virt2/api/soft/CodeMirror/lib/util/bidi.js

215 lines
8.2 KiB
JavaScript
Raw Permalink Normal View History

import { lst } from "./misc.js"
// BIDI HELPERS
export function iterateBidiSections(order, from, to, f) {
if (!order) return f(from, to, "ltr", 0)
let found = false
for (let i = 0; i < order.length; ++i) {
let part = order[i]
if (part.from < to && part.to > from || from == to && part.to == from) {
f(Math.max(part.from, from), Math.min(part.to, to), part.level == 1 ? "rtl" : "ltr", i)
found = true
}
}
if (!found) f(from, to, "ltr")
}
export let bidiOther = null
export function getBidiPartAt(order, ch, sticky) {
let found
bidiOther = null
for (let i = 0; i < order.length; ++i) {
let cur = order[i]
if (cur.from < ch && cur.to > ch) return i
if (cur.to == ch) {
if (cur.from != cur.to && sticky == "before") found = i
else bidiOther = i
}
if (cur.from == ch) {
if (cur.from != cur.to && sticky != "before") found = i
else bidiOther = i
}
}
return found != null ? found : bidiOther
}
// Bidirectional ordering algorithm
// See http://unicode.org/reports/tr9/tr9-13.html for the algorithm
// that this (partially) implements.
// One-char codes used for character types:
// L (L): Left-to-Right
// R (R): Right-to-Left
// r (AL): Right-to-Left Arabic
// 1 (EN): European Number
// + (ES): European Number Separator
// % (ET): European Number Terminator
// n (AN): Arabic Number
// , (CS): Common Number Separator
// m (NSM): Non-Spacing Mark
// b (BN): Boundary Neutral
// s (B): Paragraph Separator
// t (S): Segment Separator
// w (WS): Whitespace
// N (ON): Other Neutrals
// Returns null if characters are ordered as they appear
// (left-to-right), or an array of sections ({from, to, level}
// objects) in the order in which they occur visually.
let bidiOrdering = (function() {
// Character types for codepoints 0 to 0xff
let lowTypes = "bbbbbbbbbtstwsbbbbbbbbbbbbbbssstwNN%%%NNNNNN,N,N1111111111NNNNNNNLLLLLLLLLLLLLLLLLLLLLLLLLLNNNNNNLLLLLLLLLLLLLLLLLLLLLLLLLLNNNNbbbbbbsbbbbbbbbbbbbbbbbbbbbbbbbbb,N%%%%NNNNLNNNNN%%11NLNNN1LNNNNNLLLLLLLLLLLLLLLLLLLLLLLNLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLN"
// Character types for codepoints 0x600 to 0x6f9
let arabicTypes = "nnnnnnNNr%%r,rNNmmmmmmmmmmmrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrmmmmmmmmmmmmmmmmmmmmmnnnnnnnnnn%nnrrrmrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrmmmmmmmnNmmmmmmrrmmNmmmmrr1111111111"
function charType(code) {
if (code <= 0xf7) return lowTypes.charAt(code)
else if (0x590 <= code && code <= 0x5f4) return "R"
else if (0x600 <= code && code <= 0x6f9) return arabicTypes.charAt(code - 0x600)
else if (0x6ee <= code && code <= 0x8ac) return "r"
else if (0x2000 <= code && code <= 0x200b) return "w"
else if (code == 0x200c) return "b"
else return "L"
}
let bidiRE = /[\u0590-\u05f4\u0600-\u06ff\u0700-\u08ac]/
let isNeutral = /[stwN]/, isStrong = /[LRr]/, countsAsLeft = /[Lb1n]/, countsAsNum = /[1n]/
function BidiSpan(level, from, to) {
this.level = level
this.from = from; this.to = to
}
return function(str, direction) {
let outerType = direction == "ltr" ? "L" : "R"
if (str.length == 0 || direction == "ltr" && !bidiRE.test(str)) return false
let len = str.length, types = []
for (let i = 0; i < len; ++i)
types.push(charType(str.charCodeAt(i)))
// W1. Examine each non-spacing mark (NSM) in the level run, and
// change the type of the NSM to the type of the previous
// character. If the NSM is at the start of the level run, it will
// get the type of sor.
for (let i = 0, prev = outerType; i < len; ++i) {
let type = types[i]
if (type == "m") types[i] = prev
else prev = type
}
// W2. Search backwards from each instance of a European number
// until the first strong type (R, L, AL, or sor) is found. If an
// AL is found, change the type of the European number to Arabic
// number.
// W3. Change all ALs to R.
for (let i = 0, cur = outerType; i < len; ++i) {
let type = types[i]
if (type == "1" && cur == "r") types[i] = "n"
else if (isStrong.test(type)) { cur = type; if (type == "r") types[i] = "R" }
}
// W4. A single European separator between two European numbers
// changes to a European number. A single common separator between
// two numbers of the same type changes to that type.
for (let i = 1, prev = types[0]; i < len - 1; ++i) {
let type = types[i]
if (type == "+" && prev == "1" && types[i+1] == "1") types[i] = "1"
else if (type == "," && prev == types[i+1] &&
(prev == "1" || prev == "n")) types[i] = prev
prev = type
}
// W5. A sequence of European terminators adjacent to European
// numbers changes to all European numbers.
// W6. Otherwise, separators and terminators change to Other
// Neutral.
for (let i = 0; i < len; ++i) {
let type = types[i]
if (type == ",") types[i] = "N"
else if (type == "%") {
let end
for (end = i + 1; end < len && types[end] == "%"; ++end) {}
let replace = (i && types[i-1] == "!") || (end < len && types[end] == "1") ? "1" : "N"
for (let j = i; j < end; ++j) types[j] = replace
i = end - 1
}
}
// W7. Search backwards from each instance of a European number
// until the first strong type (R, L, or sor) is found. If an L is
// found, then change the type of the European number to L.
for (let i = 0, cur = outerType; i < len; ++i) {
let type = types[i]
if (cur == "L" && type == "1") types[i] = "L"
else if (isStrong.test(type)) cur = type
}
// N1. A sequence of neutrals takes the direction of the
// surrounding strong text if the text on both sides has the same
// direction. European and Arabic numbers act as if they were R in
// terms of their influence on neutrals. Start-of-level-run (sor)
// and end-of-level-run (eor) are used at level run boundaries.
// N2. Any remaining neutrals take the embedding direction.
for (let i = 0; i < len; ++i) {
if (isNeutral.test(types[i])) {
let end
for (end = i + 1; end < len && isNeutral.test(types[end]); ++end) {}
let before = (i ? types[i-1] : outerType) == "L"
let after = (end < len ? types[end] : outerType) == "L"
let replace = before == after ? (before ? "L" : "R") : outerType
for (let j = i; j < end; ++j) types[j] = replace
i = end - 1
}
}
// Here we depart from the documented algorithm, in order to avoid
// building up an actual levels array. Since there are only three
// levels (0, 1, 2) in an implementation that doesn't take
// explicit embedding into account, we can build up the order on
// the fly, without following the level-based algorithm.
let order = [], m
for (let i = 0; i < len;) {
if (countsAsLeft.test(types[i])) {
let start = i
for (++i; i < len && countsAsLeft.test(types[i]); ++i) {}
order.push(new BidiSpan(0, start, i))
} else {
let pos = i, at = order.length
for (++i; i < len && types[i] != "L"; ++i) {}
for (let j = pos; j < i;) {
if (countsAsNum.test(types[j])) {
if (pos < j) order.splice(at, 0, new BidiSpan(1, pos, j))
let nstart = j
for (++j; j < i && countsAsNum.test(types[j]); ++j) {}
order.splice(at, 0, new BidiSpan(2, nstart, j))
pos = j
} else ++j
}
if (pos < i) order.splice(at, 0, new BidiSpan(1, pos, i))
}
}
if (direction == "ltr") {
if (order[0].level == 1 && (m = str.match(/^\s+/))) {
order[0].from = m[0].length
order.unshift(new BidiSpan(0, 0, m[0].length))
}
if (lst(order).level == 1 && (m = str.match(/\s+$/))) {
lst(order).to -= m[0].length
order.push(new BidiSpan(0, len - m[0].length, len))
}
}
return direction == "rtl" ? order.reverse() : order
}
})()
// Get the bidi ordering for the given line (and cache it). Returns
// false for lines that are fully left-to-right, and an array of
// BidiSpan objects otherwise.
export function getOrder(line, direction) {
let order = line.order
if (order == null) order = line.order = bidiOrdering(line.text, direction)
return order
}