directory.

Set convert_charrefs=False in normalize.py.

Merge pull request #84 from zmwangx/python3.5-compat-fix

Check for NULL pointer in _scan_at.

Updated spec

Version bump to 0.22.0.

Updated changelog.

Added iframe to list of blocktags, as per spec change.

Updated spec.txt.

Merge pull request #77 from kainjow/master

Don't include static in CMARK_INLINE.

Merge pull request #76 from kainjow/msvc-warnings

Merge pull request #75 from kainjow/master

Fixed bug with HRULE after blank line.

Check for empty string before trying to look at line ending.

Merge pull request #74 from nwellnhof/pdb_clash

Make sure every line fed to S_process_line ends with `\n`.

Improved cmark_strbuf_normalize_whitespace.

Treat line ending with EOF as ending with newline.

Removed pre from blocktags scanner.

Use -10 priority with renice.

Fixed `--hardbreaks` with CRLF line breaks.

Added api_test case for #68.

cmark.3 man page - regenerated with smart.

Merge pull request #70 from kainjow/master

Disallow list item starting with multiple blank lines.

Updated spec.

Use clang-format, llvm style, for formatting.

Updated spec.txt.

Updated test/spec.txt.

Updated test/spec.txt.

Removed cmark_strbuf_printf and cmark_strbuf_vprintf.

xml renderer - use sprintf to avoid strbuf_sprintf.

html renderer - use sprintf to avoid strbuf_printf.

Avoided another use of strbuf_printf.

commonmark renderer - use regular sprintf for list markers.

Avoid using strbuf_printf when not needed.

Improved smart_punct.txt.

Allow tabs before closing ##s in ATX header

Merge pull request #66 from omasanori/roundtrip

Set POSITION_INDEPENDENT_CODE ON for static library.

make bench: allow overriding BENCHFILE.

Improved clang-check target.

Improved `make_autolink`.

Use full CMARK_NODE_x instead of NODE_x alias.

Added clang-check target.

Updated changelog.

Removed unused ref link defn from README.

Updated benchmarks.

Removed algorithm.md; it has been added to spec as appendix.

Bump version to 0.21.0.

Updated spec.

Updated spec.

astyle reformatting.

Limit 'start' to 8 digits to avoid undefined behavior (overflows).

afl target: removed '-m none', added CMARK_OPTS.

Removed assertion that broke debug build.

Added `CMARK_OPT_SAFE` option and `--safe` command-line flag.

Updated spec.

Second try on ppa for travis.

travis - Added ppa that has recent cmake and is whitelisted.

Modified .travis.yml to use new containerized build system.

commonmark renderer - escape !.

Fixed scanner for link url.

commonmark writer - escape all #'s, not just at beginning of line.

Fixed soft breaks in commonmark writer.

Small cleanups.

Use cmark_render_code_point in renderers.

Fixed type on cmark_render_code_point.

Added cmark_render_code_point.

latex renderer - simplified using cmark_render_ascii.

man renderer: Use ASCII not LIT in outc.

cmark.3 - fixed botched escaping.

man renderer: Fixed character escaping, use new cmark_render_ascii.

render: added cmark_render_ascii, to be used in char escapers.

Limit generated man page to 72 character line width.

Added width parameter to render_man.

Removed options field from renderer struct.

Removed enumlevel field of renderer.

latex: Removed need to access render->in_tight_list_item.

cmark_render: ensure final newline.

commonmark renderer - don't need to manually adjust need_cr.

Fixed some windows warnings.

Restructured common renderer code.

Rename cmark_render_state -> cmark_renderer.

render: Simplified code, avoiding some allocations.

Factored out common bits of rendering into separate render module.

Allow non-initial hyphens in html tag names.

Updated spec.

Removed debug.h.

Removed dependence on debug.h.

Updated spec

Updated spec.txt.

Updated spec.txt.

Updates for new HTML block spec.

Implemented new treatment of dashes with `--smart`.

Updated test/smart_punct.txt.

Fixed small bug in test runner.

spec_tests.py: use stdout.out.buffer.write instead of print throughout.

Updated spec_tests.py from jgm/CommonMark.

Fixed compiler warnings.

Use sprintf rather than snprintf because we know it won't overflow.

Another try at getting snprintf part to compile on Windows.

Fixed typo in Windows fix.

LaTeX writer: fixed some compile errors on MSVC.

LaTeX writer: Fixed straight single, double quote.

LaTeX renderer: \textbf for strong.

astyle formatting improvements.

Added LaTeX renderer.

commonmark writer: correctly handle email autolinks.

Use utf-8 aware re2c.

`spec_test.py`: changed default to `spec.txt` (not `test/spec.txt`).

html renderer: properly split info on first space char.

Merge pull request #61 from txdv/version-functions-ffi

Added algorithm.md, informal explanation of emph/link parsing.

process_emphasis: Fixed setting lower bound to potential openers.

Rename potential_openers -> openers_bottom.

Rename start_delim -> stack_bottom.

Minor astyle reformatting.

Added case for #59 to pathological_test.py.

Fixed emphasis/link parsing bug.

Fixed off-by-one error in line splitting routine.

Added assertion to peek_char to catch any stray NULLs.

README: added 'make afl' instructions.

Small changes to README and benchmark files.

Renamed entities.h -> entities.inc.

Updated benchmarks.

Added `CMARK_OPT_VALIDATE_UTF8` option.

Small code simplification in inlines.c.

is_blank: recognize tab as a blank character.

skip_spaces: skip tabs too.

Don't rtrim in subject_from_buffer.

Removed utf8 validation tests.

Removed utf8 validation.

Updated spec for new tab behavior.

Renamed utf8proc_detab as utf8proc_check, removed detabbing function.

Preliminary changes for new tab handling.

spec_tests.py: allow → for tab in HTML examples.

normalize.py: don't collapse whitespace in pre contexts.

Simpler approach for entity lookup.

make_entities_h.py: confirm there are no hash collisions.

Revert "Rebuild src/entities.h when the generating python program changes."

Rebuild src/entities.h when the generating python program changes.

Mark entity data structures as const.

entities: Make the first entity in the array (TripleDot) work.

astyle formatting changes.

Added explanatory note about entities.h in Makefile.

Replace gperf-based entity lookup with binary tree lookup.

Fixed cases like

inlines.c: Added utility functions to skip spaces and line endings.

Updated spec.txt.

Fixed backslashes in link destinations that are not part of escapes.

Updated spec.txt.

Updated spec.txt.

Fixed entity lookup table.

Removed "add newline if line doesn't have one."

pathological_tests: removed timeout stuff.

Small logic fixes and a simplification in process_emphasis.

Updated benchmarks.md.

Fixed `process_emphasis` to handle new pathological cases.

Added another case to pathological tests.

Added timetouts to pathological tests.

More code simplification.

Code simplification.

Added more pathological tests.

Added pathological test case for jgm/commonmark#43.

process_inlines: remove closers from delim stack when possible.

Revert "Merge pull request #58 from nwellnhof/optimize_utf8proc_detab"

Updated spec.

Merge pull request #58 from nwellnhof/optimize_utf8proc_detab

Merge pull request #57 from nwellnhof/optimize_utf8proc_detab

Updated spec.

Updated changelog.

Merge pull request #56 from nwellnhof/bufsize_t

Added benchmark for hoedown.

Merge pull request #55 from nwellnhof/strbuf_error_handling

Updated changelog.

Merge pull request #54 from bryant/ensure-size-typo

Updated changelog.

Updated version to 0.20.0.

Rename `is_line_end_char` to `S_is_line_end_char`.

Factored out `S_find_first_nonspace` in `S_proces_line`.

astyle formatting changes.

Allow new list item container indented > 4 spaces.

Updated spec.

Updated spec.

Don't allow link labels with no non-whitespace characters.

Updated spec.

Updated spec.txt.

Added test for different line endings to api_test.

Cleaned up with is_line_end_char function.

Revert "Hacks to spec_tests.py to test CRLF support."

Revised "add newline to end if missing" for performance.

Merge branch 'master' of https://github.com/btrask/cmark into btrask-master

Updated spec.

Added note on 64-bit timings in benchmarks.md.

Fixed bug in list item parsing when items indented >= 4 spaces.

Updated spec.txt.

Fixed `is_autolink`.

Merge pull request #49 from jeroenooms/master

Merge pull request #45 from nwellnhof/windows_snprintf

Removed -pg from Debug build flags.

Revert "Ubsan: add -fsanitize=undefined to link flags too."

Ubsan: add -fsanitize=undefined to link flags too.

Added Ubsan build target, to check for undefined behavior.

Merge pull request #42 from jeroenooms/master

Merge pull request #38 from nwellnhof/chunks_for_links

Added link to CommonMarker, ruby wrapper, in docs.

cmark_consolidate_text_nodes: Avoid an unnecessary string allocation.

Merge pull request #37 from nwellnhof/numeric_entities

Updated benchmarks.

cmark_consolidate_text_nodes: avoid some unnecessary allocation.

README: add links to go, perl, haskell wrapper libraries.

Added note to wrapper.rkt, requires racket 5.3+.

Revised luajit wrapper.

wrapper.rb: added argument for options.

Removed wrapper3.py, made wrapper.py work with python 2/3.

Added build status badges to README.md.

Merge pull request #36 from elibarzilay/racket-wrapper

Better solution for memory leak in `cmark_consolidate_text_nodes`.

Straightforward fix for memory leak #32.

Improved 'make leakcheck'.

Updated spec.txt.

Updated changelog.txt.

Updated spec.

Bump version to 0.19.0.

Fixed _ emphasis parsing to conform to spec.

Updated spec.txt.

Merge pull request #29 from nwellnhof/static_library_flags

Appveyor - use 64-bit python.

Use os.path.join in test/cmark.py.

Makefile.nmake fixes.

Fixed spec location in Makefile.nmake.

appveyor - use nmake for build and test.

Appveyor - don't cd in test_script.

appveyor - removed pushd/popd

appveyor - add python to path

Revert "CMakeLists.txt: quote variable to avoid warning."

Really fixed cmake warning.

CMakeLists.txt: quote variable to avoid warning.

appveyor.yml - set up for using nmake.

Fixed typo in build script for appveyor

Fixed appveyor build_script

Fixed tabs in appveyor.yml

Tried build_script in appveyor.yml

Added build_script to appveyor

Trying configuration in appveyor.yml

Made environment a mapping in appveyor.yaml

Fixed yaml error in appveyor.yml

Added appveyor.yml

astyle reformatting.

Added to afl_dictionary.

make afl: use afl_dictionary, raise timeout for hangs.

Added test/afl_dictionary.

Merge pull request #28 from nwellnhof/accept_unicode_nonchars

Added "Standardized" bullet to README.

Improved README.

Check return status of utf8proc_iterate. Closes #27.

Added afl to .PHONY in Makefile.

Bump version to 0.18.3.

Updated changelog.

Merge pull request #25 from nwellnhof/include_patchlevel_in_soname

Merge pull request #24 from yangcha/master

Use strbuf_printf instead of snprintf.

Fixed memory issue.

Merge pull request #22 from yangcha/master

Modified copyright message with --version.

Updated man page.

Updated changelog.

Updated version to 0.18.2

Travis: only do roundtrip test on linux.

Fixed roundtrip.sh.

Merge branch 'master' of github.com:jgm/cmark

Updated cmark path in roundtrip.sh.

commonmark renderer - render links as autolinks when appropriate.

Scanners: added _scan_scheme, scan_scheme.

Added roundtrip_test Makefile target.

Fixed bug in cmark_strbuf_unescape (buffer.c).

Don't wrap inline HTML.

commonmark renderer - improved tight list detection.

commonmark writer - more tight list fixes.

commonmark - don't use indented code if first thing in list item.

commonmark renderer - don't escape in fenced code blocks.

commonmark - properly escape literal backticks.

commonmark renderer - honor softbreak if width == 0.

commonmark renderer - fixed memory access error.

commonmark - improved tight list rendering.

commonmark - use * for bullet list to avoid clash with setext header.

Fixes one-off error in previous commit.

commonmark - use space after delim if code is empty.

commonmark - use fenced code blocks if code starts/ends with blank.

commonmark renderer: special case EMPH(EMPH(x)).

commonmark renderer - better inline code.

commonmark renderer - fixed calculation of number of backticks in code.

Removed obsolete TODO comment.

commonmark renderer - two blank lines after list before code/list.

commonmark renderer - escaping improvements.

commonmark rendere: more fine-grained control over escaping.

commonmark renderer: better escaping of `#`.

More TODOs in commonmark renderer.

Added TODO comment to commonmark renderer.

Removed an unnecessary check.

Merge pull request #21 from nwellnhof/update_buggy_re2c_versions

Added shell script for round-trip spec test.

Fixed HTML blocks.

commonmark renderer: handle tight/loose list distinction.

commonmark renderer: proper numbering for ordered lists.

Merge pull request #20 from nwellnhof/re2c_warning

Added TODO on tight lists.

Made CommonMark renderer sensitive to CMARK_OPT_HARDBREAKS.

Initialize width.

CommonMark renderer: Added 'width' parameter.

CommonMark renderer: use right number of backticks in quoting code.

CommonMark writer: support title on link, image.

CommonMark renderer: improved code blocks.

CommonMark renderer: don't wrap headers.

Commonmark renderer: improved escaping.

Update README on output formats.

Updated output formats in man page.

Added commonmark renderer.

test/cmark.py: set default options for markdown_to_html.

leakcheck with all formats

Removed obsolete reference to CMARK_NODE_LINK_LABEL.

Added options parameter to cmark_markdown_to_html.

Merge pull request #13 from nwellnhof/documentation-fix

Check for CMAKE_C_COMPILER (not CC_COMPILER) when setting C flags.

Updated changelog.

Version to 0.18.1.

Use libcmark.a for static library on *nix.

cmark.h: Add argument to cmark_parser_new.

Build static version of library in default build.

astyle changes.

Updated version to 0.18.

Updated changelog.

Updated spec in test.

Merge pull request #10 from nwellnhof/afl

Cleaned up some comments.

Fixed use-after-free error.

Added 'Asan' build type.

Fixed use-after-free bug.

Repacked delimiter struct to avoid hole.

Updated benchmarks.

Packed cmark_node struct to fit into 128 bytes.

Moved a test outside a loop.

Added more --smart test cases.

Rename CMARK_OPT_SMARTPUNCT -> CMARK_OPT_SMART.

Made 'options' an int rather than a long.

Use ascii escapes for punctuation characters for portability.

Don't require python3 to make cmark.3 man page.

Move normalization step from main to cmark_parser_finish.

Properly handle " with no matching close quote with --smart.

Fixed man escaping for unicode characters.

Proper escaping of smart punctuation in man writer.

Removed unneeded include.

Removed outdated statement in man page.

Moved handling of --smart from renderer to parser.

Added options parameter to cmark_parse_document, cmark_parse_file.

Moved normalization step from XML renderer to main.

Fixed segfault with --normalize (closes #7).

Man page updates.

Implemented --smart for man output.

Further improvements and small optimizations in smart punctuation code.

Optimized --smart.

Added smart.c, smart.h with function abstracting smart punct rendering.

astyle changes (code formatting only).

Updated man page for --smart.

Improved smart quote recognition.

Added test for --smart option.

Added CMARK_OPT_SMARTPUNCT and --smart option.

Updated scanners for new whitespace definition.

Updated test/spec.txt.

Use try.commonmark.org for dingus address.

Merge pull request #5 from nwellnhof/zap_strnlen

Updated spec.txt from CommonMark repo.

Fixed autolink scanner to conform to the spec.

Update cmark.3 man page.

Added Profile build type, 'make prof' target.

Adjust benchmarks for commonmark.js.

Fixed license in README.md.

Fixed link to spec.

COPYING - added license for test/spec.txt and python code in test/.

Use 2-clause BSD license.

Removed JavaScript specific licenses in COPYING.

Improved README with note on wrappers.

Added reference to benchmarks in README.

Removed non-C related parts of README.md.

Makefile: Removed spec- and website- related parts.

Makefile: Added update-spec target.

Moved spec.txt to test/ directory.

Removed spec-specific files (DTD, spec generation tools).

Renamed changelog.c.txt -> changelog.txt.

Added note on documentation in man pages.

Updated changelog.c.txt.

Updated version in CMakeLists.txt.

Bump spec version to 0.17, updated changelog for spec.

Removed obsolete PACKAGEDIR in Makefile.

Removed JS specific stuff from .editorconfig.

Removed some old targets from .PHONY.

Makefile: Removed JS related stuff.

Makefile: replaced PROG with CMARK and PROG.

Merge pull request #297 from nwellnhof/makefile_fixes

Removed JS implementation, which is moving to its own repo:

Removed unused variable.

Tweaked uglifyjs options.

Added target for minified js/dist/commonmark.min.js.

Put generated commonmark.js in js/dist/ rather than js/.

Added spec with link break inside pointy braces (no link).

Set version in CMakeLists to 0.16.0.

Merge pull request #296 from nwellnhof/version_info

Fixed Xcode install example.

Avoid free(0) for link titles and URLs.

Removed unnecessary strbuf_free's.

Avoid trying to free string_contents for inlines.

Adjust lastLineLength before returning.

Renamed first_nonspace -> next_nonspace.

Added currentLine property to Parser.

Moved continuation checks & finalizers into blocks property.

Removed unnecessary check.

Revert "Reverted addition of offset property to DocParser."

Propagate lastLineBlank up through parents.

Removed unnecessary check.

Adjust lastLineLength before returning after close fence.

Merge pull request #293 from robinst/use-void-return

Merge pull request #292 from robinst/top-property-access

Merge pull request #291 from robinst/addChild-call-arguments

Merge pull request #290 from robinst/remove-unused-assignment-lastLineBlank

Merge pull request #289 from nwellnhof/user_data

Fixed escaping error in CDATA regex.

Man writer: ensure we properly escape multiline strings.

Bigger TODO comment in js/README.md.

html.js - explicitly specify second parameter of escapeXml.

Readjust parser->current after closing fenced block.

Changed rule for `_` emphasis and strong emphasis.

Renamed DocParser -> Parser.

Fixed typo.

Added new fine-grained JS benchmarks.

Reverted addition of offset property to DocParser.

Moved place where we increment line number.

Removed a check in addLine.

eslint: turned off no-constant-condition warning.

addLine: use this.offset; removed offset parameter.

Added offset property to DocParser.

Removed some unneeded tests (code clarity).

Small code clarification.

Put check for fence close with the other checks for end-of-block.

Moved check for closing fence to close-block-check section.

Minor code simplification.

Updated benchmarks.

Performance optimization - avoid repeating scan for nonspace.

Slightly adjusted performance optimization in new block starts.

Fixed linter errors.

Minor code cleanup.

Fixed #285 in cmark.

Added spec example: loose list with blank line after fenced code.

Defer closing of fenced code blocks.

Removed unnecessary setting of default tight=true in finalize.

Check for blank line before checking indent in Item.

bin/commonmark: don't use the now-removed toObject for --ast.

Nonrecursive rewrite of ends_with_blank_line.

Renamed parameters cmark_node -> node.

Rename bullet_char -> bulletChar.

README.md: refer to js/README.md for details on JS implementation.

Improved js README.md.

Provide getters and setters for public properties of Node.

Fixed template path in tools/makespec.py.

Fixed (interact) link in template.html.

Re-added tools/{makespec.py, template.html, template.tex}.

Rename sourcepos -> _sourcepos, added sourcepos() accessor.

Removed Node.toObject().

Renamed .t to ._type to mark it as "hidden."

Removed setType(), replaced getType() with type().

eslint fix.

eslint fixes.

normalize-reference.js: disable eslint spacing warnings.

Removed vestigial ReferenceDef node type.

Added getType(), setType() to node.js.

Removed artificial distinction btw FencedCode, IndentedCode

Updated benchmarks.

Added normalize-reference.js.

Bump package.json version to 0.16.0.

Added changelog.c.txt and js/changelog.js.txt.

Bump spec version to 0.16.

Updated changelog.

Slightly more efficient test for lazy continuation.

Fixed minor linter issues.

Added pathological tests to test.js.

Added (still commented-out) pathological tests to test.js.

Rewrote beginning of Entities section.

Inserted space (typo).

Removed defective Example 449.

Removed an implicit cast in node.js 'next'.

Updated benchmarks.

Initialize fields in objects to null rather than undefined.

blocks.js - avoid an implicit boolean cast.

Updated benchmarks.

Simplified some of the new-block-start logic.

Assign marker_offset to list data at initialization.

last_matched_container -> lastMatchedContainer.

Ensure that closeUnmatchedBlocks is not called unnecessarily.

Fixed start column for blockquotes.

Added lineNumber proper to DocParser.

Fixed typo in spec.

Updated benchmarks.

Made closeUnmatchedBlocks a method of DocParser...

Another try at .travis.yml.

More travis fixes.

Fixed typo in travis script.

Revert "Moved dingus to js directory."

Fixed travis (I hope).

Improved README.

Cleaned up Makefile.

Some astyle reformatting.

Moved spec2js.js to tools/.

Moved templates, makespec.py, and specfilter.hs -> tools/.

Moved mkcasefold.pl to tools/.

Moved eslint.json to js/ subdir.

Moved python, rb, lua wrappers to wrappers subdirectory.

Moved dingus to js directory.

Removed Makefile.old.

.travis.yml: don't run valgrind on osx.

Simplified release archive.

Updated benchmarks with new multilingual progit sample.

For benchmarks, read file rather than piping from stdin.

Fixed ANSI color 'reset' in js/test.js.

Use much more multilingual benchmark source.

Fixed .travis.yml for osx.

Fixed sh syntax in .travis.yml.

Wrap decodeURI in a try.

html.js, xml.js: avoid using concat.

Use new with Document and NodeWalker constructors.

js/bench.js: use capital letter for Showdown so eslint doesn't complain.

Improved eslint.json.

Updated travis script to install osx dependencies using brew.

Enable testing on osx with travis.

Use decodeURI instead of the deprecated 'unescape'.

Relaxed some eslint rules that don't make sense for us.

Fixed some linter errors.

Moved "use strict" to top of modules.

Moved common xml escaping code from xml.js, html.js to common.js.

Factored out normalizeURI into a single function in common.js.

Factored out unescapeString into new module, js/common.js.

A few changes to JS so its xml matches cmark's.

More xml.js improvements.

Small improvements to xml.js.

js: 'ListItem' -> 'Item'.

xml.c: Include level attribute in header.

Dingus: XMLRenderer -> XmlRenderer.

node.js: toAST -> toObject.

Fixed url -> destination in xml.js.

node.js: Added title and destination to list of props to show in toAST.

xml format: use 'destination' instead of 'url', like the spec.

Handle more attributes with xml.js.

Merge pull request #281 from nwellnhof/shrink_entity_table

Added intended two-line hard-breaks in Examples 521, 523.

XML writer - implemented list attributes.

XMLRenderer -> XmlRenderer.

Updated dingus to use XML for the AST pane.

Removed JS library's dependency on util.

xml.js: basic structure working.

Start on xml.js.

node.js: use switch instead of if-else.

Use order-only dependency of cmark.3 on PROG.

Makefile tweaks.

Non-verbose output on 'make testarchive', unless error occurs.

Removed circular build dependency, improved testarchive.

Makefile: require PROG for making man3 page.

Run the spec tests only if python3 is found.

'make test' - don't use verbose output.

Added cmark.3 man page to repository and archive.

Small reorganization of man/CMakeLists.txt.

Make sure library is built before man page.

'make leakcheck': check xml and man renderers as well.

Merge pull request #277 from nwellnhof/iterator

Added python markdown, parsedown to benchmark comparisons.

Updated benchmarks.

More efficient detab.

Timings: added 'preparing input'.

Added options to DocParser.

Removed a redundant clause from switch statement.

Added --time option to js/bin/commonmark.

Include a mini ansi implementation in test.js.

Removed ansi from js source tree.

Made 'options' a parameter of the renderer constructors...

Removed an unnecessary variable from cloneUnmatchedBlocks.

html.js: removed blocksep and innersep, no longer used.

HTML renderer: throw error on unknown tag type.

Put another regex in a variable.

Another "test before replace" optimization.

Small performance optimization in dealing with final newline.

Updated benchmarks.

HTML renderer: Test for characters that need escaping before substituting.

Updated benchmarks.

More HTML rendering efficiency improvements.

Don't use array buffer in HTML rendering.

Reordered some definitions, so regex is defined before being used.

Removed unused import.

Made Node a local variable.

Made NodeWalker a regular function returning an object.

Removed unnec. comment.

Regex optimizations in inlines.js.

Regex optimizations in blocks.js.

Moved more regexes into variables.

Put backtick regexes in variables.

Removed redundant condition.

Removed a redundant condition.

Use a switch instead of if-then-else.

Small optimization.

Updated benchmarks.

JS linter improvements.

Improved unescapeString performance.

Improved newline parsing efficiency.

Put limit on AST display as string.

Non-recursive version of toAST().

Updated dingus.

Updated renderAST in lib.

JS: Renamed 'c' property to 'literal' to match libcmark.

Added options to render, implemented sourcepos option.

JS: sourceloc -> sourcepos for confirmity with cmark.

Replaced 'pos' with 'sourceloc', making it an array.

Added end column position to pos for blocks.

JS: Improved end lines.

Don't include string_content in output of toAST().

JS: use c rather than string_content property for code blocks, html.

Updated benchmarks.

Rewrote HTML escaping for big performance gain.

Updated benchmarks.md.

Comment fix.

Simplified reMain for more performance gains.

Simplified reMain, with AST manipulation for 2-space hardbreak.

Don't initialize strings, string_content to defined value.

Made `tight` a property of `list_data`.

Initialize more properties of Node. Even better performance!

Initialize more fields in Node. Better performance.

Initialize more fields in Node.

Removed makeBlock. Use new Node directly. Initialize more fields.

JS: replace NUL characters with U+FFFD, as per spec.

More JS linter fixes.

JS linter improvements.

Rename js/lib/html-renderer.js -> js/lib/html.js.

JS linter cleanups.

Minor code reformatting.

Use linked list instead of arrays for AST.

js/bench.js: use npm, added markdown-it.

xml writer: add list attributes.

Added `cmark_iter_reset` and a note about handling destructive updates.

js: Use children instead of inline_content for Paragraph.

js: Use children rather than 'c' for Emph and Strong contents.

js: Changed 'label' in Link, Image to 'children'.

Fixed 'make alltest.md'.

cmark: Add function & option to normalize text nodes.

Merge pull request #273 from zmwangx/master

Merge pull request #269 from jgm/license

Clarify that brackets in general don't take precedence over emph.

Clarify that final newline is removed from paragraph content.

Talk of "info string" rather than "attributes" for code blocks.

Clarified precedence of code spans, HTML tags, autolinks.

Reformatted code consistently with astyle.

Added 'make astyle' to reformat code consistently.

Fixed a number of links in spec.

makespec.py: put pluralizing s inside link for better visuals.

makespec.py: put ids on headers, not separate anchor.

Removed pandoc from travis build deps.

Strip {-} off unnumbered sections in HTML version of spec.

Moved spec.html error checks into makespec.py.

Fixed spec.pdf target.

Fixed some duplicate references.

Removed second parameter of makespec.py.

Use cmark itself to build spec.html.

makespec.py: create toc without relying on pandoc.

Improved spec.pdf production.

spec2md.py -> makespec.py.

Linkify "info string" in spec.

Use shortcut reference links when possible in spec.txt.

xml writer - fixed issues with empty images.

Improved DTD readability with entities.

spec2md.py: Add reference link definitions for internal anchors.

Rewrote spec2md in python.

Fixed CDATA regex in js.

Moved non-comment examples (1e789d0f) to inline tag section.

Improved js regex for html comments.

scanners.re: More accurate regex for HTML comments.

spec: Added examples of non-comments.

spec: use HTML5 spec's rule for HTML comments.

Make clear that any sequence of characters is a valid document.

Spec: "is preferred" -> "takes precedence".

Minor reformatting.

Regularized "non-space character" spelling and added links.

Use 4 spaces rather than 5 to show "four spaces is too much."

Updated package.json.

Updated changelog.

Spec to 0.15.

Fixed some left-flanking/right-flanking reversals in spec for emphasis.

Updated dingus - incorporated changes made in CommonMark-site.

Updated spec changelog.

Omit footer on template.

Bump spec version to 0.14.

Added CC-BY-SA license to documentation.

Merge pull request #256 from nwellnhof/windows_fixes

Revert "Remove unneeded va_copy"

Revert "man: use a variable to store the char * from node_get_url."

man: use a variable to store the char * from node_get_url.

Require cmake >= 2.8.9, use a ppa for travis.

Attempted optimization of cmark_ctype.

Renamed a couple variables.

.travis.yml: install python3 for build.

Added cmark_ prefix to functions in cmark_ctype.

Fixed leakcheck for python3.

Use python3 for bench in Makefile.

Added cmark_ctype.h with locale-independent isspace, ispunct, etc.

spec: Define ascii punctuation character separately.

Added wrapper3.py (python3 sample wrapper).

Updated tests to use python3.

Fixed md2man for python3.

Avoid warning about unused options parameter in man writer.

Make `--sourcepos` affect xml writer too.

Added options parameter to renderers.

Added cmark_node_set_list_delim, cmark_node_get_list_delim.

Added CMARK_NO_DELIM to cmark_delim_type

Removed old 'ast' format, now that we have 'xml'.

Adjust benchmarks for cmark.

Improved end column/end line calculations in finalize.

Added end_column to cmark_node struct.

blocks.c - removed unneeded start_line parameter from make_block.

blocks.c: removed redundant line_number param in finalize.

Rename CMARK_NODE_LIST_ITEM -> CMARK_NODE_ITEM.

Revamped xml writer and dtd to use lowercase element names.

Rename cmark_node_type_string -> cmark_node_get_type_string.

Added xml writer, to dump the AST in XML.

Expose cmark_node_type_string in node.h (private header).

pathological_tests.py: limit to 65000 reps.

Spec: added example with two blank lines in fenced code in sublist.

Clarify precedence of empty list item over setext header line.

Makefile: Added spec target with check of internal links/anchors.

spec: Added rule for empty list item.

Spec: reworded principle of uniformity for more generality.

spec: added another example of unclarity in Markdown syntax description.

Added pathological tests for nulls in input and nested blockquotes.

Minor whitespace fixes.

Fixed shadowing error.

Updated benchmark for cmark.

Use python, not python2, as name of python interpreter.

Optimized the change in 72a556fd3bd17feaffba2fb6faa8c8f2a3931643.

Revert "Makefile: specify python2 in places where it's needed."

Fixed links to HTML5 spec.

Fixed some broken anchors in the spec.

Linkified "whitespace" in spec.

Makefile: specify python2 in places where it's needed.

Updated C parser to handle #252.

Added test case for link-in-link-in-image (#252).

JS: fixed parsing of link-in-link-in-image.

Fixed infinite loop in JS parser for link-in-link-in-image.

Improved rules for emphasis and strong emphasis.

Use python2 for man, test scripts.

Fixed make_man_page.py so it works with both python2 and python3.

Added definitions of whitespace and other character classes.

Made spec_tests.py more DRY.

Added HTML normalization tests to test/CMakeLists.txt.

Fixed normalization bug, added more doctests for normalization.

spec_tests.py: make --dump-tests sensitive to --pattern, --number.

Merge pull request #248 from cirosantilli/short-options

Merge pull request #247 from cirosantilli/run-single-test-by-number

Slight improvement to #250.

Merge pull request #250 from cirosantilli/deal-invalid-unicode

Added a few more doctests for HTML normalization (#245).

Fixed whitespace normalization (#245).

Merge pull request #245 from cirosantilli/normalize-whitespace-fail

Merge pull request #244 from petere/patch-1

Merge pull request #243 from maxlinc/patch-1

More Makefile adjustments.

Updated Makefile for changes in man page generation.

Create cmark.3 in cmake build process.

Merge pull request #240 from cirosantilli/ignore-pyc

scanners.re: use cmark_chunk rather than chunk.

Added 'literal' field to 'code' struct.

HTML writer: fixed bug in detection of fenced code blocks.

print: print fenced and fence_length for code blocks.

Re-added cmark_ prefix to strbuf and chunk.

Clarify that unicode whitespace counts as whitespace in emph rules.

Added comment.

Use UTF-8 sensitive functions to check for spaces around emph markers.

Rename get/set_string_content -> get/set_literal.

Use cmark_iter to avoid stack allocation in process_inlines.

man writer: use chunk literal in code blocks.

Added GNU profiler flags to debug build.

Use chunk for fenced code info, instead of strbuf.

Use as.literal instead of string_content for HTML and code blocks.

Added empty clauses for DOCUMENT node to html and man writers.

Rename CMARK_NODE_INLINE_CODE -> CMARK_NODE_CODE.

Clarified spec on indented code blocks.

Added API documentation to src/cmark.h. Closes #224.

Removed cmark_node_handler - it was used by cmark_walk, now deleted.

Makefile fixes for man page.

Makefile: add targets for html versions of man pages.

Use typewriter font for signatures in man pages.

Write API docs in cmark.h using markdown.

Added cmark_render_man (man page writer).