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

@ -70,9 +70,6 @@ const MentionLink = {
highlightClass () { highlightClass () {
if (this.highlight) return highlightClass(this.user) if (this.highlight) return highlightClass(this.user)
}, },
oldPlace () {
return !this.mergedConfig.mentionsOwnLine
},
oldStyle () { oldStyle () {
return !this.mergedConfig.mentionsNewStyle return !this.mergedConfig.mentionsNewStyle
}, },

View File

@ -1,7 +1,6 @@
<template> <template>
<span <span
class="MentionLink" class="MentionLink"
:class="{ '-oldPlace': oldPlace }"
> >
<!-- eslint-disable vue/no-v-html --> <!-- eslint-disable vue/no-v-html -->
<a <a

View File

@ -1,7 +1,7 @@
import Vue from 'vue' import Vue from 'vue'
import { unescape, flattenDeep } from 'lodash' import { unescape, flattenDeep } from 'lodash'
import { convertHtml, getTagName, processTextForEmoji, getAttrs } from 'src/services/mini_html_converter/mini_html_converter.service.js' import { convertHtmlToTree, getTagName, processTextForEmoji, getAttrs } from 'src/services/html_converter/html_tree_converter.service.js'
import { processHtml } from 'src/services/tiny_post_html_processor/tiny_post_html_processor.service.js' import { convertHtmlToLines } from 'src/services/html_converter/html_line_converter.service.js'
import StillImage from 'src/components/still-image/still-image.vue' import StillImage from 'src/components/still-image/still-image.vue'
import MentionLink from 'src/components/mention_link/mention_link.vue' import MentionLink from 'src/components/mention_link/mention_link.vue'
@ -31,11 +31,24 @@ export default Vue.component('RichContent', {
required: false, required: false,
type: Boolean, type: Boolean,
default: false default: false
},
// Whether to hide last mentions (hellthreads)
hideLastMentions: {
required: false,
type: Boolean,
default: false
},
// Whether to hide first mentions
hideFirstMentions: {
required: false,
type: Boolean,
default: false
} }
}, },
render (h) { render (h) {
// Pre-process HTML // Pre-process HTML
const html = this.greentext ? addGreentext(this.html) : this.html const html = preProcessPerLine(this.html, this.greentext, this.hideLastMentions)
console.log(this.hideFirstMentions, this.hideLastMentions)
const renderImage = (tag) => { const renderImage = (tag) => {
return <StillImage return <StillImage
@ -45,18 +58,20 @@ export default Vue.component('RichContent', {
} }
const renderMention = (attrs, children, encounteredText) => { const renderMention = (attrs, children, encounteredText) => {
return <MentionLink return (this.hideFirstMentions && !encounteredText)
url={attrs.href} ? ''
content={flattenDeep(children).join('')} : <MentionLink
firstMention={!encounteredText} url={attrs.href}
/> content={flattenDeep(children).join('')}
firstMention={!encounteredText}
/>
} }
// We stop treating mentions as "first" ones when we encounter // We stop treating mentions as "first" ones when we encounter
// non-whitespace text // non-whitespace text
let encounteredText = false let encounteredText = false
// Processor to use with mini_html_converter // Processor to use with mini_html_converter
const processItem = (item) => { const processItem = (item, index, array, what) => {
// Handle text nodes - just add emoji // Handle text nodes - just add emoji
if (typeof item === 'string') { if (typeof item === 'string') {
const emptyText = item.trim() === '' const emptyText = item.trim() === ''
@ -69,7 +84,7 @@ export default Vue.component('RichContent', {
encounteredText = true encounteredText = true
} }
if (item.includes(':')) { if (item.includes(':')) {
return processTextForEmoji( unescapedItem = processTextForEmoji(
unescapedItem, unescapedItem,
this.emoji, this.emoji,
({ shortcode, url }) => { ({ shortcode, url }) => {
@ -81,9 +96,8 @@ export default Vue.component('RichContent', {
/> />
} }
) )
} else {
return unescapedItem
} }
return unescapedItem
} }
// Handle tag nodes // Handle tag nodes
@ -98,6 +112,8 @@ export default Vue.component('RichContent', {
const attrs = getAttrs(opener) const attrs = getAttrs(opener)
if (attrs['class'] && attrs['class'].includes('mention')) { if (attrs['class'] && attrs['class'].includes('mention')) {
return renderMention(attrs, children, encounteredText) return renderMention(attrs, children, encounteredText)
} else if (attrs['class'] && attrs['class'].includes('hashtag')) {
return item // We'll handle it later
} else { } else {
attrs.target = '_blank' attrs.target = '_blank'
return <a {...{ attrs }}> return <a {...{ attrs }}>
@ -116,43 +132,129 @@ export default Vue.component('RichContent', {
} }
} }
} }
// Processor for back direction (for finding "last" stuff, just easier this way)
let encounteredTextReverse = false
const renderHashtag = (attrs, children, encounteredTextReverse) => {
attrs.target = '_blank'
if (!encounteredTextReverse) {
attrs['data-parser-last'] = true
}
return <a {...{ attrs }}>
{ children.map(processItem) }
</a>
}
const processItemReverse = (item, index, array, what) => {
// Handle text nodes - just add emoji
if (typeof item === 'string') {
const emptyText = item.trim() === ''
if (emptyText) return encounteredTextReverse ? item : item.trim()
if (!encounteredTextReverse) encounteredTextReverse = true
return item
} else if (Array.isArray(item)) {
// Handle tag nodes
const [opener, children] = item
const Tag = getTagName(opener)
switch (Tag) {
case 'a': // replace mentions with MentionLink
if (!this.handleLinks) break
const attrs = getAttrs(opener)
// should only be this
if (attrs['class'] && attrs['class'].includes('hashtag')) {
return renderHashtag(attrs, children, encounteredTextReverse)
}
}
}
return item
}
return <span class="RichContent"> return <span class="RichContent">
{ this.$slots.prefix } { this.$slots.prefix }
{ convertHtml(html).map(processItem) } { convertHtmlToTree(html).map(processItem).reverse().map(processItemReverse).reverse() }
{ this.$slots.suffix } { this.$slots.suffix }
</span> </span>
} }
}) })
export const addGreentext = (html) => { /** Pre-processing HTML
try { *
if (html.includes('&gt;')) { * Currently this does two things:
// This checks if post has '>' at the beginning, excluding mentions so that @mention >impying works * - add green/cyantexting
return processHtml(html, (string) => { * - wrap and mark last line containing only mentions as ".lastMentionsLine" for
if ( * more compact hellthreads.
string.includes('&gt;') && string *
.replace(/<[^>]+?>/gi, '') // remove all tags * @param {String} html - raw HTML to process
.replace(/@\w+/gi, '') // remove mentions (even failed ones) * @param {Boolean} greentext - whether to enable greentexting or not
.trim() * @param {Boolean} removeLastMentions - whether to remove last mentions
.startsWith('&gt;') */
) { export const preProcessPerLine = (html, greentext, removeLastMentions) => {
return `<span class='greentext'>${string}</span>` // Only mark first (last) encounter
} else { let lastMentionsMarked = false
return string
} return convertHtmlToLines(html).reverse().map((item, index, array) => {
}) if (!item.text) return item
} else { const string = item.text
return html
// Greentext stuff
if (greentext && (string.includes('&gt;') || string.includes('&lt;'))) {
const cleanedString = string.replace(/<[^>]+?>/gi, '') // remove all tags
.replace(/@\w+/gi, '') // remove mentions (even failed ones)
.trim()
if (cleanedString.startsWith('&gt;')) {
return `<span class='greentext'>${string}</span>`
} else if (cleanedString.startsWith('&lt;')) {
return `<span class='cyantext'>${string}</span>`
}
} }
} catch (e) {
console.error('Failed to process status html', e) const tree = convertHtmlToTree(string)
return html
} // If line has loose text, i.e. text outside a mention or a tag
// we won't touch mentions.
let hasLooseText = false
let hasMentions = false
const process = (item) => {
if (Array.isArray(item)) {
const [opener, children, closer] = item
const tag = getTagName(opener)
if (tag === 'a') {
const attrs = getAttrs(opener)
if (attrs['class'] && attrs['class'].includes('mention')) {
hasMentions = true
return [opener, children, closer]
} else {
hasLooseText = true
return [opener, children, closer]
}
} else if (tag === 'span' || tag === 'p') {
return [opener, [...children].reverse().map(process).reverse(), closer]
} else {
hasLooseText = true
return [opener, children, closer]
}
}
if (typeof item === 'string') {
if (item.trim() !== '') {
hasLooseText = true
}
return item
}
}
const result = [...tree].reverse().map(process).reverse()
if (removeLastMentions && hasMentions && !hasLooseText && !lastMentionsMarked) {
lastMentionsMarked = true
return ''
} else {
return flattenDeep(result).join('')
}
}).reverse().join('')
} }
export const getHeadTailLinks = (html) => { export const getHeadTailLinks = (html) => {
// Exported object properties // Exported object properties
const firstMentions = [] // Mentions that appear in the beginning of post body const firstMentions = [] // Mentions that appear in the beginning of post body
const lastMentions = [] // Mentions that appear at the end of post body
const lastTags = [] // Tags that appear at the end of post body const lastTags = [] // Tags that appear at the end of post body
const writtenMentions = [] // All mentions that appear in post body const writtenMentions = [] // All mentions that appear in post body
const writtenTags = [] // All tags that appear in post body const writtenTags = [] // All tags that appear in post body
@ -170,7 +272,7 @@ export const getHeadTailLinks = (html) => {
} }
} }
// Processor to use with mini_html_converter // Processor to use with html_tree_converter
const processItem = (item) => { const processItem = (item) => {
// Handle text nodes - stop treating mentions as "first" when text encountered // Handle text nodes - stop treating mentions as "first" when text encountered
if (typeof item === 'string') { if (typeof item === 'string') {
@ -182,6 +284,7 @@ export const getHeadTailLinks = (html) => {
} }
// Encountered text? That means tags we've been collectings aren't "last"! // Encountered text? That means tags we've been collectings aren't "last"!
lastTags.splice(0) lastTags.splice(0)
lastMentions.splice(0)
return return
} }
// Handle tag nodes // Handle tag nodes
@ -197,6 +300,7 @@ export const getHeadTailLinks = (html) => {
firstMentions.push(linkData) firstMentions.push(linkData)
} }
writtenMentions.push(linkData) writtenMentions.push(linkData)
lastMentions.push(linkData)
} else if (attrs['class'].includes('hashtag')) { } else if (attrs['class'].includes('hashtag')) {
lastTags.push(linkData) lastTags.push(linkData)
writtenTags.push(linkData) writtenTags.push(linkData)
@ -206,6 +310,6 @@ export const getHeadTailLinks = (html) => {
children && children.forEach(processItem) children && children.forEach(processItem)
} }
} }
convertHtml(html).forEach(processItem) convertHtmlToTree(html).forEach(processItem)
return { firstMentions, writtenMentions, writtenTags, lastTags } return { firstMentions, writtenMentions, writtenTags, lastTags, lastMentions }
} }

View File

@ -196,6 +196,9 @@ const Status = {
hasMentionsLine () { hasMentionsLine () {
return this.mentionsLine.length > 0 return this.mentionsLine.length > 0
}, },
hideLastMentions () {
return this.headTailLinks.firstMentions.length === 0
},
muted () { muted () {
if (this.statusoid.user.id === this.currentUser.id) return false if (this.statusoid.user.id === this.currentUser.id) return false
const { status } = this const { status } = this

View File

@ -306,6 +306,7 @@
:highlight="highlight" :highlight="highlight"
:focused="isFocused" :focused="isFocused"
:hide-first-mentions="mentionsOwnLine && isReply" :hide-first-mentions="mentionsOwnLine && isReply"
:hide-last-mentions="hideLastMentions"
:head-tail-links="headTailLinks" :head-tail-links="headTailLinks"
@mediaplay="addMediaPlaying($event)" @mediaplay="addMediaPlaying($event)"
@mediapause="removeMediaPlaying($event)" @mediapause="removeMediaPlaying($event)"

View File

@ -30,7 +30,8 @@ const StatusContent = {
// if this was computed at upper level it can be passed here, otherwise // if this was computed at upper level it can be passed here, otherwise
// it will be in this component // it will be in this component
'headTailLinks', 'headTailLinks',
'hideFirstMentions' 'hideFirstMentions',
'hideLastMentions'
], ],
data () { data () {
return { return {
@ -80,9 +81,12 @@ const StatusContent = {
attachmentTypes () { attachmentTypes () {
return this.status.attachments.map(file => fileType.fileType(file.mimetype)) return this.status.attachments.map(file => fileType.fileType(file.mimetype))
}, },
mentions () { mentionsFirst () {
return this.headTailLinksComputed.firstMentions return this.headTailLinksComputed.firstMentions
}, },
mentionsLast () {
return this.headTailLinksComputed.lastMentions
},
...mapGetters(['mergedConfig']) ...mapGetters(['mergedConfig'])
}, },
components: { components: {

View File

@ -49,11 +49,19 @@
:emoji="status.emojis" :emoji="status.emojis"
:handle-links="true" :handle-links="true"
:greentext="mergedConfig.greentext" :greentext="mergedConfig.greentext"
:hide-first-mentions="hideFirstMentions"
:hide-last-mentions="hideLastMentions"
> >
<template v-slot:prefix> <template v-slot:prefix>
<MentionsLine <MentionsLine
v-if="!hideFirstMentions" v-if="!hideFirstMentions && mentionsFirst"
:mentions="mentions" :mentions="mentionsFirst"
/>
</template>
<template v-slot:suffix>
<MentionsLine
v-if="!hideFirstMentions && mentionsLast"
:mentions="mentionsLast"
/> />
</template> </template>
</RichContent> </RichContent>

View File

@ -33,6 +33,7 @@ const StatusContent = {
'fullContent', 'fullContent',
'singleLine', 'singleLine',
'hideFirstMentions', 'hideFirstMentions',
'hideLastMentions',
'headTailLinks' 'headTailLinks'
], ],
computed: { computed: {

View File

@ -5,7 +5,8 @@
:status="status" :status="status"
:single-line="singleLine" :single-line="singleLine"
:hide-first-mentions="hideFirstMentions" :hide-first-mentions="hideFirstMentions"
:headTailLinks="headTailLinks" :hide-last-mentions="hideLastMentions"
:head-tail-links="headTailLinks"
> >
<div v-if="status.poll && status.poll.options"> <div v-if="status.poll && status.poll.options">
<poll :base-poll="status.poll" /> <poll :base-poll="status.poll" />

View File

@ -1,18 +1,26 @@
/** /**
* This is a tiny purpose-built HTML parser/processor. This basically detects any type of visual newline and * This is a tiny purpose-built HTML parser/processor. This basically detects
* allows it to be processed, useful for greentexting, mostly * 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 * known issue: doesn't handle CDATA so nested CDATA might not work well
* *
* @param {Object} input - input data * @param {Object} input - input data
* @param {(string) => string} processor - function that will be called on every line * @return {(string|{ text: string })[]} processed html in form of a list.
* @return {string} processed html
*/ */
export const processHtml = (html, processor) => { export const convertHtmlToLines = (html) => {
const handledTags = new Set(['p', 'br', 'div']) const handledTags = new Set(['p', 'br', 'div'])
const openCloseTags = new Set(['p', '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 const level = [] // How deep we are in tags and which tags were there
let textBuffer = '' // Current line content let textBuffer = '' // Current line content
let tagBuffer = null // Current tag buffer, if null = we are not currently reading a tag 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 const flush = () => { // Processes current line buffer, adds it to output buffer and clears line buffer
if (textBuffer.trim().length > 0) { if (textBuffer.trim().length > 0) {
buffer += processor(textBuffer) buffer.push({ text: textBuffer })
} else { } else {
buffer += textBuffer buffer.push(textBuffer)
} }
textBuffer = '' textBuffer = ''
} }
const handleBr = (tag) => { // handles single newlines/linebreaks/selfclosing const handleBr = (tag) => { // handles single newlines/linebreaks/selfclosing
flush() flush()
buffer += tag buffer.push(tag)
} }
const handleOpen = (tag) => { // handles opening tags const handleOpen = (tag) => { // handles opening tags
flush() flush()
buffer += tag buffer.push(tag)
level.push(tag) level.push(tag)
} }
const handleClose = (tag) => { // handles closing tags const handleClose = (tag) => { // handles closing tags
flush() flush()
buffer += tag buffer.push(tag)
if (level[level.length - 1] === tag) { if (level[level.length - 1] === tag) {
level.pop() level.pop()
} }

View File

@ -1,15 +1,23 @@
/** /**
* This is a not-so-tiny purpose-built HTML parser/processor. It was made for use * This is a not-so-tiny purpose-built HTML parser/processor. This parses html
* with StatusBody component for purpose of replacing tags with vue components * 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 {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 * @return {string} processed html
*/ */
export const convertHtml = (html) => { export const convertHtmlToTree = (html) => {
// Elements that are implicitly self-closing // Elements that are implicitly self-closing
// https://developer.mozilla.org/en-US/docs/Glossary/empty_element // https://developer.mozilla.org/en-US/docs/Glossary/empty_element
const emptyElements = new Set([ const emptyElements = new Set([

View File

@ -1,46 +1,64 @@
import { processHtml } from 'src/services/tiny_post_html_processor/tiny_post_html_processor.service.js' import { convertHtmlToLines } from 'src/services/html_converter/html_line_converter.service.js'
const mapOnlyText = (processor) => (input) => input.text ? processor(input.text) : input
describe('TinyPostHTMLProcessor', () => { describe('TinyPostHTMLProcessor', () => {
describe('with processor that keeps original line should not make any changes to HTML when', () => { describe('with processor that keeps original line should not make any changes to HTML when', () => {
const processorKeep = (line) => line const processorKeep = (line) => line
it('fed with regular HTML with newlines', () => { it('fed with regular HTML with newlines', () => {
const inputOutput = '1<br/>2<p class="lol">3 4</p> 5 \n 6 <p > 7 <br> 8 </p> <br>\n<br/>' const inputOutput = '1<br/>2<p class="lol">3 4</p> 5 \n 6 <p > 7 <br> 8 </p> <br>\n<br/>'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with possibly broken HTML with invalid tags/composition', () => { it('fed with possibly broken HTML with invalid tags/composition', () => {
const inputOutput = '<feeee dwdwddddddw> <i>ayy<b>lm</i>ao</b> </section>' const inputOutput = '<feeee dwdwddddddw> <i>ayy<b>lm</i>ao</b> </section>'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with very broken HTML with broken composition', () => { it('fed with very broken HTML with broken composition', () => {
const inputOutput = '</p> lmao what </div> whats going on <div> wha <p>' const inputOutput = '</p> lmao what </div> whats going on <div> wha <p>'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with sorta valid HTML but tags aren\'t closed', () => { it('fed with sorta valid HTML but tags aren\'t closed', () => {
const inputOutput = 'just leaving a <div> hanging' const inputOutput = 'just leaving a <div> hanging'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with not really HTML at this point... tags that aren\'t finished', () => { it('fed with not really HTML at this point... tags that aren\'t finished', () => {
const inputOutput = 'do you expect me to finish this <div class=' const inputOutput = 'do you expect me to finish this <div class='
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with dubiously valid HTML (p within p and also div inside p)', () => { it('fed with dubiously valid HTML (p within p and also div inside p)', () => {
const inputOutput = 'look ma <p> p \nwithin <p> p! </p> and a <br/><div>div!</div></p>' const inputOutput = 'look ma <p> p \nwithin <p> p! </p> and a <br/><div>div!</div></p>'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with maybe valid HTML? self-closing divs and ps', () => { it('fed with maybe valid HTML? self-closing divs and ps', () => {
const inputOutput = 'a <div class="what"/> what now <p aria-label="wtf"/> ?' const inputOutput = 'a <div class="what"/> what now <p aria-label="wtf"/> ?'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
it('fed with valid XHTML containing a CDATA', () => { it('fed with valid XHTML containing a CDATA', () => {
const inputOutput = 'Yes, it is me, <![CDATA[DIO]]>' const inputOutput = 'Yes, it is me, <![CDATA[DIO]]>'
expect(processHtml(inputOutput, processorKeep)).to.eql(inputOutput) const result = convertHtmlToLines(inputOutput)
const comparableResult = result.map(mapOnlyText(processorKeep)).join('')
expect(comparableResult).to.eql(inputOutput)
}) })
}) })
describe('with processor that replaces lines with word "_" should match expected line when', () => { describe('with processor that replaces lines with word "_" should match expected line when', () => {
@ -48,49 +66,65 @@ describe('TinyPostHTMLProcessor', () => {
it('fed with regular HTML with newlines', () => { it('fed with regular HTML with newlines', () => {
const input = '1<br/>2<p class="lol">3 4</p> 5 \n 6 <p > 7 <br> 8 </p> <br>\n<br/>' const input = '1<br/>2<p class="lol">3 4</p> 5 \n 6 <p > 7 <br> 8 </p> <br>\n<br/>'
const output = '_<br/>_<p class="lol">_</p>_\n_<p >_<br>_</p> <br>\n<br/>' const output = '_<br/>_<p class="lol">_</p>_\n_<p >_<br>_</p> <br>\n<br/>'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with possibly broken HTML with invalid tags/composition', () => { it('fed with possibly broken HTML with invalid tags/composition', () => {
const input = '<feeee dwdwddddddw> <i>ayy<b>lm</i>ao</b> </section>' const input = '<feeee dwdwddddddw> <i>ayy<b>lm</i>ao</b> </section>'
const output = '_' const output = '_'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with very broken HTML with broken composition', () => { it('fed with very broken HTML with broken composition', () => {
const input = '</p> lmao what </div> whats going on <div> wha <p>' const input = '</p> lmao what </div> whats going on <div> wha <p>'
const output = '</p>_</div>_<div>_<p>' const output = '</p>_</div>_<div>_<p>'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with sorta valid HTML but tags aren\'t closed', () => { it('fed with sorta valid HTML but tags aren\'t closed', () => {
const input = 'just leaving a <div> hanging' const input = 'just leaving a <div> hanging'
const output = '_<div>_' const output = '_<div>_'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with not really HTML at this point... tags that aren\'t finished', () => { it('fed with not really HTML at this point... tags that aren\'t finished', () => {
const input = 'do you expect me to finish this <div class=' const input = 'do you expect me to finish this <div class='
const output = '_' const output = '_'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with dubiously valid HTML (p within p and also div inside p)', () => { it('fed with dubiously valid HTML (p within p and also div inside p)', () => {
const input = 'look ma <p> p \nwithin <p> p! </p> and a <br/><div>div!</div></p>' const input = 'look ma <p> p \nwithin <p> p! </p> and a <br/><div>div!</div></p>'
const output = '_<p>_\n_<p>_</p>_<br/><div>_</div></p>' const output = '_<p>_\n_<p>_</p>_<br/><div>_</div></p>'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with maybe valid HTML? self-closing divs and ps', () => { it('fed with maybe valid HTML? self-closing divs and ps', () => {
const input = 'a <div class="what"/> what now <p aria-label="wtf"/> ?' const input = 'a <div class="what"/> what now <p aria-label="wtf"/> ?'
const output = '_<div class="what"/>_<p aria-label="wtf"/>_' const output = '_<div class="what"/>_<p aria-label="wtf"/>_'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
it('fed with valid XHTML containing a CDATA', () => { it('fed with valid XHTML containing a CDATA', () => {
const input = 'Yes, it is me, <![CDATA[DIO]]>' const input = 'Yes, it is me, <![CDATA[DIO]]>'
const output = '_' const output = '_'
expect(processHtml(input, processorReplace)).to.eql(output) const result = convertHtmlToLines(input)
const comparableResult = result.map(mapOnlyText(processorReplace)).join('')
expect(comparableResult).to.eql(output)
}) })
}) })
}) })

View File

@ -1,10 +1,10 @@
import { convertHtml, processTextForEmoji, getAttrs } from 'src/services/mini_html_converter/mini_html_converter.service.js' import { convertHtmlToTree, processTextForEmoji, getAttrs } from 'src/services/html_converter/html_tree_converter.service.js'
describe('MiniHtmlConverter', () => { describe('MiniHtmlConverter', () => {
describe('convertHtml', () => { describe('convertHtmlToTree', () => {
it('converts html into a tree structure', () => { it('converts html into a tree structure', () => {
const input = '1 <p>2</p> <b>3<img src="a">4</b>5' const input = '1 <p>2</p> <b>3<img src="a">4</b>5'
expect(convertHtml(input)).to.eql([ expect(convertHtmlToTree(input)).to.eql([
'1 ', '1 ',
[ [
'<p>', '<p>',
@ -26,7 +26,7 @@ describe('MiniHtmlConverter', () => {
}) })
it('converts html to tree while preserving tag formatting', () => { it('converts html to tree while preserving tag formatting', () => {
const input = '1 <p >2</p><b >3<img src="a">4</b>5' const input = '1 <p >2</p><b >3<img src="a">4</b>5'
expect(convertHtml(input)).to.eql([ expect(convertHtmlToTree(input)).to.eql([
'1 ', '1 ',
[ [
'<p >', '<p >',
@ -47,7 +47,7 @@ describe('MiniHtmlConverter', () => {
}) })
it('converts semi-broken html', () => { it('converts semi-broken html', () => {
const input = '1 <br> 2 <p> 42' const input = '1 <br> 2 <p> 42'
expect(convertHtml(input)).to.eql([ expect(convertHtmlToTree(input)).to.eql([
'1 ', '1 ',
['<br>'], ['<br>'],
' 2 ', ' 2 ',
@ -59,7 +59,7 @@ describe('MiniHtmlConverter', () => {
}) })
it('realistic case 1', () => { it('realistic case 1', () => {
const input = '<p><span class="h-card"><a class="u-url mention" data-user="9wRC6T2ZZiKWJ0vUi8" href="https://cawfee.club/users/benis" rel="ugc">@<span>benis</span></a></span> <span class="h-card"><a class="u-url mention" data-user="194" href="https://shigusegubu.club/users/hj" rel="ugc">@<span>hj</span></a></span> nice</p>' const input = '<p><span class="h-card"><a class="u-url mention" data-user="9wRC6T2ZZiKWJ0vUi8" href="https://cawfee.club/users/benis" rel="ugc">@<span>benis</span></a></span> <span class="h-card"><a class="u-url mention" data-user="194" href="https://shigusegubu.club/users/hj" rel="ugc">@<span>hj</span></a></span> nice</p>'
expect(convertHtml(input)).to.eql([ expect(convertHtmlToTree(input)).to.eql([
[ [
'<p>', '<p>',
[ [
@ -112,7 +112,7 @@ describe('MiniHtmlConverter', () => {
}) })
it('realistic case 2', () => { it('realistic case 2', () => {
const inputOutput = 'Country improv: give me a city<br/>Audience: Memphis<br/>Improv troupe: come on, a better one<br/>Audience: el paso' const inputOutput = 'Country improv: give me a city<br/>Audience: Memphis<br/>Improv troupe: come on, a better one<br/>Audience: el paso'
expect(convertHtml(inputOutput)).to.eql([ expect(convertHtmlToTree(inputOutput)).to.eql([
'Country improv: give me a city', 'Country improv: give me a city',
[ [
'<br/>' '<br/>'