✨Feature/plugin non breaking spaces #55
Conversation
Relates #45
| const replaced = line.value.replace(pattern, replacement); | ||
|
|
||
| line.setValue(replaced); |
There was a problem hiding this comment.
Better to use this syntax
| const replaced = line.value.replace(pattern, replacement); | |
| line.setValue(replaced); | |
| line.setValue(value => value.replace(pattern, replacement)); |
| } else { | ||
| options.logger.warn( | ||
| `Pattern for current language ${language} was not found` | ||
| ); |
There was a problem hiding this comment.
- Are you sure you want to log warning for each line in language that doesn't have pattern to replace?
- What if I want use same patterns for all supported languages, am I forced to define same patterns for each language separately?
There was a problem hiding this comment.
- Good point, no. Moved it to
transformFullOutput - Using same pattern for all languages could not be wise. Languages have different rules. Sure it makes sense in english uk and us, but thats about it. So I don't know, what do you suggest?
There was a problem hiding this comment.
- Ok, I understand, lets keep it as it is, I didn't realize the gramatics differencies the feature is based on.
We could maybe provide default definitions for some most used languages like EN/DE. Please create an enhancement ticket for it 🙏
There was a problem hiding this comment.
Alright, makes sense.
|
|
||
| if (Object.keys(patterns).includes(language.toLowerCase())) { | ||
| const pattern = patterns[language.toLowerCase()]; |
There was a problem hiding this comment.
| if (Object.keys(patterns).includes(language.toLowerCase())) { | |
| const pattern = patterns[language.toLowerCase()]; | |
| const pattern = patterns[language.toLowerCase()]; | |
| if (pattern) { |
| import { Patterns, CustomPatterns } from "."; | ||
|
|
||
| export const lowerCaseKeys = (patterns: CustomPatterns) => | ||
| Object.keys(patterns).reduce((prev, current) => { | ||
| prev[current.toLowerCase()] = patterns[current]; | ||
| return prev; | ||
| }, {} as CustomPatterns); | ||
|
|
||
| const convertStringToRegex = (string: string) => new RegExp(string, "gim"); | ||
|
|
||
| export const regexifyValues = (patterns: CustomPatterns) => { | ||
| const regexified: Patterns = {}; | ||
|
|
||
| Object.keys(patterns).forEach((key) => { | ||
| regexified[key] = convertStringToRegex(patterns[key]); | ||
| }); | ||
|
|
||
| return regexified; | ||
| }; |
There was a problem hiding this comment.
It's pretty fine to use lodash in lokse as it is cli utility. With it you can simplify your utils 💡
import { lowerCase, mapKeys, mapValues } from 'lodash';
export const lowerCaseKeys = (patterns: CustomPatterns) =>
mapKeys(patterns, lowerCase);
export const regexifyValues = (patterns: CustomPatterns) =>
mapValues(patterns, convertStringToRegex);There was a problem hiding this comment.
Thanks, used a bit differently tho
| // provide custom regex patter without flags | ||
| // default flag is gim | ||
| // use the language code as key (has to be the same as your lang in the spreadsheet) | ||
| "ad-HD": "(\\s|^)(a|i|k|o|s|u|v|z)(\\s+)" |
There was a problem hiding this comment.
do we need to repeat (\\s|^)( and )(\\s+) in each definition? Isn't it a common pattern when looking for a prepositions (or other non standable characters) that could be included in the plugin?
| } | ||
| ] | ||
| } | ||
| ``` |
There was a problem hiding this comment.
Please document default patterns that work out of the box 🙏
| @@ -0,0 +1,12 @@ | |||
| import { mapKeys, mapValues } from "lodash"; | |||
|
|
|||
| import { CustomPatterns } from "."; | |||
There was a problem hiding this comment.
Are you sure about this import from '.' ? 🙂
|
|
||
| import nonBreakingSpacesPlugin from ".."; | ||
|
|
||
| import Transformer from "../../../core/lib/transformer/json"; |
There was a problem hiding this comment.
Please don't break packages encapsulation 🙏
| import Transformer from "../../../core/lib/transformer/json"; | |
| import { transformersByFormat, OutputFormat } from "@lokse/core"; | |
| const Transformer = transformersByFormat[OutputFormat.JSON]; |
| } else { | ||
| options.logger.warn( | ||
| `Pattern for current language ${language} was not found` | ||
| ); |
There was a problem hiding this comment.
- Ok, I understand, lets keep it as it is, I didn't realize the gramatics differencies the feature is based on.
We could maybe provide default definitions for some most used languages like EN/DE. Please create an enhancement ticket for it 🙏
Thanks :) |
— Adds plugin for replacing single-letter characters with non-breaking spaces
— So far only for Czech language, however patterns list could be extend via plugin options
— Later on we could base general find and replace plugin on this one