NOTE: currently we support interpreted mode and Python source code generation.
It's EASY to compile compiled_unit into source code for C#, F# and others. See CodeGen.Python.fs for how to write a custom backend.
Fable.Sedlex
Thanks to the Fable.Python compiler, this project managed to modify code from the amazing OCaml sedlex, implementing an ultimate lexer generator for (currently) both Python and F#.
The most impressive feature of sedlex is that Sedlex statically analyses regular expressions built with lexer combinators, correctly ordering/merging/optimizing the lexical rules. In short, sedlex is correct and efficient.
For Python users: the generated Python code is located in fable_sedlex directory, and you can directly copy them to your Python package to access such lexer generator.
For .NET(C#, F#) users: copying Sedlex.fs will give you such lexer generator in .NET.
For users in other programming languages: you might access compiled_unit and write a code generator.
from fable_sedlex.sedlex import *
from fable_sedlex.code_gen_python import codegen_python
from fable_sedlex.code_gen import show_doc
digit = pinterval(ord('0'), ord('9'))
dquote = pchar('"')
backslash = pchar('\\')
dot = pchar('.')
exp = pseq([por(pchar('E'), pchar('e')), pplus(digit)])
flt = pseq([pstar(digit), dot, pplus(digit), popt(exp)])
integral = pseq([pplus(digit), popt(exp)])
string_lit = pseq([dquote, pstar(por(pcompl(dquote), pseq([backslash, pany]))), dquote])
space = pchars(["\n", "\t", "\r", ' '])
EOF_ID = 0
cu = build(
[
(space, Lexer_discard),
(flt, Lexer_tokenize(1)),
(integral, Lexer_tokenize(2)),
(string_lit, Lexer_tokenize(3)),
(pstring("+"), Lexer_tokenize(4)),
(pstring("+="), Lexer_tokenize(4)),
(peof, Lexer_tokenize(EOF_ID))
], "my error")
f = compile_inline_thread(cu)
# we can also generate Python source code for equivalent lexer via:
# `code = show_doc(codegen_python(cu))`
# see `test_codegen_py.py`, `generated.py` and `test_run_generated.py`
# for more details
buf = from_ustring(r'123 2345 + += 2.34E5 "sada\"sa" ')
tokens = []
while True:
try:
x = f(buf)
except Exception as e:
print(e)
raise
print(x)
if x is None:
continue
if x.token_id == EOF_ID:
break
tokens.append(x)
print(tokens)
> python test.py
{ token_id = 2
lexeme = 123
line = 0
col = 3
span = 3
offset = 0
file = }
...
154 Jan 6, 2023
5 Feb 18, 2022
6 Jan 7, 2022
1 Nov 24, 2021
10 May 12, 2022
32 Jun 11, 2022
6 Aug 22, 2021
1 Dec 25, 2021
1 Nov 8, 2021
1 Nov 04, 2021
15 Dec 17, 2022
1 Sep 08, 2022
3 Oct 08, 2021
2 Jan 18, 2022
67 Jan 10, 2022
7 Nov 21, 2022
2 Oct 08, 2021
3 Sep 26, 2022
1 Jan 07, 2022
1 Dec 30, 2021
11 Dec 02, 2021
3 Dec 26, 2022
2 Mar 20, 2022
16 Dec 30, 2022
82 Aug 08, 2022
4 Aug 03, 2022
13 Nov 17, 2022
16 Nov 17, 2022