Install

Just copy replacement.js to Tampermonkey / Userscripts.
Change url to your config.

Idea

Here is small js script for word replacements.

• replacements stored at json file, which loaded by url
• support regular expressions
• flexible json design - create categories / subcategories
• flexible url matching - you can specify web_book chapters for specific replacements as example
• split json on multiple json files and use include, you can add configs from other ppl

Script Config

it's small config at the beginning of the scirpt

const config = {
"json_url": "https://api.npoint.io/adca32df1622919ca5bd",
"traditional_to_simple_chinese": true,
"default_priority_level": 1,
"binds": {
"click_interval": 3000,
"n": 3,
"m": 2,
"events": {
[known_events.config_update]: [4, 4, 6, 6],
[known_events.turn_on_off]: [6, 6, 4, 4],
}
}
};

json_url

url, with your config

Where you can store json?

• [https://www.npoint.io/]
• you can create gist on [https://github.com/], or create repo with configs and make links to raw file

traditional_to_simple_chinese

true, false
I readed chinese book with translation dubbing. I saw as original, as translated text.
So I added option to convert hard readable hieroglyphs to there simple version

default_priority_level

optional, useful if you're experimenting with the order of substitutions
read about __level in json

binds

const known_events = {
config_update: "forse_update_replacements",
turn_on_off: "swap_on_off",
}
// ....
"binds": {
"click_interval": 3000,
"n": 3,
"m": 2,
"events": {
[known_events.config_update]: [4, 4, 6, 6],
[known_events.turn_on_off]: [6, 6, 4, 4],
}
}

Binds - settings - where and how click to run some options.
I want to work them as on PC, as on IOS - and don't want to add GUI buttons.

Idea is following:

• Screen splitted on n x m rectangles
• each rectangle get it's own number 1, ... n x m
• from left-> right, from up->down
  

Example for n=3, m=2

   -------------
| 1 | 2 | 3 |
|---|---|---|
| 4 | 5 | 6 |
-------------

Next step is to set option - what to do and order - which block clicks expected.

click_interval

It's interval, where you have to do all clicks. "click_interval": 3000 - 3 seconds

indexes

[4, 4, 6, 6]
Expected order of clicks.
At this case - you should tap twice in left bottom screen
And after that - twice in right bottom.

known options

• config_update - script would load json config again and update
  replacements (If you change config and don't want to reload page)
• turn_on_off
  off - script discard changes and stop works on text updates,
  on - works again, usefull if you want to check original values before script.

Json Config

config looks this way:

{
"__urls": [
"https://some_novel_site/my_novel/chapter_*"
],
"smthA": {
"a": "b",
"smthB": {
"smthC": {
"x": "y",
}
}
},
"smthD": {
"/val(.+)/": "test $1"
}
}

Check here mine current config:
[https://www.npoint.io/docs/adca32df1622919ca5bd]
[https://www.npoint.io/docs/97ef86b1e99dfc26c72d]

You can create so much nested nodes with any names, as you want.
Script recursively parse json and collect pairs

arr = [
["a", "b"],
["x", "y"],
["/val(.+)/", "test $1"],
]

Replacements

Replacement have 2 options.

• basic form: string: string
• random form: string: [str1, str2, ...str_n]

Only this things would be collected.

Random replacements

"x": ["a", "b", "c"]
"老三", ["Third Bro", "Younger Bro"],
Random value would be select for replacement.
"AAx" -> "AAb"
"AAx" -> "AAc"

Regex support

Example:
"/(\\d)o/": "$10" - change o after digit to 0
2o -> 20, 9o -> 90

• start and end with /
• escape with \\
• all flags supported i, u, e.t.c.
• flag global g auto inserted
• multiline would work only if all text lays in one html element - paragraph as example, depends on page

google javascript regex if want more examples

Space problem

• you have string ab
• you created replacements "a": "Alice", "b": "Bob" basicaly r###lt would be AliceBob
  but script converts it to Alice Bob
  usefull for hieroglyphs without spaces
  

rules to convertion:

• replacement like Alice - start from letter, not hieroglyph
• it search - where to insert Alice "XXX<--aXXX"
• if there is letter too - add space

Key Words

There are following key words:

• __urls - ignore node if not passed
• __include - add other nodes from higher level, or url to load another json
• __level - set priority level, probably some replacements should work first, another later

__urls

example:

{
"common_replacements": {
...
"__urls": ["*"],
}
"MyNovel1": {
...
"arc 5": {
"specific per concrete chapters": {
"x": "y",
// chapters 1159-1212
"__urls": [
"1159",
"/11[6-9][0-9]/",
"/120[0-9]/",
"/121[0-2]/",
],
}
}
"__urls": [
"https://some_novel_site/my_novel/chapter_*",
"https://another_source_with_novel/chapt*"
]
}
}

how script work with "__url", nested behaviour

Script recursively look in each node {}.
If it see __urls and they passed - it started to collect replacements in that node.

You can add __urls in any subnode too.
If they not passed - that subnode would be ignored.
Usefull for specific chapters of book/ pages, check example.

url format

easy way *

use *, it solves 99% cases
https://tw.wa01.com/novel/pagea/lunhuileyuan-nayizhiwenzi_*
if you really have * in your url (some extreme cases) - then escape it \\*
https://tw.wa01.com/novel/\\*/lunhuileyuan-nayizhiwenzi_*

hard way full regex

• easy way - use unique part from url, no mess with /
/lunhuileyuan-nayizhiwenzi_.*/
• full way - escape []^&$.()?/\+{}|* with \\ https:\\/\\/tw\\.wa01\\.com\\/novel\\/pagea\\/lunhuileyuan-nayizhiwenzi_
• add bounds /url/ - to point that this is regular expression
  /https:\\/\\/tw\\.wa01\\.com\\/novel\\/pagea\\/lunhuileyuan-nayizhiwenzi_/
• add your regex part
  /https:\\/\\/tw\\.wa01\\.com\\/novel\\/pagea\\/lunhuileyuan-nayizhiwenzi_.+/
• example for specific chapters
  [1][0-9] - chapters 10 - 19
  /https:\\/\\/tw\\.wa01\\.com\\/novel\\/pagea\\/lunhuileyuan-nayizhiwenzi_[1][0-9]/

__include

example:

{
"commonA": {...},
"commonB": {...},
"Book1": {
...
"__include": {
"smth_name": "commonA",
"blockB": "commonB",
"engeneer terms": "https://some_server/engeneer_terms.json"
}
},
"news_portal": {
...
"__include": {"commonA": "commonA"}
}
}

As you see - it allows to include outer nodes or even load other jsons.

how it works

Firstly script recursively check all nodes
and collect all names on all depth like "name": {...}
Later script just insert in node with __include all data by aliases

Example:
json by url1

{"x": "y", "z": {"a": "b"}} // url json

{
"commonA": {"nodeA": {"n": "m"}},
"ExampleNode": {
...
"__include": {
"alias1": "nodeA",
"engeneer_terms": "url1",
}
}
}

ExampleNode converted to

{
...
"ExampleNode": {
...
"alias1": {"n": "m"},
"engeneer_terms": {"x": "y", "z": {"a": "b"}},
"__include": {...},
}
}

__level & order

about order of replacements

As earlier was mentioned - script build simple array from json

arr = [
["a1", "b1"],
["a2", "b2"],
["a3", "b3"],
]

and apply replacements in same order

automatic order

Example:

text = "xxAAxx"
json = {"A": "B", "AA": "CC"}
replacements = [
["A", "B"],
["AA", "CC"],
]

we expect xxCCxx
basically r###lt would be xxBBxx

But this problem is solved.

1. if adding key contains one of the earlier added - it would be moved before that point.
   json [ ..., -> ["A", "B"], ..., ]
2. if exactly same key exists, new won't be added

Problem But regular expressions can't be auto checked.

{
"100": "hundred",
"/o\d/": "0$1",
}

we expected "1o0" -> "hundred"
but if replacements collected in wrong order (which json not guaranteed)

replacements = [
["100", "hundred"],
["/o\d/", "o$1"],
]

here __level comes

{
"100": "hundred",
fixes: {
"/o\d/": "0$1",
"__level": 0,
}
}

__level

Example:

{
"first_order_fixes": {
...,
"__level": 0,
},
...
"source_which_override_mine_replacements": {
"__include": {
"fiction terms": "some_url",
}
"__level": 2,
}
}

__level allow you to solve problems with order
Default level set in javascript config:

let config = {
...
"default_priority_level":  1,
}

All replacements collected independently by this levels.

let arr = [
"-10": [['a1', 'a2'], ['a3', 'a4'], ...]
"0": [['b1', 'b2'], ['b3', 'b4'], ...]
"1": [['c1', 'c2'], ['c3', 'c4'], ...]
"5": [['d', 'd2'], ['d3', 'd4'], ...]
]

Autoorder, checking same key - works only on the same lvl
Finally they squashed in one array, ordered by levels

[
'a1','a2',
...,
'b1','b2',
...,
'd1','d2',
...
]

Inheriting Level is inherited by child nodes.
Here is example.

{
"a1": "a2", // default level 1
"a": {
"b1": "b2" // level 1
"b": {
"c1": "c2", // level -100
"c": {
"d1": "d2", // level -100
"d": {
"e1": "e2", // level 5
"f": {
"f1": f2, // level 5
}
"__level": 5
}
}
"__level": -100
}
},
"x": {
"x1": "x2", // level 3
"__level": 3
}
}