Re: [PATCH v19 0/5] MR2091: Draft: windows.media.speech: Initial Unixlib work.

[Jinoh Kang (＠iamahuman)]

Jinoh Kang (@iamahuman) commented about dlls/windows.media.speech/unixlib.c:
> - while ((dirent = readdir(dir)))
> - {
> -
>    if (dirent->d_type != DT_DIR)
> -
>        continue;
> -
> -
>    dir_name = dirent->d_name;
> -
> -
>    if (!strncmp(dir_name, vosk_model_identifier_small, ident_small_len))
> -
>        dir_name += ident_small_len;
> -
>    else if (!strncmp(dir_name, vosk_model_identifier, ident_len))
> -
>        dir_name += ident_len;
> -
>    else
> -
>        continue;
> -
> -
>    /* First match for lang and region (en-us), then only lang (en). */
> -
>    if (strncmp(dir_name, locale, len) && strncmp(dir_name, locale, str - locale))

If the `en-gb` model comes before the `en-us` model in the directory, `en-gb` will match the lang part (`en`) and `find_model_by_locale_and_path` will proceed with it.

Note that the `strncmp(dir_name, locale, len) && ` part can be omitted since lang-region match _implies_ lang match, and lang mismatch _implies_ lang-region mismatch.  Do you mind explaining how it worked for your case?