Virastar Versions Save

• Added: removes spaces between dots on fix_three_dots
• Added: new option normalize_dates to re-order date parts with slash as delimiter
• Added: new option fix_misc_spacing to remove space before braces containing numbers
• Added: new option remove_diacritics to remove all diacritic characters #4
• Added: removes markdown link spaces inside normal parentheses
• Changed: early fix persian glyphs
• Changed: removing space between different/same marks
• Fixed: lazy seek before dashes on frontmatter preserving
• Fixed: combined preserving markdown links
• Fixed: avoid new-lines as whitespaces
• Fixed: combined pattern for removing kashidas
• Fixed: prevent removing spaces after punctuations on the end of lines
• Fixed: prevent removing double spaces on the end of lines
• Fixed: prevent removing new-lines on normalize braces

• Added: cleanup zwnj before/after ellipsis
• Added: converting kashida between numbers to ndash
• Added: new option for converting arabic hamzeh
• Added: replaces spaces after ellipsis
• Added: support for markdown images
• Added: support for more punctuations before/after zwnjs cleanup
• Changed: also removes nbsp from beginning of new-lines
• Changed: late checks for zwnjs
• Changed: revert to the old and big uri pattern, ref
• Fixed: account for punctuations after domain tlds
• Fixed: account for space before image opening brace within links
• Fixed: also single padding on the beginning of the text
• Fixed: check for non-space after space and prefixes
• Fixed: check for persian chars before suffix spacings
• Fixed: optional space after preservers

• Added: cleaning more than one of diacritic chars on fix_diacritics, props @languagetool-org
• Added: extra method for converting persian numbers back
• Added: fix another arabic kaf char on fix_misc_non_persian_chars
• Added: removes space before common domain tlds on fix_spacing_for_punctuations
• Added: replace comma between numbers to thousands separators on fix_numeral_symbols
• Added: support for man tan shan suffixes
• Changed: begin/end space cleanup after preservers
• Changed: extract diacritics fixes as new option: fix_diacritics
• Changed: yet another pattern for preserving URIs, (ref)
• Fixed: fix ha haye before other suffixes
• Fixed: support for more punctuation types after suffixes

• Added: (undocumented) fix heh + ye, alternative to fix_hamzeh
• Added: cleaning whitespace/zwnj between new-lines on cleanup_spacing
• Added: new option fix_numeral_symbols to replace percent signs and decimal separators, props @ebraminio/persiantools
• Added: new option fix_persian_glyphs to replace glyph chars, props @ebraminio/persiantools
• Added: new option fix_suffix_misc to fix hamza with double yeh, props @ebraminio/persiantools
• Added: new option markdown_normalize_braces
• Added: new option markdown_normalize_lists
• Added: new option preserve_frontmatter to preserve frontmatter data
• Added: padding the end of string while cleanup
• Added: re-ordering extra marks: ?! into !?
• Added: removing space between same marks
• Added: removing unnecessary zwnj on start/end of each line, props @ebraminio/persiantools
• Added: replacing more than one english question mark with just one
• Added: skip cleanup if text is empty or whitespace
• Added: storing markdown links separetly to help space cleanup working
• Changed: deprecated aggresive option
• Changed: moved cleaning whitespaces before newlines to cleanup_begin_and_end
• Changed: new option fix_spacing_for_punctuations extracted from fix_spacing_for_braces_and_quotes
• Changed: simpler pattern for preserving URIs
• Changed: some options disabled by default: preserve_braces, preserve_brackets, skip_markdown_ordered_lists_numbers_conversion
• Fixed: account for punctuations, braces and quots after suffixes
• Fixed: account for zwnj after yeh on fixing hamzeh
• Fixed: copy options object before parsing
• Fixed: putting back correct whitespace on cleaning zwnjs
• Fixed: unescaped char on space after dots in numbers

• Added: new option normalize_ellipsis to replace more than one ellipsis with one
• Added: convert all soft hyphens into zwnj, on cleanup_zwnj
• Added: remove direction marks from begin and end of text, on cleanup_begin_and_end
• Added: extra method for flipping punctuations
• Fixed: fix three dots as standalone method

• Added: initial check for type of the input
• Added: new option cleanup_line_breaks to remove more than two contiguous line breaks
• Added: new option preserve_entities to preserve html non decoded entities
• Added: new option preserve_comments to preserve html comments
• Added: new option preserve_nbsps to preserve no-break spaces
• Fixed: also fix heh plus standalone hamza
• Fixed: clean spaces before diacritic characters
• Fixed: clean ZWNJs before diacritic characters
• Fixed: putting back chars after suffix white spaces
• Fixed: better pattern for preserving html tags
• Fixed: decode all decimal, hex and some selected entities
• Fixed: decode html entity as standalone method

• Added: support more misc non Persian chars, props @ebraminio/persiantools
• Added: new option: fix_punctuations
• Fixed: better prototype structure
• Fixed: cleanup zwnj as standalone method

• Added: should not replace sprintf directives
• Added: extra method for swaping incorrect quotes
• Fixed: word tokenizer accounts for wraping chars
• Fixed: no space after dots in numbers
• Fixed: no space in time parts in English
• Fixed: mismatch options for Arabic/English numbers

• Added: support prefix: bi*
• Added: support suffix: *am, *at, *ash, *ei, *eid, *eem, *and
• Added: support suffix: *hayee, *hayam, *hayat, *hayash, *hayetan, *hayeman, *hayeshan
• Fixed: check for space befor suffix: *tar, *tari, *tarin

props @zoghal

• Added: convert back quot/apos entities
• Added: new option: decode_htmlentities
• Fixed: test suite