dymokomi/luce-prism
Source of release 0.6.1
0.6.1 / src / editor / highlight.lucb
## Line-local byte spans; applications choose colors and typography.
import interop
import memory
from codecs.api import Codec
from codecs.common import find
from storage import List
from types import invalid
pub enum TokenClass as u8:
normal = 0
heading = 1
emphasis = 2
code = 3
marker = 4
comment = 5
keyword = 6
string = 7
number = 8
entity = 9
note = 10
pub struct TokenSpan:
pub var start: usize
pub var end: usize
pub var kind: TokenClass
pub struct TokenSpans:
var spans: List[TokenSpan]
pub func init():
return
pub func count() -> usize:
return self.spans.count
pub func at(index: usize) -> TokenSpan!:
if index >= self.spans.count:
error(invalid, "Prism token span index is out of bounds")
return self.spans.items[index]
pub func class_at(column: usize) -> TokenClass:
for span in self.spans.span():
if column >= span.start and column < span.end:
return span.kind
return TokenClass.normal
func add(start: usize, end: usize, kind: TokenClass) -> !:
try self.spans.push(TokenSpan(start, end, kind))
pub func close():
self.spans.close()
pub let token_spans_type = interop.Type[TokenSpans]("Prism token spans", TokenSpans.close, closeable = true)
func digit(byte: u8) -> bool:
return byte >= '0' and byte <= '9'
func alpha(byte: u8) -> bool:
return (byte >= 'A' and byte <= 'Z') or (byte >= 'a' and byte <= 'z')
func word(byte: u8) -> bool:
return alpha(byte) or digit(byte) or byte == '_'
func markdown_inline(line: str, start: usize, output: TokenSpans*) -> !:
var i = start
while i < line.length:
if line.bytes[i] == '`':
if let end = find(line, "`", i + 1):
try output.add(i, end + 1, TokenClass.code)
i = end + 1
continue
if i + 1 < line.length and (str)line.bytes[i..<i + 2] == "**":
if let end = find(line, "**", i + 2):
try output.add(i, end + 2, TokenClass.emphasis)
i = end + 2
continue
i += 1
func markdown(line: str, output: TokenSpans*) -> !:
var i: usize = 0
let n = line.length
while i < n and line.bytes[i] == ' ':
i += 1
if i < n and line.bytes[i] == '#':
var h = i
while h < n and line.bytes[h] == '#':
h += 1
if h - i <= 6 and (h == n or line.bytes[h] == ' '):
try output.add(i, h, TokenClass.marker)
if h < n:
try output.add(h, n, TokenClass.heading)
return
if i + 2 < n and (str)line.bytes[i..<i + 3] == "```":
try output.add(i, n, TokenClass.marker)
return
if i < n and line.bytes[i] == '>':
try output.add(i, i + 1, TokenClass.marker)
if i + 1 < n:
try output.add(i + 1, n, TokenClass.comment)
return
if i + 1 < n and (line.bytes[i] == '-' or line.bytes[i] == '*' or line.bytes[i] == '+') and line.bytes[i + 1] == ' ':
try output.add(i, i + 1, TokenClass.marker)
try markdown_inline(line, i + 2, output)
return
var d = i
while d < n and digit(line.bytes[d]):
d += 1
if d > i and d + 1 < n and line.bytes[d] == '.' and line.bytes[d + 1] == ' ':
try output.add(i, d + 1, TokenClass.marker)
try markdown_inline(line, d + 2, output)
return
try markdown_inline(line, i, output)
func source(line: str, json: bool, output: TokenSpans*) -> !:
var i: usize = 0
var first = true
let n = line.length
while i < n:
let byte = line.bytes[i]
if not json and byte == '#':
try output.add(i, n, TokenClass.comment)
return
if byte == '"':
var j = i + 1
while j < n and line.bytes[j] != '"':
if json and line.bytes[j] == '\\':
j += 1
j += 1
let end = j + 1 if j < n else n
var k = end
while k < n and line.bytes[k] == ' ':
k += 1
try output.add(i, end, TokenClass.keyword if json and k < n and line.bytes[k] == ':' else TokenClass.string)
i = end
continue
if not json and byte == '<':
if let end = find(line, ">", i + 1):
try output.add(i, end + 1, TokenClass.marker)
i = end + 1
continue
if digit(byte) or (byte == '-' and i + 1 < n and digit(line.bytes[i + 1])):
var j = i + 1 if byte == '-' else i
while j < n:
let c = line.bytes[j]
if not alpha(c) and not digit(c) and c != '.' and not (json and (c == '+' or c == '-')):
break
j += 1
try output.add(i, j, TokenClass.number)
i = j
continue
if alpha(byte) if json else word(byte):
var j = i
while j < n and (alpha(line.bytes[j]) if json else (word(line.bytes[j]) or line.bytes[j] == '.')):
j += 1
let token = (str)line.bytes[i..<j]
if (token == "true" or token == "false" or token == "null") if json else (first or token == "def" or token == "over"):
try output.add(i, j, TokenClass.keyword)
first = false
i = j
continue
i += 1
func html(line: str, output: TokenSpans*) -> !:
var i: usize = 0
while i < line.length:
if line.bytes[i] == '<':
let comment = i + 4 <= line.length and (str)line.bytes[i..<i + 4] == "<!--"
var end = line.length
if let position = find(line, "-->" if comment else ">", i + 1):
end = position + (3 if comment else 1)
try output.add(i, end, TokenClass.comment if comment else TokenClass.marker)
i = end
else:
i += 1
pub func highlight_line(line: str, codec: Codec) -> interop.Reference[TokenSpans]!:
var spans = TokenSpans()
errdefer spans.close()
if codec == Codec.markdown:
try markdown(line, &spans)
elif codec == Codec.prisma or codec == Codec.prism or codec == Codec.json:
try source(line, codec == Codec.json, &spans)
elif codec == Codec.html or codec == Codec.svg:
try html(line, &spans)
let owned = try new TokenSpans() in memory.heap
owned[0] = spans
spans = TokenSpans()
return try interop.Reference[TokenSpans].adopt(owned, token_spans_type, memory.heap)