Hellthread(tm) Certified

This commit is contained in:
Henry Jameson 2021-06-10 18:52:01 +03:00
parent 0f73e96194
commit cc00af7a31
13 changed files with 257 additions and 89 deletions

View file

@ -1,18 +1,26 @@
/**
* This is a tiny purpose-built HTML parser/processor. This basically detects any type of visual newline and
* allows it to be processed, useful for greentexting, mostly
* This is a tiny purpose-built HTML parser/processor. This basically detects
* any type of visual newline and converts entire HTML into a array structure.
*
* Text nodes are represented as object with single property - text - containing
* the visual line. Intended usage is to process the array with .map() in which
* map function returns a string and resulting array can be converted back to html
* with a .join('').
*
* Generally this isn't very useful except for when you really need to either
* modify visual lines (greentext i.e. simple quoting) or do something with
* first/last line.
*
* known issue: doesn't handle CDATA so nested CDATA might not work well
*
* @param {Object} input - input data
* @param {(string) => string} processor - function that will be called on every line
* @return {string} processed html
* @return {(string|{ text: string })[]} processed html in form of a list.
*/
export const processHtml = (html, processor) => {
export const convertHtmlToLines = (html) => {
const handledTags = new Set(['p', 'br', 'div'])
const openCloseTags = new Set(['p', 'div'])
let buffer = '' // Current output buffer
let buffer = [] // Current output buffer
const level = [] // How deep we are in tags and which tags were there
let textBuffer = '' // Current line content
let tagBuffer = null // Current tag buffer, if null = we are not currently reading a tag
@ -25,27 +33,27 @@ export const processHtml = (html, processor) => {
const flush = () => { // Processes current line buffer, adds it to output buffer and clears line buffer
if (textBuffer.trim().length > 0) {
buffer += processor(textBuffer)
buffer.push({ text: textBuffer })
} else {
buffer += textBuffer
buffer.push(textBuffer)
}
textBuffer = ''
}
const handleBr = (tag) => { // handles single newlines/linebreaks/selfclosing
flush()
buffer += tag
buffer.push(tag)
}
const handleOpen = (tag) => { // handles opening tags
flush()
buffer += tag
buffer.push(tag)
level.push(tag)
}
const handleClose = (tag) => { // handles closing tags
flush()
buffer += tag
buffer.push(tag)
if (level[level.length - 1] === tag) {
level.pop()
}

View file

@ -1,15 +1,23 @@
/**
* This is a not-so-tiny purpose-built HTML parser/processor. It was made for use
* with StatusBody component for purpose of replacing tags with vue components
* This is a not-so-tiny purpose-built HTML parser/processor. This parses html
* and converts it into a tree structure representing tag openers/closers and
* children.
*
* known issue: doesn't handle CDATA so nested CDATA might not work well
* Structure follows this pattern: [opener, [...children], closer] except root
* node which is just [...children]. Text nodes can only be within children and
* are represented as strings.
*
* Intended use is to convert HTML structure and then recursively iterate over it
* most likely using a map. Very useful for dynamically rendering html replacing
* tags with JSX elements in a render function.
*
* known issue: doesn't handle CDATA so CDATA might not work well
* known issue: doesn't handle HTML comments
*
* @param {Object} input - input data
* @param {(string) => string} lineProcessor - function that will be called on every line
* @param {{ key[string]: (string) => string}} tagProcessor - map of processors for tags
* @return {string} processed html
*/
export const convertHtml = (html) => {
export const convertHtmlToTree = (html) => {
// Elements that are implicitly self-closing
// https://developer.mozilla.org/en-US/docs/Glossary/empty_element
const emptyElements = new Set([