TODO.otl

> Goals

TODO
	[_] FIXME: angle brackets break format for xhtml and fb2
	[_] FIXME: save indendent scroll position in local storage
	[_] FIXME: dld/tl bugs found on re:zero
	[_] HACK deal with connection interruptions (don't freeze as it's now)
	[_] Simplify dicts format as one file
	[$] Split todo.otl in multiple docs/*.nou.
  2016-07-02 [X] Move 'js' parts into aeternum
	Split development in different orphan branches
	[_] FIX: when specifed direct --url to chapter w/o number -- need process chapter anyway
		:= Currently it's not -- because loop won't run with chapters=[]
		: http://stackoverflow.com/questions/15301147/python-argparse-default-value-or-specified-value
	[_] Autogenerated gh-pages
		http://stackoverflow.com/questions/5296236/how-to-add-a-git-repo-as-a-submodule-of-itself-or-how-to-generate-github-page
		Add links to all involved sub-repos
	[_] Releases
		https://help.github.com/articles/creating-releases/
    Allow world-wide editable wiki for users to place their own lexics/sub-plugins/TL-stories


TRY:(syntax)
	Make with python -- converge statistics from separate words, then crop by
	threshold >3, filter frequent particles '-to', filter list of already
	highlighted words (contained several yaml's -- specific and general),
	show.  Then using this statistics on new files you can speedily determine
	necessary new words to add. Then you can generate .vimrc with necessary
	syntax.

	Second merit -- using statistics, you can show translation only under
	first 3 word entries, ignoring all others -- for text space economy when
	prominent knowing language.

		$ cat arifureta-130.otl | sed -rn '/\|\s*(\w+)\s*\|.*/s//\1/p' | sort -n | uniq -c | sort -n


Static formats generators
	xhtml
		[_] THINK? use xml2dict parser for forming page?
			: https://github.com/martinblech/xmltodict

		TODO: add splitting by regex, generate content by js?
			Regex
				http://www.w3schools.com/jsref/jsref_obj_regexp.asp
				(i) https://developer.mozilla.org/en/docs/Web/JavaScript/Guide/Regular_Expressions
			File reading
				http://www.html5rocks.com/en/tutorials/file/dndfiles/
				http://stackoverflow.com/questions/3582671/how-to-open-a-local-disk-file-with-javascript
				https://gist.github.com/thiagodebastos/08ea551b97892d585f17
				NOTE: loaded db must be pre-generated from all translators, dicts, etc purposely for this WN
					? how to unite db's for subtitution / highlighting
				THINK:(-vs-) db format
					sqlite
						better access time
						compressed size
						query
						! CHECK: sqlite in js
					plain text
						readability
						editability
						distribution
						vcs
			Format
				BUT: how to load yaml? Or js can natively (but js isn't Ruby)?
					: http://stackoverflow.com/questions/4335804/javascript-yaml-parser
					https://github.com/tj/js-yaml
					https://github.com/jeremyfa/yaml.js
					https://code.google.com/archive/p/javascript-yaml-parser
				ALT:(own format) simplified yaml -- use code snips from yaml parser
					keys: ^\s*([a-z_]+\.?)+:\s*$
						MAYBE use '.' for flattened nesting
							object-like k1.k2.k3:
							path-like k1/k2/k3:
							THEN all next stuff until another key goes into last key
								+ no need to awfully control indent in 920 editor on android
								+ useful for long lists to find out hierarchy coords
								+ easy split/join throughout multiple files
								- miswritten path parts -> disjointed hierarchy
					single-line statement: \s*
						origin, sound, tr1, tr2, ...
					indent parsed only for keys
					no mixed keys/statements on one level
						all statements go to last key independently from indent
					[_] Document dict/style formats in README

		[_] Table folding / hiding
		[_] DEV Copy to clipboard on press (individual words) in table column
			http://stackoverflow.com/questions/3325867/folding-unfolding-table-with-html-javascript
		>>[_] http://www.html5rocks.com/en/mobile/responsivedesign/
		[_] DEV popup meaning/transcription on click
			http://www.sitepoint.com/xhtml-strict-popups/
			http://accessify.com/features/tutorials/the-perfect-popup/archive/
		[_] How about clicking on syntax words -> links to some chapter of jap grammar book?
			< likewise some figure of speech which I forget often
		THINK
			DEV integrate yarxi.pl
				[_] Copy articles inside xhtml (one article -- many references)
				[_] toggle visibility of section with article from yarxi by click on word
			[_] reuse same article (content is one copy) for all same words (construct dictionary with back-refs)
				Store db reduced for particular novel instead of universal dicts
			[_] jump to glossary at bottom of page / another embedded page / popup window with content


NOTE:DEV:(js)
	2016-02-04 [X] FIXME keep scroll position on resize and browser restart
	[_] THINK: onresize and onbeforeunload -- enough to storePos()? But if crash...
	[_] FIXME: If text wraps, then on resize relative position isn't correct
	[_] BUG: call for storePos(); inside onResize() don't work as expected
		:= It somehow saves new position after resize instead of old!
		:? CHECK: Maybe window 'h' is updated before 'y'?
	[_] BUG:WARNING: key scrollRelPos in localStorage is shared between all arifureta-*.xhtml
		: Therefore when loading new page it will jump to new old inherited position!


THINK
	2016-02-03 [X] TODO: continue line-by-line translation from last response onward
	[_] FIXME: don't split line in source/web.py, if before/after regex <=6 symbols
		: Also unite all splitting in one place -- move additional splitting at ','
	[_] FIXME: split line -- when url for GET must be less then 2000 characters
	[_] Make statistics on words
	: choose set by threshold and highlight -- or hi! by gradations of like
	: pink or better last gray palette. Others same as normal text/kana.
	:= This will allow to focus on most often used kanji to remember
	Export vim syntax highlighting to phone
		hardcopy > ss.ps (no japanese) and then ps2pdf ss.ps
		:OR :TOhtml and then make epub?
		:> Now I work with .xhtml creation

	[_] Reformat synonyms table to easily copy entries into yml as whole row
		:= can be also kept hidden and being copied in yml format only on click

	IDEA: different processing, based on absolute paths?
		E.g. -- treat group 'Characters' as whole words only
		:> No sense: as I could use patterns in fields. Firstly, accumulate some lexics.

	IDEA: in xhtml make words in table as links
		en-ru -- translate in online dictionary
		jap-kanji -- search in yarxi/stardict

	IDEA: choose word group directly from radio list after click on word
		:= It allows to fill db in much more faster/comfortable way

	IDEA: using Qt I can edit TL's textboxes with fakevim


REGRESSION
	[_] FIX generation from yml vim syntax highlightings
	[_] THINK if possible to fix auto indentation for yml
	[_] Concealing in yml of '|' and replacing with bold ', '
	[_] FIX: remove jap '[' ']' from TR table


DEV plugin for vim: specific highlighting and editing optimization
	:= To eliminate necessity in votl and logically use some own keymaps.
	:+ Then I could use python inside vim to regenerate and reload syntax files
	> IDEA (cornerstones):
		Lexems -- fast extraction into *.yml
		Grammar -- group and split difficult sentences
	IDEA: edits in vim -> export back into db to benefit all other formats/viewers

	Consider as the best idea, especially for ranobe translation process.
	[_] Shortcuts to fast populate *.yml with new words
		:? open split with local .yml, fast open ',.[0-9]' common .yml
	[_] Regex convert
		SEE:(vim) :h perl-patterns
		https://github.com/othree/eregex.vim/blob/master/plugin/eregex.vim
		THINK: prefer perl (or what compatible with js) instead of vimre?
	[_] THINK?: Attach account from yarp for grammar parsing


DEV
	[_] USE cfg/*.yml for common lexics and <ranobe>.yml or <ranobe>/*.yml for specific
	[_] Generate vimrc from all yml cfgs into ~/.cache/forestanza.vim
		: OR load by :pythondo, assuming already installed pyYaml
	[_] Load syntax by 'source forestanza.vim' inside local <ranobe>/.vimrc
		: OR inside <ranobe>/ place script to launch: vim -c 'source ...'

Style
	256 term -> RGB
		http://www.calmar.ws/vim/256-xterm-24bit-rgb-color-chart.html
	[_] Split lexics and colorscheme.
		[_] Treat whole key-path inside yml as identifier for group
		[_] Inherit colors in colorschemes from parent level
		[_] Override colors by direct path
		[_] Set several color variations for term and gui/pdf