Encode character by character with error handling.
(
code_page: u32,
s: &PyStrRef,
errors: &str,
encoding_name: &str,
vm: &VirtualMachine,
)
| 870 | |
| 871 | /// Encode character by character with error handling. |
| 872 | fn encode_code_page_errors( |
| 873 | code_page: u32, |
| 874 | s: &PyStrRef, |
| 875 | errors: &str, |
| 876 | encoding_name: &str, |
| 877 | vm: &VirtualMachine, |
| 878 | ) -> PyResult<(Vec<u8>, usize)> { |
| 879 | use crate::builtins::{PyBytes, PyStr, PyTuple}; |
| 880 | use windows_sys::Win32::Globalization::WideCharToMultiByte; |
| 881 | |
| 882 | let char_len = s.char_len(); |
| 883 | let flags = encode_code_page_flags(code_page, errors); |
| 884 | let use_default_char = code_page != 65001 && code_page != 65000; |
| 885 | let encoding_str = vm.ctx.new_str(encoding_name); |
| 886 | let reason_str = vm.ctx.new_str("invalid character"); |
| 887 | |
| 888 | // For strict mode, find the first unencodable character and raise |
| 889 | if errors == "strict" { |
| 890 | // Find the failing position by trying each character |
| 891 | let mut fail_pos = 0; |
| 892 | for cp in s.as_wtf8().code_points() { |
| 893 | let ch = cp.to_u32(); |
| 894 | if (0xD800..=0xDFFF).contains(&ch) { |
| 895 | break; |
| 896 | } |
| 897 | let mut wchars = [0u16; 2]; |
| 898 | let wchar_len = if ch < 0x10000 { |
| 899 | wchars[0] = ch as u16; |
| 900 | 1 |
| 901 | } else { |
| 902 | wchars[0] = ((ch - 0x10000) >> 10) as u16 + 0xD800; |
| 903 | wchars[1] = ((ch - 0x10000) & 0x3FF) as u16 + 0xDC00; |
| 904 | 2 |
| 905 | }; |
| 906 | let mut used_default_char: i32 = 0; |
| 907 | let pused = if use_default_char { |
| 908 | &mut used_default_char as *mut i32 |
| 909 | } else { |
| 910 | core::ptr::null_mut() |
| 911 | }; |
| 912 | let outsize = unsafe { |
| 913 | WideCharToMultiByte( |
| 914 | code_page, |
| 915 | flags, |
| 916 | wchars.as_ptr(), |
| 917 | wchar_len, |
| 918 | core::ptr::null_mut(), |
| 919 | 0, |
| 920 | core::ptr::null(), |
| 921 | pused, |
| 922 | ) |
| 923 | }; |
| 924 | if outsize <= 0 || (use_default_char && used_default_char != 0) { |
| 925 | break; |
| 926 | } |
| 927 | fail_pos += 1; |
| 928 | } |
| 929 | return Err(vm.new_unicode_encode_error_real( |
no test coverage detected