use std::char;
use std::fmt;
use std::mem::size_of;
use std::ops::Range;
use std::string::ToString;

use crossbeam_utils::atomic::AtomicCell;
use itertools::Itertools;
use num_traits::ToPrimitive;
use unic_ucd_bidi::BidiClass;
use unic_ucd_category::GeneralCategory;
use unic_ucd_ident::{is_xid_continue, is_xid_start};
use unicode_casing::CharExt;

use super::objbytes::{PyBytes, PyBytesRef};
use super::objdict::PyDict;
use super::objint::{PyInt, PyIntRef};
use super::objiter;
use super::objsequence::{PySliceableSequence, SequenceIndex};
use super::objtype::{self, PyClassRef};
use crate::exceptions::IntoPyException;
use crate::format::{FormatSpec, FormatString, FromTemplate};
use crate::function::{OptionalArg, OptionalOption, PyFuncArgs};
use crate::pyobject::{
    BorrowValue, IdProtocol, IntoPyObject, ItemProtocol, PyClassImpl, PyContext, PyIterable,
    PyObjectRef, PyRef, PyResult, PyValue, TryFromObject, TryIntoRef, TypeProtocol,
};
use crate::pystr::{
    self, adjust_indices, PyCommonString, PyCommonStringContainer, PyCommonStringWrapper,
};
use crate::vm::VirtualMachine;
use rustpython_common::hash;

/// str(object='') -> str
/// str(bytes_or_buffer[, encoding[, errors]]) -> str
///
/// Create a new string object from the given object. If encoding or
/// errors is specified, then the object must expose a data buffer
/// that will be decoded using the given encoding and error handler.
/// Otherwise, returns the result of object.__str__() (if defined)
/// or repr(object).
/// encoding defaults to sys.getdefaultencoding().
/// errors defaults to 'strict'."
#[pyclass(module = false, name = "str")]
#[derive(Debug)]
pub struct PyString {
    value: String,
    hash: AtomicCell<Option<hash::PyHash>>,
    len: AtomicCell<Option<usize>>,
}

impl<'a> BorrowValue<'a> for PyString {
    type Borrowed = &'a str;

    fn borrow_value(&'a self) -> Self::Borrowed {
        &self.value
    }
}

impl AsRef<str> for PyString {
    fn as_ref(&self) -> &str {
        &self.value
    }
}

impl<T> From<&T> for PyString
where
    T: AsRef<str> + ?Sized,
{
    fn from(s: &T) -> PyString {
        s.as_ref().to_owned().into()
    }
}

impl From<String> for PyString {
    fn from(s: String) -> PyString {
        PyString {
            value: s,
            hash: AtomicCell::default(),
            len: AtomicCell::default(),
        }
    }
}

pub type PyStringRef = PyRef<PyString>;

impl fmt::Display for PyString {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        fmt::Display::fmt(&self.value, f)
    }
}

impl TryIntoRef<PyString> for String {
    fn try_into_ref(self, vm: &VirtualMachine) -> PyResult<PyRef<PyString>> {
        Ok(PyString::from(self).into_ref(vm))
    }
}

impl TryIntoRef<PyString> for &str {
    fn try_into_ref(self, vm: &VirtualMachine) -> PyResult<PyRef<PyString>> {
        Ok(PyString::from(self).into_ref(vm))
    }
}

#[pyclass(module = false, name = "str_iterator")]
#[derive(Debug)]
pub struct PyStringIterator {
    pub string: PyStringRef,
    position: AtomicCell<usize>,
}

impl PyValue for PyStringIterator {
    fn class(vm: &VirtualMachine) -> PyClassRef {
        vm.ctx.types.str_iterator_type.clone()
    }
}

#[pyimpl]
impl PyStringIterator {
    #[pymethod(name = "__next__")]
    fn next(&self, vm: &VirtualMachine) -> PyResult<String> {
        // TODO: use something more performant than chars().nth() that's still atomic
        let pos = self.position.fetch_add(1);

        if let Some(c) = self.string.value.chars().nth(pos) {
            Ok(c.to_string())
        } else {
            Err(objiter::new_stop_iteration(vm))
        }
    }

    #[pymethod(name = "__iter__")]
    fn iter(zelf: PyRef<Self>) -> PyRef<Self> {
        zelf
    }
}

#[pyclass(module = false, name = "str_reverseiterator")]
#[derive(Debug)]
pub struct PyStringReverseIterator {
    pub position: AtomicCell<isize>,
    pub string: PyStringRef,
}

impl PyValue for PyStringReverseIterator {
    fn class(vm: &VirtualMachine) -> PyClassRef {
        vm.ctx.types.str_reverseiterator_type.clone()
    }
}

#[pyimpl]
impl PyStringReverseIterator {
    #[pymethod(name = "__next__")]
    fn next(&self, vm: &VirtualMachine) -> PyResult<String> {
        let pos = self.position.fetch_sub(1);
        if pos > 0 {
            if let Some(c) = self.string.value.chars().nth(pos as usize - 1) {
                return Ok(c.to_string());
            }
        }
        Err(objiter::new_stop_iteration(vm))
    }

    #[pymethod(name = "__iter__")]
    fn iter(zelf: PyRef<Self>) -> PyRef<Self> {
        zelf
    }
}

#[derive(FromArgs)]
struct StrArgs {
    #[pyarg(positional_or_keyword, optional = true)]
    object: OptionalArg<PyObjectRef>,
    #[pyarg(positional_or_keyword, optional = true)]
    encoding: OptionalArg<PyStringRef>,
    #[pyarg(positional_or_keyword, optional = true)]
    errors: OptionalArg<PyStringRef>,
}

#[pyimpl(flags(BASETYPE))]
impl PyString {
    #[pyslot]
    fn tp_new(cls: PyClassRef, args: StrArgs, vm: &VirtualMachine) -> PyResult<PyStringRef> {
        let string: PyStringRef = match args.object {
            OptionalArg::Present(input) => {
                if let OptionalArg::Present(enc) = args.encoding {
                    vm.decode(input, Some(enc.clone()), args.errors.into_option())?
                        .downcast()
                        .map_err(|obj| {
                            vm.new_type_error(format!(
                                "'{}' decoder returned '{}' instead of 'str'; use codecs.encode() to \
                                 encode arbitrary types",
                                enc,
                                obj.lease_class().name,
                            ))
                        })?
                } else {
                    vm.to_str(&input)?
                }
            }
            OptionalArg::Missing => {
                PyString::from(String::new()).into_ref_with_type(vm, cls.clone())?
            }
        };
        if string.class().is(&cls) {
            Ok(string)
        } else {
            PyString::from(string.borrow_value()).into_ref_with_type(vm, cls)
        }
    }

    #[pymethod(name = "__add__")]
    fn add(&self, other: PyObjectRef, vm: &VirtualMachine) -> PyResult<String> {
        if objtype::isinstance(&other, &vm.ctx.types.str_type) {
            Ok(self.value.py_add(borrow_value(&other)))
        } else {
            Err(vm.new_type_error(format!("Cannot add {} and {}", self, other)))
        }
    }

    #[pymethod(name = "__bool__")]
    fn bool(&self) -> bool {
        !self.value.is_empty()
    }

    #[pymethod(name = "__eq__")]
    fn eq(&self, rhs: PyObjectRef, vm: &VirtualMachine) -> PyObjectRef {
        if objtype::isinstance(&rhs, &vm.ctx.types.str_type) {
            vm.ctx.new_bool(self.value == borrow_value(&rhs))
        } else {
            vm.ctx.not_implemented()
        }
    }

    #[pymethod(name = "__ne__")]
    fn ne(&self, rhs: PyObjectRef, vm: &VirtualMachine) -> PyObjectRef {
        if objtype::isinstance(&rhs, &vm.ctx.types.str_type) {
            vm.ctx.new_bool(self.value != borrow_value(&rhs))
        } else {
            vm.ctx.not_implemented()
        }
    }

    #[pymethod(name = "__contains__")]
    fn contains(&self, needle: PyStringRef) -> bool {
        self.value.contains(&needle.value)
    }

    #[pymethod(name = "__getitem__")]
    fn getitem(&self, needle: SequenceIndex, vm: &VirtualMachine) -> PyResult {
        match needle {
            SequenceIndex::Int(pos) => {
                let index: usize = if pos.is_negative() {
                    (self.value.chars().count() as isize + pos) as usize
                } else {
                    pos.abs() as usize
                };

                if let Some(character) = self.value.chars().nth(index) {
                    Ok(vm.ctx.new_str(character.to_string()))
                } else {
                    Err(vm.new_index_error("string index out of range".to_owned()))
                }
            }
            SequenceIndex::Slice(slice) => {
                let string = self.get_slice_items(vm, &slice)?;
                Ok(vm.ctx.new_str(string))
            }
        }
    }

    #[pymethod(name = "__gt__")]
    fn gt(&self, other: PyStringRef) -> bool {
        self.value > other.value
    }

    #[pymethod(name = "__ge__")]
    fn ge(&self, other: PyStringRef) -> bool {
        self.value >= other.value
    }

    #[pymethod(name = "__lt__")]
    fn lt(&self, other: PyStringRef) -> bool {
        self.value < other.value
    }

    #[pymethod(name = "__le__")]
    fn le(&self, other: PyStringRef) -> bool {
        self.value <= other.value
    }

    #[pymethod(name = "__hash__")]
    pub(crate) fn hash(&self) -> hash::PyHash {
        self.hash.load().unwrap_or_else(|| {
            let hash = hash::hash_str(&self.value);
            self.hash.store(Some(hash));
            hash
        })
    }

    #[pymethod(name = "__len__")]
    #[inline]
    fn len(&self) -> usize {
        self.len.load().unwrap_or_else(|| {
            let len = self.value.chars().count();
            self.len.store(Some(len));
            len
        })
    }

    #[pymethod(name = "__sizeof__")]
    fn sizeof(&self) -> usize {
        size_of::<Self>() + self.value.capacity() * size_of::<u8>()
    }

    #[pymethod(name = "__mul__")]
    #[pymethod(name = "__rmul__")]
    fn mul(&self, value: isize) -> String {
        self.value.repeat(value.to_usize().unwrap_or(0))
    }

    #[pymethod(name = "__str__")]
    fn str(zelf: PyRef<Self>) -> PyStringRef {
        zelf
    }

    #[pymethod(name = "__repr__")]
    pub(crate) fn repr(&self, vm: &VirtualMachine) -> PyResult<String> {
        let in_len = self.value.len();
        let mut out_len = 0usize;
        // let mut max = 127;
        let mut squote = 0;
        let mut dquote = 0;

        for ch in self.value.chars() {
            let incr = match ch {
                '\'' => {
                    squote += 1;
                    1
                }
                '"' => {
                    dquote += 1;
                    1
                }
                '\\' | '\t' | '\r' | '\n' => 2,
                ch if ch < ' ' || ch as u32 == 0x7f => 4, // \xHH
                ch if ch.is_ascii() => 1,
                ch if char_is_printable(ch) => {
                    // max = std::cmp::max(ch, max);
                    1
                }
                ch if (ch as u32) < 0x100 => 4,   // \xHH
                ch if (ch as u32) < 0x10000 => 6, // \uHHHH
                _ => 10,                          // \uHHHHHHHH
            };
            if out_len > (std::isize::MAX as usize) - incr {
                return Err(vm.new_overflow_error("string is too long to generate repr".to_owned()));
            }
            out_len += incr;
        }

        let (quote, unchanged) = {
            let mut quote = '\'';
            let mut unchanged = out_len == in_len;
            if squote > 0 {
                unchanged = false;
                if dquote > 0 {
                    // Both squote and dquote present. Use squote, and escape them
                    out_len += squote;
                } else {
                    quote = '"';
                }
            }
            (quote, unchanged)
        };

        out_len += 2; // quotes

        let mut repr = String::with_capacity(out_len);
        repr.push(quote);
        if unchanged {
            repr.push_str(&self.value);
        } else {
            for ch in self.value.chars() {
                if ch == quote || ch == '\\' {
                    repr.push('\\');
                    repr.push(ch);
                } else if ch == '\n' {
                    repr.push_str("\\n")
                } else if ch == '\t' {
                    repr.push_str("\\t");
                } else if ch == '\r' {
                    repr.push_str("\\r");
                } else if ch < ' ' || ch as u32 == 0x7F {
                    repr.push_str(&format!("\\x{:02x}", ch as u32));
                } else if ch.is_ascii() {
                    repr.push(ch);
                } else if !char_is_printable(ch) {
                    let code = ch as u32;
                    let escaped = if code < 0xff {
                        format!("\\x{:02x}", code)
                    } else if code < 0xffff {
                        format!("\\u{:04x}", code)
                    } else {
                        format!("\\U{:08x}", code)
                    };
                    repr.push_str(&escaped);
                } else {
                    repr.push(ch)
                }
            }
        }

        repr.push(quote);
        Ok(repr)
    }

    #[pymethod]
    fn lower(&self) -> String {
        self.value.to_lowercase()
    }

    // casefold is much more aggressive than lower
    #[pymethod]
    fn casefold(&self) -> String {
        caseless::default_case_fold_str(&self.value)
    }

    #[pymethod]
    fn upper(&self) -> String {
        self.value.to_uppercase()
    }

    #[pymethod]
    fn capitalize(&self) -> String {
        let mut chars = self.value.chars();
        if let Some(first_char) = chars.next() {
            format!(
                "{}{}",
                first_char.to_uppercase(),
                &chars.as_str().to_lowercase(),
            )
        } else {
            "".to_owned()
        }
    }

    #[pymethod]
    fn split(&self, args: SplitArgs, vm: &VirtualMachine) -> PyResult {
        let elements = self.value.py_split(
            args,
            vm,
            |v, s, vm| v.split(s).map(|s| vm.ctx.new_str(s)).collect(),
            |v, s, n, vm| v.splitn(n, s).map(|s| vm.ctx.new_str(s)).collect(),
            |v, n, vm| v.py_split_whitespace(n, |s| vm.ctx.new_str(s)),
        )?;
        Ok(vm.ctx.new_list(elements))
    }

    #[pymethod]
    fn rsplit(&self, args: SplitArgs, vm: &VirtualMachine) -> PyResult {
        let mut elements = self.value.py_split(
            args,
            vm,
            |v, s, vm| v.rsplit(s).map(|s| vm.ctx.new_str(s)).collect(),
            |v, s, n, vm| v.rsplitn(n, s).map(|s| vm.ctx.new_str(s)).collect(),
            |v, n, vm| v.py_rsplit_whitespace(n, |s| vm.ctx.new_str(s)),
        )?;
        // Unlike Python rsplit, Rust rsplitn returns an iterator that
        // starts from the end of the string.
        elements.reverse();
        Ok(vm.ctx.new_list(elements))
    }

    #[pymethod]
    fn strip(&self, chars: OptionalOption<PyStringRef>) -> String {
        self.value
            .py_strip(
                chars,
                |s, chars| s.trim_matches(|c| chars.contains(c)),
                |s| s.trim(),
            )
            .to_owned()
    }

    #[pymethod]
    fn lstrip(&self, chars: OptionalOption<PyStringRef>) -> String {
        self.value
            .py_strip(
                chars,
                |s, chars| s.trim_start_matches(|c| chars.contains(c)),
                |s| s.trim_start(),
            )
            .to_owned()
    }

    #[pymethod]
    fn rstrip(&self, chars: OptionalOption<PyStringRef>) -> String {
        self.value
            .py_strip(
                chars,
                |s, chars| s.trim_end_matches(|c| chars.contains(c)),
                |s| s.trim_end(),
            )
            .to_owned()
    }

    #[pymethod]
    fn endswith(&self, args: pystr::StartsEndsWithArgs, vm: &VirtualMachine) -> PyResult<bool> {
        self.value.py_startsendswith(
            args,
            "endswith",
            "str",
            |s, x: &PyStringRef| s.ends_with(x.borrow_value()),
            vm,
        )
    }

    #[pymethod]
    fn startswith(&self, args: pystr::StartsEndsWithArgs, vm: &VirtualMachine) -> PyResult<bool> {
        self.value.py_startsendswith(
            args,
            "startswith",
            "str",
            |s, x: &PyStringRef| s.starts_with(x.borrow_value()),
            vm,
        )
    }

    /// removeprefix($self, prefix, /)
    ///
    ///
    /// Return a str with the given prefix string removed if present.
    ///
    /// If the string starts with the prefix string, return string[len(prefix):]
    /// Otherwise, return a copy of the original string.
    #[pymethod]
    fn removeprefix(&self, pref: PyStringRef) -> String {
        self.value
            .py_removeprefix(&pref.value, pref.value.len(), |s, p| s.starts_with(p))
            .to_owned()
    }

    /// removesuffix(self, prefix, /)
    ///
    ///
    /// Return a str with the given suffix string removed if present.
    ///
    /// If the string ends with the suffix string, return string[:len(suffix)]
    /// Otherwise, return a copy of the original string.
    #[pymethod]
    fn removesuffix(&self, suff: PyStringRef) -> String {
        self.value
            .py_removesuffix(&suff.value, suff.value.len(), |s, p| s.ends_with(p))
            .to_owned()
    }

    #[pymethod]
    fn isalnum(&self) -> bool {
        !self.value.is_empty() && self.value.chars().all(char::is_alphanumeric)
    }

    #[pymethod]
    fn isnumeric(&self) -> bool {
        !self.value.is_empty() && self.value.chars().all(char::is_numeric)
    }

    #[pymethod]
    fn isdigit(&self) -> bool {
        // python's isdigit also checks if exponents are digits, these are the unicodes for exponents
        let valid_unicodes: [u16; 10] = [
            0x2070, 0x00B9, 0x00B2, 0x00B3, 0x2074, 0x2075, 0x2076, 0x2077, 0x2078, 0x2079,
        ];

        !self.value.is_empty()
            && self
                .value
                .chars()
                .filter(|c| !c.is_digit(10))
                .all(|c| valid_unicodes.contains(&(c as u16)))
    }

    #[pymethod]
    fn isdecimal(&self) -> bool {
        !self.value.is_empty() && self.value.chars().all(|c| c.is_ascii_digit())
    }

    #[pymethod(name = "__mod__")]
    fn modulo(&self, values: PyObjectRef, vm: &VirtualMachine) -> PyResult<String> {
        let formatted = self.value.py_cformat(values, vm)?;
        Ok(formatted)
    }

    #[pymethod(name = "__rmod__")]
    fn rmod(&self, _values: PyObjectRef, vm: &VirtualMachine) -> PyResult {
        Ok(vm.ctx.not_implemented())
    }

    #[pymethod]
    fn format(&self, args: PyFuncArgs, vm: &VirtualMachine) -> PyResult<String> {
        match FormatString::from_str(&self.value) {
            Ok(format_string) => format_string.format(&args, vm),
            Err(err) => Err(err.into_pyexception(vm)),
        }
    }

    /// S.format_map(mapping) -> str
    ///
    /// Return a formatted version of S, using substitutions from mapping.
    /// The substitutions are identified by braces ('{' and '}').
    #[pymethod]
    fn format_map(&self, mapping: PyObjectRef, vm: &VirtualMachine) -> PyResult<String> {
        match FormatString::from_str(&self.value) {
            Ok(format_string) => format_string.format_map(&mapping, vm),
            Err(err) => Err(err.into_pyexception(vm)),
        }
    }

    #[pymethod(name = "__format__")]
    fn format_str(&self, spec: PyStringRef, vm: &VirtualMachine) -> PyResult<String> {
        match FormatSpec::parse(spec.borrow_value())
            .and_then(|format_spec| format_spec.format_string(&self.value))
        {
            Ok(string) => Ok(string),
            Err(err) => Err(vm.new_value_error(err.to_string())),
        }
    }

    /// Return a titlecased version of the string where words start with an
    /// uppercase character and the remaining characters are lowercase.
    #[pymethod]
    fn title(&self) -> String {
        let mut title = String::with_capacity(self.value.len());
        let mut previous_is_cased = false;
        for c in self.value.chars() {
            if c.is_lowercase() {
                if !previous_is_cased {
                    title.extend(c.to_titlecase());
                } else {
                    title.push(c);
                }
                previous_is_cased = true;
            } else if c.is_uppercase() || c.is_titlecase() {
                if previous_is_cased {
                    title.extend(c.to_lowercase());
                } else {
                    title.push(c);
                }
                previous_is_cased = true;
            } else {
                previous_is_cased = false;
                title.push(c);
            }
        }
        title
    }

    #[pymethod]
    fn swapcase(&self) -> String {
        let mut swapped_str = String::with_capacity(self.value.len());
        for c in self.value.chars() {
            // to_uppercase returns an iterator, to_ascii_uppercase returns the char
            if c.is_lowercase() {
                swapped_str.push(c.to_ascii_uppercase());
            } else if c.is_uppercase() {
                swapped_str.push(c.to_ascii_lowercase());
            } else {
                swapped_str.push(c);
            }
        }
        swapped_str
    }

    #[pymethod]
    fn isalpha(&self) -> bool {
        !self.value.is_empty() && self.value.chars().all(char::is_alphabetic)
    }

    #[pymethod]
    fn replace(&self, old: PyStringRef, new: PyStringRef, count: OptionalArg<isize>) -> String {
        match count {
            OptionalArg::Present(maxcount) if maxcount >= 0 => {
                if maxcount == 0 || self.value.is_empty() {
                    // nothing to do; return the original bytes
                    return self.value.clone();
                }
                self.value
                    .replacen(&old.value, &new.value, maxcount as usize)
            }
            _ => self.value.replace(&old.value, &new.value),
        }
    }

    /// Return true if all characters in the string are printable or the string is empty,
    /// false otherwise.  Nonprintable characters are those characters defined in the
    /// Unicode character database as `Other` or `Separator`,
    /// excepting the ASCII space (0x20) which is considered printable.
    ///
    /// All characters except those characters defined in the Unicode character
    /// database as following categories are considered printable.
    ///   * Cc (Other, Control)
    ///   * Cf (Other, Format)
    ///   * Cs (Other, Surrogate)
    ///   * Co (Other, Private Use)
    ///   * Cn (Other, Not Assigned)
    ///   * Zl Separator, Line ('\u2028', LINE SEPARATOR)
    ///   * Zp Separator, Paragraph ('\u2029', PARAGRAPH SEPARATOR)
    ///   * Zs (Separator, Space) other than ASCII space('\x20').
    #[pymethod]
    fn isprintable(&self) -> bool {
        self.value
            .chars()
            .all(|c| c == '\u{0020}' || char_is_printable(c))
    }

    #[pymethod]
    fn isspace(&self) -> bool {
        if self.value.is_empty() {
            return false;
        }
        use unic_ucd_bidi::bidi_class::abbr_names::*;
        self.value.chars().all(|c| {
            GeneralCategory::of(c) == GeneralCategory::SpaceSeparator
                || matches!(BidiClass::of(c), WS | B | S)
        })
    }

    // Return true if all cased characters in the string are lowercase and there is at least one cased character, false otherwise.
    #[pymethod]
    fn islower(&self) -> bool {
        self.value.py_iscase(char::is_lowercase, char::is_uppercase)
    }

    // Return true if all cased characters in the string are uppercase and there is at least one cased character, false otherwise.
    #[pymethod]
    fn isupper(&self) -> bool {
        self.value.py_iscase(char::is_uppercase, char::is_lowercase)
    }

    #[pymethod]
    fn isascii(&self) -> bool {
        self.value.is_ascii()
    }

    #[pymethod]
    fn splitlines(&self, args: pystr::SplitLinesArgs, vm: &VirtualMachine) -> PyObjectRef {
        vm.ctx
            .new_list(self.value.py_splitlines(args, |s| vm.ctx.new_str(s)))
    }

    #[pymethod]
    fn join(&self, iterable: PyIterable<PyStringRef>, vm: &VirtualMachine) -> PyResult<String> {
        let iter = iterable.iter(vm)?;
        self.value.py_join(iter)
    }

    #[inline]
    fn _find<F>(&self, args: FindArgs, find: F) -> Option<usize>
    where
        F: Fn(&str, &str) -> Option<usize>,
    {
        let (sub, range) = args.get_value(self.len());
        self.value.py_find(&sub.value, range, find)
    }

    #[pymethod]
    fn find(&self, args: FindArgs) -> isize {
        self._find(args, |r, s| r.find(s))
            .map_or(-1, |v| v as isize)
    }

    #[pymethod]
    fn rfind(&self, args: FindArgs) -> isize {
        self._find(args, |r, s| r.rfind(s))
            .map_or(-1, |v| v as isize)
    }

    #[pymethod]
    fn index(&self, args: FindArgs, vm: &VirtualMachine) -> PyResult<usize> {
        self._find(args, |r, s| r.find(s))
            .ok_or_else(|| vm.new_value_error("substring not found".to_owned()))
    }

    #[pymethod]
    fn rindex(&self, args: FindArgs, vm: &VirtualMachine) -> PyResult<usize> {
        self._find(args, |r, s| r.rfind(s))
            .ok_or_else(|| vm.new_value_error("substring not found".to_owned()))
    }

    #[pymethod]
    fn partition(&self, sep: PyStringRef, vm: &VirtualMachine) -> PyResult {
        let (front, has_mid, back) =
            self.value
                .py_partition(&sep.value, || self.value.splitn(2, &sep.value), vm)?;
        Ok(vm.ctx.new_tuple(vec![
            vm.ctx.new_str(front),
            if has_mid {
                sep.into_object()
            } else {
                vm.ctx.new_str("")
            },
            vm.ctx.new_str(back),
        ]))
    }

    #[pymethod]
    fn rpartition(&self, sep: PyStringRef, vm: &VirtualMachine) -> PyResult {
        let (back, has_mid, front) =
            self.value
                .py_partition(&sep.value, || self.value.rsplitn(2, &sep.value), vm)?;
        Ok(vm.ctx.new_tuple(vec![
            vm.ctx.new_str(front),
            if has_mid {
                sep.into_object()
            } else {
                vm.ctx.new_str("")
            },
            vm.ctx.new_str(back),
        ]))
    }

    /// Return `true` if the sequence is ASCII titlecase and the sequence is not
    /// empty, `false` otherwise.
    #[pymethod]
    fn istitle(&self) -> bool {
        if self.value.is_empty() {
            return false;
        }

        let mut cased = false;
        let mut previous_is_cased = false;
        for c in self.value.chars() {
            if c.is_uppercase() || c.is_titlecase() {
                if previous_is_cased {
                    return false;
                }
                previous_is_cased = true;
                cased = true;
            } else if c.is_lowercase() {
                if !previous_is_cased {
                    return false;
                }
                previous_is_cased = true;
                cased = true;
            } else {
                previous_is_cased = false;
            }
        }
        cased
    }

    #[pymethod]
    fn count(&self, args: FindArgs) -> usize {
        let (needle, range) = args.get_value(self.len());
        self.value
            .py_count(&needle.value, range, |h, n| h.matches(n).count())
    }

    #[pymethod]
    fn zfill(&self, width: isize) -> String {
        // this is safe-guaranteed because the original self.value is valid utf8
        unsafe { String::from_utf8_unchecked(self.value.py_zfill(width)) }
    }

    #[inline]
    fn _pad(
        &self,
        width: isize,
        fillchar: OptionalArg<PyStringRef>,
        pad: fn(&str, usize, char, usize) -> String,
        vm: &VirtualMachine,
    ) -> PyResult<String> {
        let fillchar = match fillchar {
            OptionalArg::Present(ref s) => s.value.chars().exactly_one().map_err(|_| {
                vm.new_type_error(
                    "The fill character must be exactly one character long".to_owned(),
                )
            }),
            OptionalArg::Missing => Ok(' '),
        }?;
        Ok(if self.len() as isize >= width {
            String::from(&self.value)
        } else {
            pad(&self.value, width as usize, fillchar, self.len())
        })
    }

    #[pymethod]
    fn center(
        &self,
        width: isize,
        fillchar: OptionalArg<PyStringRef>,
        vm: &VirtualMachine,
    ) -> PyResult<String> {
        self._pad(width, fillchar, PyCommonString::<char>::py_center, vm)
    }

    #[pymethod]
    fn ljust(
        &self,
        width: isize,
        fillchar: OptionalArg<PyStringRef>,
        vm: &VirtualMachine,
    ) -> PyResult<String> {
        self._pad(width, fillchar, PyCommonString::<char>::py_ljust, vm)
    }

    #[pymethod]
    fn rjust(
        &self,
        width: isize,
        fillchar: OptionalArg<PyStringRef>,
        vm: &VirtualMachine,
    ) -> PyResult<String> {
        self._pad(width, fillchar, PyCommonString::<char>::py_rjust, vm)
    }

    #[pymethod]
    fn expandtabs(&self, args: pystr::ExpandTabsArgs) -> String {
        let tab_stop = args.tabsize();
        let mut expanded_str = String::with_capacity(self.value.len());
        let mut tab_size = tab_stop;
        let mut col_count = 0 as usize;
        for ch in self.value.chars() {
            match ch {
                '\t' => {
                    let num_spaces = tab_size - col_count;
                    col_count += num_spaces;
                    let expand = " ".repeat(num_spaces);
                    expanded_str.push_str(&expand);
                }
                '\r' | '\n' => {
                    expanded_str.push(ch);
                    col_count = 0;
                    tab_size = 0;
                }
                _ => {
                    expanded_str.push(ch);
                    col_count += 1;
                }
            }
            if col_count >= tab_size {
                tab_size += tab_stop;
            }
        }
        expanded_str
    }

    #[pymethod]
    fn isidentifier(&self) -> bool {
        let mut chars = self.value.chars();
        let is_identifier_start = chars.next().map_or(false, |c| c == '_' || is_xid_start(c));
        // a string is not an identifier if it has whitespace or starts with a number
        is_identifier_start && chars.all(is_xid_continue)
    }

    // https://docs.python.org/3/library/stdtypes.html#str.translate
    #[pymethod]
    fn translate(&self, table: PyObjectRef, vm: &VirtualMachine) -> PyResult<String> {
        vm.get_method_or_type_error(table.clone(), "__getitem__", || {
            format!("'{}' object is not subscriptable", table.lease_class().name)
        })?;

        let mut translated = String::new();
        for c in self.value.chars() {
            match table.get_item((c as u32).into_pyobject(vm), vm) {
                Ok(value) => {
                    if let Some(text) = value.payload::<PyString>() {
                        translated.push_str(&text.value);
                    } else if let Some(bigint) = value.payload::<PyInt>() {
                        let ch = bigint
                            .borrow_value()
                            .to_u32()
                            .and_then(std::char::from_u32)
                            .ok_or_else(|| {
                                vm.new_value_error(
                                    "character mapping must be in range(0x110000)".to_owned(),
                                )
                            })?;
                        translated.push(ch as char);
                    } else if !vm.is_none(&value) {
                        return Err(vm.new_type_error(
                            "character mapping must return integer, None or str".to_owned(),
                        ));
                    }
                }
                _ => translated.push(c),
            }
        }
        Ok(translated)
    }

    #[pymethod]
    fn maketrans(
        dict_or_str: PyObjectRef,
        to_str: OptionalArg<PyStringRef>,
        none_str: OptionalArg<PyStringRef>,
        vm: &VirtualMachine,
    ) -> PyResult {
        let new_dict = vm.ctx.new_dict();
        if let OptionalArg::Present(to_str) = to_str {
            match dict_or_str.downcast::<PyString>() {
                Ok(from_str) => {
                    if to_str.len() == from_str.len() {
                        for (c1, c2) in from_str.value.chars().zip(to_str.value.chars()) {
                            new_dict.set_item(
                                vm.ctx.new_int(c1 as u32),
                                vm.ctx.new_int(c2 as u32),
                                vm,
                            )?;
                        }
                        if let OptionalArg::Present(none_str) = none_str {
                            for c in none_str.value.chars() {
                                new_dict.set_item(vm.ctx.new_int(c as u32), vm.get_none(), vm)?;
                            }
                        }
                        Ok(new_dict.into_pyobject(vm))
                    } else {
                        Err(vm.new_value_error(
                            "the first two maketrans arguments must have equal length".to_owned(),
                        ))
                    }
                }
                _ => Err(vm.new_type_error(
                    "first maketrans argument must be a string if there is a second argument"
                        .to_owned(),
                )),
            }
        } else {
            // dict_str must be a dict
            match dict_or_str.downcast::<PyDict>() {
                Ok(dict) => {
                    for (key, val) in dict {
                        if let Some(num) = key.payload::<PyInt>() {
                            new_dict.set_item(
                                num.borrow_value().to_i32().into_pyobject(vm),
                                val,
                                vm,
                            )?;
                        } else if let Some(string) = key.payload::<PyString>() {
                            if string.len() == 1 {
                                let num_value = string.value.chars().next().unwrap() as u32;
                                new_dict.set_item(num_value.into_pyobject(vm), val, vm)?;
                            } else {
                                return Err(vm.new_value_error(
                                    "string keys in translate table must be of length 1".to_owned(),
                                ));
                            }
                        }
                    }
                    Ok(new_dict.into_pyobject(vm))
                }
                _ => Err(vm.new_value_error(
                    "if you give only one argument to maketrans it must be a dict".to_owned(),
                )),
            }
        }
    }

    #[pymethod]
    fn encode(zelf: PyRef<Self>, args: EncodeArgs, vm: &VirtualMachine) -> PyResult<PyBytesRef> {
        encode_string(zelf, args.encoding, args.errors, vm)
    }

    #[pymethod(magic)]
    fn iter(zelf: PyRef<Self>) -> PyStringIterator {
        PyStringIterator {
            position: AtomicCell::new(0),
            string: zelf,
        }
    }

    #[pymethod(magic)]
    fn reversed(zelf: PyRef<Self>) -> PyStringReverseIterator {
        let begin = zelf.value.chars().count();

        PyStringReverseIterator {
            position: AtomicCell::new(begin as isize),
            string: zelf,
        }
    }
}

#[derive(FromArgs)]
struct EncodeArgs {
    #[pyarg(positional_or_keyword, default = "None")]
    encoding: Option<PyStringRef>,
    #[pyarg(positional_or_keyword, default = "None")]
    errors: Option<PyStringRef>,
}

pub(crate) fn encode_string(
    s: PyStringRef,
    encoding: Option<PyStringRef>,
    errors: Option<PyStringRef>,
    vm: &VirtualMachine,
) -> PyResult<PyBytesRef> {
    vm.encode(s.into_object(), encoding.clone(), errors)?
        .downcast::<PyBytes>()
        .map_err(|obj| {
            vm.new_type_error(format!(
                "'{}' encoder returned '{}' instead of 'bytes'; use codecs.encode() to \
                 encode arbitrary types",
                encoding.as_ref().map_or("utf-8", |s| s.borrow_value()),
                obj.lease_class().name,
            ))
        })
}

impl PyValue for PyString {
    fn class(vm: &VirtualMachine) -> PyClassRef {
        vm.ctx.types.str_type.clone()
    }
}

impl IntoPyObject for String {
    fn into_pyobject(self, vm: &VirtualMachine) -> PyObjectRef {
        vm.ctx.new_str(self)
    }
}

impl IntoPyObject for &str {
    fn into_pyobject(self, vm: &VirtualMachine) -> PyObjectRef {
        vm.ctx.new_str(self)
    }
}

impl IntoPyObject for &String {
    fn into_pyobject(self, vm: &VirtualMachine) -> PyObjectRef {
        vm.ctx.new_str(self.clone())
    }
}

impl TryFromObject for std::ffi::CString {
    fn try_from_object(vm: &VirtualMachine, obj: PyObjectRef) -> PyResult<Self> {
        let s = PyStringRef::try_from_object(vm, obj)?;
        Self::new(s.borrow_value().to_owned())
            .map_err(|_| vm.new_value_error("embedded null character".to_owned()))
    }
}

impl TryFromObject for std::ffi::OsString {
    fn try_from_object(vm: &VirtualMachine, obj: PyObjectRef) -> PyResult<Self> {
        use std::str::FromStr;

        let s = PyStringRef::try_from_object(vm, obj)?;
        Ok(std::ffi::OsString::from_str(s.borrow_value()).unwrap())
    }
}

type SplitArgs<'a> = pystr::SplitArgs<'a, PyStringRef, str, char>;

#[derive(FromArgs)]
pub struct FindArgs {
    #[pyarg(positional_only, optional = false)]
    sub: PyStringRef,
    #[pyarg(positional_only, default = "None")]
    start: Option<PyIntRef>,
    #[pyarg(positional_only, default = "None")]
    end: Option<PyIntRef>,
}

impl FindArgs {
    fn get_value(self, len: usize) -> (PyStringRef, std::ops::Range<usize>) {
        let range = adjust_indices(self.start, self.end, len);
        (self.sub, range)
    }
}

pub fn init(ctx: &PyContext) {
    PyString::extend_class(ctx, &ctx.types.str_type);

    PyStringIterator::extend_class(ctx, &ctx.types.str_iterator_type);
    PyStringReverseIterator::extend_class(ctx, &ctx.types.str_reverseiterator_type);
}

pub fn clone_value(obj: &PyObjectRef) -> String {
    obj.payload::<PyString>().unwrap().value.clone()
}

pub fn borrow_value(obj: &PyObjectRef) -> &str {
    &obj.payload::<PyString>().unwrap().value
}

impl PySliceableSequence for PyString {
    type Sliced = String;

    fn do_slice(&self, range: Range<usize>) -> Self::Sliced {
        self.value
            .chars()
            .skip(range.start)
            .take(range.end - range.start)
            .collect()
    }

    fn do_slice_reverse(&self, range: Range<usize>) -> Self::Sliced {
        let count = self.len();
        self.value
            .chars()
            .rev()
            .skip(count - range.end)
            .take(range.end - range.start)
            .collect()
    }

    fn do_stepped_slice(&self, range: Range<usize>, step: usize) -> Self::Sliced {
        self.value
            .chars()
            .skip(range.start)
            .take(range.end - range.start)
            .step_by(step)
            .collect()
    }

    fn do_stepped_slice_reverse(&self, range: Range<usize>, step: usize) -> Self::Sliced {
        let count = self.len();
        self.value
            .chars()
            .rev()
            .skip(count - range.end)
            .take(range.end - range.start)
            .step_by(step)
            .collect()
    }

    fn empty() -> Self::Sliced {
        String::new()
    }

    fn len(&self) -> usize {
        self.len()
    }

    fn is_empty(&self) -> bool {
        self.value.is_empty()
    }
}

// According to python following categories aren't printable:
// * Cc (Other, Control)
// * Cf (Other, Format)
// * Cs (Other, Surrogate)
// * Co (Other, Private Use)
// * Cn (Other, Not Assigned)
// * Zl Separator, Line ('\u2028', LINE SEPARATOR)
// * Zp Separator, Paragraph ('\u2029', PARAGRAPH SEPARATOR)
// * Zs (Separator, Space) other than ASCII space('\x20').
fn char_is_printable(c: char) -> bool {
    let cat = GeneralCategory::of(c);
    !(cat.is_other() || cat.is_separator())
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn str_title() {
        let tests = vec![
            (" Hello ", " hello "),
            ("Hello ", "hello "),
            ("Hello ", "Hello "),
            ("Format This As Title String", "fOrMaT thIs aS titLe String"),
            ("Format,This-As*Title;String", "fOrMaT,thIs-aS*titLe;String"),
            ("Getint", "getInt"),
            ("Greek Ωppercases ...", "greek ωppercases ..."),
            ("Greek ῼitlecases ...", "greek ῳitlecases ..."),
        ];
        for (title, input) in tests {
            assert_eq!(PyString::from(input).title().as_str(), title);
        }
    }

    #[test]
    fn str_istitle() {
        let pos = vec![
            "A",
            "A Titlecased Line",
            "A\nTitlecased Line",
            "A Titlecased, Line",
            "Greek Ωppercases ...",
            "Greek ῼitlecases ...",
        ];

        for s in pos {
            assert!(PyString::from(s).istitle());
        }

        let neg = vec![
            "",
            "a",
            "\n",
            "Not a capitalized String",
            "Not\ta Titlecase String",
            "Not--a Titlecase String",
            "NOT",
        ];
        for s in neg {
            assert!(!PyString::from(s).istitle());
        }
    }

    #[test]
    fn str_maketrans_and_translate() {
        let vm: VirtualMachine = Default::default();

        let table = vm.ctx.new_dict();
        table.set_item("a", vm.ctx.new_str("🎅"), &vm).unwrap();
        table.set_item("b", vm.get_none(), &vm).unwrap();
        table.set_item("c", vm.ctx.new_str("xda"), &vm).unwrap();
        let translated = PyString::maketrans(
            table.into_object(),
            OptionalArg::Missing,
            OptionalArg::Missing,
            &vm,
        )
        .unwrap();
        let text = PyString::from("abc");
        let translated = text.translate(translated, &vm).unwrap();
        assert_eq!(translated, "🎅xda".to_owned());
        let translated = text.translate(vm.ctx.new_int(3), &vm);
        assert_eq!(
            translated.unwrap_err().lease_class().name,
            "TypeError".to_owned()
        );
    }
}

impl PyCommonStringWrapper<str> for PyStringRef {
    fn as_ref(&self) -> &str {
        self.value.as_str()
    }
}

impl PyCommonStringContainer<str> for String {
    fn new() -> Self {
        String::new()
    }

    fn with_capacity(capacity: usize) -> Self {
        String::with_capacity(capacity)
    }

    fn push_str(&mut self, other: &str) {
        String::push_str(self, other)
    }
}

impl<'s> PyCommonString<'s, char> for str {
    type Container = String;
    type CharIter = std::str::Chars<'s>;
    type ElementIter = std::str::Chars<'s>;

    fn element_bytes_len(c: char) -> usize {
        c.len_utf8()
    }

    fn to_container(&self) -> Self::Container {
        self.to_owned()
    }

    fn as_bytes(&self) -> &[u8] {
        self.as_bytes()
    }

    fn as_utf8_str(&self) -> Result<&str, std::str::Utf8Error> {
        Ok(self)
    }

    fn chars(&'s self) -> Self::CharIter {
        str::chars(self)
    }

    fn elements(&'s self) -> Self::ElementIter {
        str::chars(self)
    }

    fn get_bytes<'a>(&'a self, range: std::ops::Range<usize>) -> &'a Self {
        &self[range]
    }

    fn get_chars<'a>(&'a self, range: std::ops::Range<usize>) -> &'a Self {
        rustpython_common::str::get_chars(self, range)
    }

    fn is_empty(&self) -> bool {
        Self::is_empty(self)
    }

    fn bytes_len(&self) -> usize {
        Self::len(self)
    }

    fn py_split_whitespace<F>(&self, maxsplit: isize, convert: F) -> Vec<PyObjectRef>
    where
        F: Fn(&Self) -> PyObjectRef,
    {
        // CPython split_whitespace
        let mut splited = Vec::new();
        let mut last_offset = 0;
        let mut count = maxsplit;
        for (offset, _) in self.match_indices(|c: char| c.is_ascii_whitespace() || c == '\x0b') {
            if last_offset == offset {
                last_offset += 1;
                continue;
            }
            if count == 0 {
                break;
            }
            splited.push(convert(&self[last_offset..offset]));
            last_offset = offset + 1;
            count -= 1;
        }
        if last_offset != self.len() {
            splited.push(convert(&self[last_offset..]));
        }
        splited
    }

    fn py_rsplit_whitespace<F>(&self, maxsplit: isize, convert: F) -> Vec<PyObjectRef>
    where
        F: Fn(&Self) -> PyObjectRef,
    {
        // CPython rsplit_whitespace
        let mut splited = Vec::new();
        let mut last_offset = self.len();
        let mut count = maxsplit;
        for (offset, _) in self.rmatch_indices(|c: char| c.is_ascii_whitespace() || c == '\x0b') {
            if last_offset == offset + 1 {
                last_offset -= 1;
                continue;
            }
            if count == 0 {
                break;
            }
            splited.push(convert(&self[offset + 1..last_offset]));
            last_offset = offset;
            count -= 1;
        }
        if last_offset != 0 {
            splited.push(convert(&self[..last_offset]));
        }
        splited
    }
}