v4.0.5

• Fixed parsing problem for PRE element with CDATA in thread and single mode. #156
• Fixed the problem of parsing chunks when there was a script tag. #154
• Fixed parsing entity. In very rare cases there were wrong parsing. 541219b
• Fixed segfault if doctype hasn't attribute. #151
• Append link to Perl 5 wrapper module.
• Minor bug fixes

Special thanks to Kirill Zhumarin for PRs.

v4.0.4

• Updated CMakeLists.txt for cmake build. Added support for create Visual Studio Solution and the creation of packages for Linux systems. #116
• Fixed segfault if we have but not have a opening . #124
• Fixed cmake install path. #126
• Fixed rpm changelog date
• Delete Release 4.0.3
• Minor bug fixes

v4.0.2

• Grammar: change function name _pasition => _position
• Fixed infinite loop if html file is to big. Queue round not work properly - fixed. #117
• Append new function myhtml_node_is_void_element for check to see if we are dealing with a void element. #119
• Potential loss of the pointer on systems other than x86, x86_64 (Misaligned Integer Pointer)

v4.0.1

• Fix for creating a spinlock without support siplock #103
• Added two functions for detect encoding with returning found position myencoding_prescan_stream_to_determine_encoding_with_found and myencoding_extracting_character_encoding_from_charset_with_found #107
• Added automated package build and publicate on PackageCloud.io (https://packagecloud.io/modest/myhtml)
• Minor bug fixes

Special thanks for Alexander Fedyashov for help with automated package build.

v4.0.0

• API breaking changes
• MyHTML split to MyCORE, MyHTML, MyENCODING. MyCORE is a base module which include shared functions for all others modules. Each of the modules can build without other modules if he not dependent at it. It is good for those need only URL parse and not need other modules.
• Removed all io print functions to file: myhtml_tree_print_by_node, myhtml_tree_print_node_children, myhtml_tree_print_node; Use serializations instead of their
• If you use encoding enum, like MyHTML_ENCODING_UTF8, now it MyENCODING_UTF_8, i.e MyHTML_ENCODING_* => MyENCODING_*
• Functions migrated to MyCORE from MyHTML: myhtml_incoming_buffer_* => mycore_incoming_buffer_*, myhtml_string* => mycore_string*, myhtml_utils* => mycore_utils*
• Fully refactoring build system with GNU Make (Makefile), now it expects generally accepted parameters and rules, like install, clean, library and more
• Tested create a DLL library for Windows OS
• Support create ports for different OS or for simple change work with memory, io, threads (if build with threads, default)
• Support add self modules for build library
• Now all return statuses, like a myhtml_status_t, mycss_status_t changed to global mystatus_t (unsigned int)
• Added forgot '\0' if text node ends with '\r' #91
• Remove CMakeLists.txt
• Added PKG-CONFIG *.pc after make command

v3.0.1

New Release v3.0.1

Fixed broken mapping for convert encoding functions in release 3.0.0
Release 3.0.0 removed

• API breaking changes!!! See api_breaking_changes.md file
• Sync with Specification (https://html.spec.whatwg.org/multipage/)
• Fix problem with close token position in title tag (the inner essence)
• Fix problem with detect SHIFT_JIS encoding
• Added function myhtml_encoding_prescan_stream_to_determine_encoding to prescan a byte stream to determine its encoding. In other words, detect encoding in meta tag before start HTML parsing. See exapmle
• Added function myhtml_encoding_name_by_id for get encoding name by id
• Added function myhtml_encoding_extracting_character_encoding_from_charset
• Added utils/mhash.* for create a hash table
• Added function myhtml_node_tree for get current Tree from a node
• Сonsumes less memory when initializing, 3MB => 1MB with no negative impact on performance. In the future, the memory will be consumed even less.
• Now MyHTML_INSTALL_HEADER in cmake options set ON by default
• Fixed broken mapping for convert encoding functions after release 3.0.0

Thanks!

v2.0.1

• Fixed very serious problem with MyHTML::Collection in function myhtml_collection_check_size #84
• General improvements

Special thanks to Kostya that found big bug!

v2.0.0

• API Breaking Changes: Remove all functions associated with tag index: myhtml_tree_get_tag_index, myhtml_tag_index_*
• Changes for work with threads
• Removed example replacing_node_attributes_low_level.c. Example is not working correctly. Let the future
• Fix for myhtml_string_destroy function. Sometimes the resources are not free.
• Fix problem with serialization in UTF-8 (0xC2 0xA0)
• Added AVL-Tree for utils

v1.0.4

• Added possibility to set specify user data value for tree node (#67)
• Fixed gross errors in serialization (b27f9f7)
• Added myhtml_incoming_buffer_split function for split Incoming Buffer Node
• Added example for replace node attributes before begin token process replacing_node_attributes_low_level
• Changes for function name myhtml_incomming_* => myhtml_incoming_* (my bad english)
• Minor bug fixes

v1.0.3

• [https://github.com/lexborisov/myhtml/commit/873d0fa2cbe8ec4a3b8a50b649506e791f0907c7] Fixed attributes processing bug. Not cleared temp values after processing attribute key. What's in rare cases lead to segfault
• [https://github.com//issues/63] Fixes for tokenizer state problem with parse without build tree
• [https://github.com//issues/66#issuecomment-247941324] Fixed problem with append nodes to "tag index" in formatting reconstruction algorithm
• [https://github.com//issues/62] Fixed segfault if build without threads and used parse flag without process token
• Changed rules for parse flag MyHTML_TREE_PARSE_FLAGS_SKIP_WHITESPACE_TOKEN. Now we skip ws tokens, but not for RCDATA, RAWTEXT, CDATA and PLAINTEXT
• Added tree serialization by specification (see example)
• Tested by 1 billion HTML pages (by commoncrawl.org)