morethantext/src/document/field.rs

use chrono::prelude::*;
use isolang::Language;
use std::{
    cmp::Ordering,
    collections::HashMap,
    ops::{Add, AddAssign},
    str::Split,
    time::Duration,
};
use uuid::Uuid;

#[derive(Clone, Debug, Eq, Hash, Ord, PartialEq)]
pub enum Field {
    Boolean(bool),
    DateTime(DateTime<Utc>),
    Duration(Duration),
    Integer(i128),
    None,
    StaticString(String),
    Uuid(Uuid),
}

impl Field {
    pub fn get_type(&self) -> FieldType {
        self.into()
    }
}

impl Add for Field {
    type Output = Self;

    fn add(self, other: Self) -> Self {
        match self {
            Field::DateTime(value1) => match other {
                Field::Duration(value2) => { value1 + value2 }.into(),
                _ => Field::None,
            },
            Field::Duration(value1) => match other {
                Field::Duration(value2) => { value1 + value2 }.into(),
                _ => Field::None,
            },
            Field::Integer(value1) => match other {
                Field::Integer(value2) => { value1 + value2 }.into(),
                _ => Field::None,
            },
            _ => Self::None,
        }
    }
}

impl AddAssign for Field {
    fn add_assign(&mut self, other: Self) {
        *self = self.clone().add(other);
    }
}

impl From<bool> for Field {
    fn from(value: bool) -> Self {
        Self::Boolean(value)
    }
}

impl From<DateTime<Utc>> for Field {
    fn from(value: DateTime<Utc>) -> Self {
        Self::DateTime(value)
    }
}

impl From<Duration> for Field {
    fn from(value: Duration) -> Self {
        Self::Duration(value)
    }
}

impl From<String> for Field {
    fn from(value: String) -> Self {
        Self::StaticString(value)
    }
}

impl From<&str> for Field {
    fn from(value: &str) -> Self {
        Self::from(value.to_string())
    }
}

impl From<Uuid> for Field {
    fn from(value: Uuid) -> Self {
        Self::Uuid(value)
    }
}

impl From<i128> for Field {
    fn from(value: i128) -> Self {
        Self::Integer(value)
    }
}

impl From<isize> for Field {
    fn from(value: isize) -> Self {
        let data: i128 = value.try_into().unwrap();
        Self::from(data)
    }
}

impl From<i32> for Field {
    fn from(value: i32) -> Self {
        let data: i128 = value.into();
        Self::from(data)
    }
}

impl PartialOrd for Field {
    fn partial_cmp(&self, other: &Field) -> Option<Ordering> {
        match (self, other) {
            (Self::Boolean(d1), Self::Boolean(d2)) => d1.partial_cmp(d2),
            (Self::DateTime(d1), Self::DateTime(d2)) => d1.partial_cmp(d2),
            (Self::Duration(d1), Self::Duration(d2)) => d1.partial_cmp(d2),
            (Self::Integer(d1), Self::Integer(d2)) => d1.partial_cmp(d2),
            (Self::StaticString(d1), Self::StaticString(d2)) => d1.partial_cmp(d2),
            (Self::Uuid(d1), Self::Uuid(d2)) => d1.partial_cmp(d2),
            (_, _) => None,
        }
    }
}

#[cfg(test)]
mod fields {
    use super::*;
    use rand::random;

    #[test]
    fn can_create_static_string() {
        let data = Uuid::new_v4().to_string();
        let result: Field = data.clone().into();
        match result.clone() {
            Field::StaticString(output) => assert_eq!(output, data),
            _ => unreachable!("got {:?}: should have been static string", result),
        }
        assert_eq!(result.get_type(), FieldType::StaticString);
    }

    #[test]
    fn can_create_from_str() {
        let holder = ["one", "two"];
        for data in holder.into_iter() {
            let result: Field = data.into();
            match result.clone() {
                Field::StaticString(output) => assert_eq!(output, data),
                _ => unreachable!("got {:?}: should have been static string", result),
            }
            assert_eq!(result.get_type(), FieldType::StaticString);
        }
    }

    #[test]
    fn create_from_uuid() {
        let data = Uuid::new_v4();
        let result: Field = data.clone().into();
        match result.clone() {
            Field::Uuid(output) => assert_eq!(output, data),
            _ => unreachable!("got {:?}: should have been uuid", result),
        }
        assert_eq!(result.get_type(), FieldType::Uuid);
    }

    #[test]
    fn create_from_datatime() {
        let data = Utc::now();
        let result: Field = data.into();
        match result.clone() {
            Field::DateTime(output) => assert_eq!(output, data),
            _ => unreachable!("got {:?}: should have been uuid", result),
        }
        assert_eq!(result.get_type(), FieldType::DateTime);
    }

    #[test]
    fn does_adding_return_none_for_things_that_cannot_add() {
        let value1: Field = Uuid::new_v4().into();
        let value2: Field = Uuid::new_v4().into();
        assert_eq!(value1 + value2, Field::None);
    }

    #[test]
    fn can_integers_be_added() {
        let value1: i128 = random::<u8>().into();
        let value2: i128 = random::<u8>().into();
        let expected: Field = { value1 + value2 }.into();
        let value1: Field = value1.into();
        let value2: Field = value2.into();
        assert_eq!(value1 + value2, expected);
    }

    #[test]
    fn can_integer_add_mismatch_returns_none() {
        let value1: Field = 5.into();
        let value2: Field = "nope".into();
        assert_eq!(value1 + value2, Field::None);
    }

    #[test]
    fn can_durations_be_added() {
        let data1: u64 = random::<u8>().into();
        let data2: u64 = random::<u8>().into();
        let value1: Field = Duration::from_secs(data1).into();
        let value2: Field = Duration::from_secs(data2).into();
        let expected: Field = Duration::from_secs(data1 + data2).into();
        assert_eq!(value1 + value2, expected);
    }

    #[test]
    fn does_duration_mismatch_return_none() {
        let value1: Field = Duration::from_secs(20).into();
        let value2: Field = "nope".into();
        assert_eq!(value1 + value2, Field::None);
    }

    #[test]
    fn can_durations_be_added_to_datetimes() {
        let timestamp = Utc::now();
        let data: u64 = random::<u8>().into();
        let duration = Duration::from_secs(data);
        let expected: Field = { timestamp + duration }.into();
        let value1: Field = timestamp.into();
        let value2: Field = duration.into();
        assert_eq!(value1 + value2, expected);
    }

    #[test]
    fn does_datetime_mismatch_return_none() {
        let value1: Field = Utc::now().into();
        let value2: Field = "nope".into();
        assert_eq!(value1 + value2, Field::None);
    }

    #[test]
    fn do_comparason_functions_work() {
        let fields = [Field::Integer(0), Field::Integer(1), Field::Integer(2)];
        assert!(fields[1] == fields[1], "equal did not work");
        assert!(fields[1] < fields[2], "less than did not work");
        assert!(fields[1] <= fields[2], "less than equal to did not work");
        assert!(fields[1] <= fields[1], "less tahn equal 50 did not work");
        assert!(fields[1] > fields[0], "greater than did not work");
        assert!(fields[1] >= fields[0], "greater than equal to did not work");
        assert!(fields[1] >= fields[1], "greater than equal to did not work");
    }

    #[test]
    fn does_mismatched_comparason_fields_return_false() {
        let fields = [Field::Integer(0), Field::Uuid(Uuid::nil())];
        assert!(!(fields[0] == fields[1]), "equal did not work");
        assert!(!(fields[0] < fields[1]), "less than did not work");
        assert!(!(fields[0] <= fields[1]), "less than equal to did not work");
        assert!(!(fields[0] <= fields[1]), "less tahn equal 50 did not work");
        assert!(!(fields[1] > fields[0]), "greater than did not work");
        assert!(
            !(fields[1] >= fields[0]),
            "greater than equal to did not work"
        );
        assert!(
            !(fields[1] >= fields[0]),
            "greater than equal to did not work"
        );
    }
}

#[derive(Clone, Debug, PartialEq)]
pub enum FieldType {
    Boolean,
    DateTime,
    Duration,
    Integer,
    None,
    StaticString,
    Uuid,
}

impl FieldType {
    pub fn get_default(&self) -> Field {
        match self {
            FieldType::Boolean => false.into(),
            FieldType::DateTime => Utc::now().into(),
            FieldType::Duration => Duration::from_secs(0).into(),
            FieldType::Integer => 0.into(),
            FieldType::None => Field::None,
            FieldType::StaticString => "".into(),
            FieldType::Uuid => Uuid::new_v4().into(),
        }
    }
}

impl From<&Field> for FieldType {
    fn from(value: &Field) -> Self {
        match value {
            Field::Boolean(_) => Self::Boolean,
            Field::DateTime(_) => Self::DateTime,
            Field::Duration(_) => Self::Duration,
            Field::Integer(_) => Self::Integer,
            Field::None => Self::None,
            Field::StaticString(_) => Self::StaticString,
            Field::Uuid(_) => Self::Uuid,
        }
    }
}

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

    #[test]
    fn can_get_defaults_for_uuid() {
        let ftype = FieldType::Uuid;
        let mut ids: Vec<Uuid> = Vec::new();
        for _ in 0..5 {
            let result = ftype.get_default();
            match result {
                Field::Uuid(data) => {
                    assert!(
                        !ids.contains(&data),
                        "found duplicate id {:?} in {:?}",
                        data,
                        ids
                    );
                    ids.push(data.clone());
                }
                _ => unreachable!("got {:?}: should have been uuid", result),
            }
        }
    }

    #[test]
    fn can_get_defaults_for_static_string() {
        let ftype = FieldType::StaticString;
        let result = ftype.get_default();
        match result {
            Field::StaticString(data) => assert_eq!(data, ""),
            _ => unreachable!("got {:?}: should have been static string", result),
        }
    }
}

#[derive(Clone, Debug)]
enum DataOrigin {
    Human,
    Computer,
}

#[derive(Clone, Debug)]
struct ParagraphData {
    string: String,
    origin: DataOrigin,
    initial: bool,
}

impl ParagraphData {
    fn initial(string: String) -> Self {
        Self {
            string: string,
            origin: DataOrigin::Human,
            initial: true,
        }
    }

    fn new(string: String, origin: DataOrigin) -> Self {
        Self {
            string: string,
            origin: origin,
            initial: false,
        }
    }

    fn get_text(&self) -> &String {
        &self.string
    }

    fn by_human(&self) -> bool {
        match self.origin {
            DataOrigin::Human => true,
            DataOrigin::Computer => false,
        }
    }

    fn is_initial(&self) -> bool {
        self.initial
    }
}

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

    #[test]
    fn can_determine_initial_information() {
        let text = Uuid::new_v4().to_string();
        let data = ParagraphData::initial(text.clone());
        assert_eq!(data.get_text(), &text);
        assert!(data.by_human(), "{:?} should have returned true", data);
        assert!(data.is_initial(), "{:?} should have returned true", data);
    }

    #[test]
    fn can_be_made_by_humans() {
        let text = Uuid::new_v4().to_string();
        let data = ParagraphData::new(text.clone(), DataOrigin::Human);
        assert_eq!(data.get_text(), &text);
        assert!(data.by_human(), "{:?} should have returned true", data);
        assert!(!data.is_initial(), "{:?} should have returned false", data);
    }

    #[test]
    fn can_be_made_by_computers() {
        let text = Uuid::new_v4().to_string();
        let data = ParagraphData::new(text.clone(), DataOrigin::Computer);
        assert_eq!(data.get_text(), &text);
        assert!(!data.by_human(), "{:?} should have returned false", data);
        assert!(!data.is_initial(), "{:?} should have returned false", data);
    }
}

#[derive(Clone, Debug)]
struct Paragraph {
    data: HashMap<Language, ParagraphData>,
}

impl Paragraph {
    fn new(lang: Language, string: String) -> Self {
        let mut data = HashMap::new();
        data.insert(lang, ParagraphData::initial(string));
        Self { data: data }
    }

    fn add_translation(&mut self, string: String, lang: Language) {
        match self.data.get(&lang) {
            Some(_) => {}
            None => {
                self.data
                    .insert(lang, ParagraphData::new(string, DataOrigin::Computer));
            }
        };
    }

    fn improve_translation(&mut self, string: String, lang: Language) {
        self.data
            .insert(lang, ParagraphData::new(string, DataOrigin::Human));
    }

    fn get(&self, lang: &Language) -> Option<&String> {
        match self.data.get(lang) {
            Some(data) => Some(data.get_text()),
            None => None,
        }
    }

    fn get_initial(&self) -> (&Language, &String) {
        for (lang, data) in self.data.iter() {
            if data.is_initial() {
                return (lang, data.get_text());
            }
        }
        unreachable!("paragraph should initialize with data");
    }

    fn by_humans(&self) -> HashMap<Language, String> {
        let mut output = HashMap::new();
        for (lang, data) in self.data.iter() {
            if data.by_human() {
                output.insert(lang.clone(), data.get_text().clone());
            }
        }
        output
    }
}

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

    #[test]
    fn does_paragraph_store_language_information() {
        let languages = [
            Language::from_639_1("en").unwrap(),
            Language::from_639_1("ja").unwrap(),
        ];
        let data = Uuid::new_v4().to_string();
        for lang in languages.iter() {
            let result = Paragraph::new(lang.clone(), data.clone());
            assert_eq!(result.get(lang).unwrap(), &data);
        }
    }

    #[test]
    fn are_multiple_languages_stored() {
        let text = ["test", "テスト"];
        let languages = [
            Language::from_639_1("en").unwrap(),
            Language::from_639_1("ja").unwrap(),
        ];
        let mut paragraph = Paragraph::new(languages[0].clone(), text[0].clone().to_string());
        paragraph.add_translation(text[1].clone().to_string(), languages[1].clone());
        for i in 0..text.len() {
            assert_eq!(paragraph.get(&languages[i]).unwrap(), text[i]);
        }
    }

    #[test]
    fn does_add_translation_get_ignored_if_it_already_exists() {
        let text = "something";
        let lang = Language::from_639_1("en").unwrap();
        let mut paragraph = Paragraph::new(lang.clone(), text.to_string());
        paragraph.add_translation("other".to_string(), lang);
        assert_eq!(paragraph.get(&lang).unwrap(), text);
    }

    #[test]
    fn does_improve_translation_replace_existing() {
        let text = "new";
        let lang = Language::from_639_1("en").unwrap();
        let mut paragraph = Paragraph::new(lang.clone(), "old".to_string());
        paragraph.improve_translation(text.to_string(), lang.clone());
        assert_eq!(paragraph.get(&lang).unwrap(), text);
    }

    #[test]
    fn can_determine_human_text() {
        let text = "something";
        let lang = Language::from_639_1("en").unwrap();
        let paragraph = Paragraph::new(lang.clone(), text.to_string());
        let result = paragraph.by_humans();
        assert_eq!(result.len(), 1, "got wrong numnber of texts");
        assert_eq!(result.get(&lang).unwrap(), text);
    }

    #[test]
    fn add_translation_does_not_count_as_human_text() {
        let text = "test";
        let lang = Language::from_639_1("en").unwrap();
        let mut paragraph = Paragraph::new(lang.clone(), text.to_string());
        paragraph.add_translation("テスト".to_string(), Language::from_639_1("ja").unwrap());
        let result = paragraph.by_humans();
        assert_eq!(result.len(), 1, "got wrong numnber of texts");
        assert_eq!(result.get(&lang).unwrap(), text);
    }

    #[test]
    fn impove_translation_does_get_added_as_human() {
        let text = ["test", "テスト"];
        let languages = [
            Language::from_639_1("en").unwrap(),
            Language::from_639_1("ja").unwrap(),
        ];
        let mut paragraph = Paragraph::new(languages[0].clone(), text[0].clone().to_string());
        paragraph.improve_translation(text[1].clone().to_string(), languages[1].clone());
        let result = paragraph.by_humans();
        assert_eq!(result.len(), 2, "got wrong numnber of texts");
        for i in 0..text.len() {
            assert_eq!(result.get(&languages[i]).unwrap(), text[i]);
        }
    }

    #[test]
    fn can_get_original_text() {
        let text = Uuid::nil().to_string();
        let lang = Language::from_639_1("en").unwrap();
        let mut paragraph = Paragraph::new(lang.clone(), text.clone());
        paragraph.add_translation(
            Uuid::new_v4().to_string(),
            Language::from_639_1("ja").unwrap(),
        );
        paragraph.improve_translation(
            Uuid::new_v4().to_string(),
            Language::from_639_1("de").unwrap(),
        );
        let (rlang, rtext) = paragraph.get_initial();
        assert_eq!(rlang, &lang);
        assert_eq!(rtext, &text);
    }
}

#[derive(Clone, Debug)]
struct UniversalString {
    paragraphs: HashMap<Uuid, Paragraph>,
    revisions: Vec<Vec<Uuid>>,
}

impl UniversalString {
    fn new(lang: Language, text: String) -> Self {
        let mut output = Self {
            paragraphs: HashMap::new(),
            revisions: Vec::new(),
        };
        output.update(lang, text);
        output
    }

    fn get(&self, lang: &Language) -> Option<String> {
        let latest = self.revisions.len() - 1;
        self.get_revision(latest, lang)
    }

    fn get_revision(&self, rev_num: usize, lang: &Language) -> Option<String> {
        let mut output = "".to_string();
        for id in self.revisions[rev_num].iter() {
            let paragraph = self.paragraphs.get(id).unwrap();
            let text = paragraph.get(lang).unwrap();
            output += text;
            output += "\u{2029}";
        }
        Some(output)
    }

    fn revision_count(&self) -> usize {
        self.revisions.len() - 1
    }

    fn update(&mut self, lang: Language, text: String) {
        let mut version = Vec::new();
        for paragraph in text.as_str().split("\u{2029}") {
            if paragraph != "" {
                let mut id = Uuid::nil();
                for (key, value) in self.paragraphs.iter() {
                    if &paragraph == value.get(&lang).unwrap() {
                        id = key.clone();
                        break;
                    }
                }
                if id == Uuid::nil() {
                    id = Uuid::new_v4();
                    while self.paragraphs.contains_key(&id) {
                        id = Uuid::new_v4();
                    }
                    self.paragraphs.insert(
                        id.clone(),
                        Paragraph::new(lang.clone(), paragraph.to_string()),
                    );
                }
                version.push(id);
            }
        }
        self.revisions.push(version);
    }
}

#[cfg(test)]
mod universal_strings {
    use super::*;
    use rand::random_range;

    const ENGLISH_DATA: [&str; 5] = ["one", "two", "three", "four", "five"];
    const JAPANESE_DATA: [&str; 5] = ["一", "二", "三", "四", "五"];

    struct TestData;

    impl TestData {
        fn english() -> (Language, Vec<String>) {
            let lang = Language::from_639_1("en").unwrap();
            let mut data = Vec::new();
            for text in ENGLISH_DATA.iter() {
                data.push(text.to_string());
            }
            (lang, data)
        }

        fn japanese() -> (Language, Vec<String>) {
            let lang = Language::from_639_1("ja").unwrap();
            let mut data = Vec::new();
            for text in JAPANESE_DATA.iter() {
                data.push(text.to_string());
            }
            (lang, data)
        }

        fn to_input(data: Vec<String>) -> String {
            let mut output = "".to_string();
            for paragraph in data.iter() {
                output += paragraph;
                output += "\u{2029}";
            }
            output
        }
    }

    #[test]
    fn are_initial_strings_empty() {
        let text = ["test", "テスト"];
        let languages = [
            Language::from_639_1("en").unwrap(),
            Language::from_639_1("ja").unwrap(),
        ];
        for i in 0..text.len() {
            let ustr = UniversalString::new(languages[i].clone(), text[i].to_string());
            let expected = text[i].to_string() + "\u{2029}";
            assert_eq!(ustr.get(&languages[i]).unwrap(), expected);
            assert_eq!(ustr.revision_count(), 0);
            assert_eq!(ustr.paragraphs.len(), 1);
        }
    }

    #[test]
    fn accepts_strings_with_multiple_paragraphs() {
        let (lang, data) = TestData::english();
        let input = TestData::to_input(data.clone());
        let ustr = UniversalString::new(lang.clone(), input.clone());
        assert_eq!(ustr.get(&lang).unwrap(), input);
        assert_eq!(ustr.revision_count(), 0);
        assert_eq!(ustr.paragraphs.len(), data.len(), "{:?}", ustr);
    }

    #[test]
    fn can_insert_text_into_string() {
        let (lang, mut data) = TestData::english();
        let initial = TestData::to_input(data.clone());
        let mut ustr = UniversalString::new(lang.clone(), initial.clone());
        let position = random_range(..data.len());
        data.insert(position, "something".to_string());
        let expected = TestData::to_input(data.clone());
        ustr.update(lang.clone(), expected.clone());
        assert_eq!(ustr.get(&lang).unwrap(), expected);
        assert_eq!(ustr.revision_count(), 1);
        assert_eq!(ustr.paragraphs.len(), data.len(), "{:?}", ustr);
        assert_eq!(ustr.get_revision(0, &lang).unwrap(), initial);
    }

    #[test]
    fn can_a_paragraph_be_replaced() {
        let (lang, mut data) = TestData::english();
        let initial = TestData::to_input(data.clone());
        let mut ustr = UniversalString::new(lang.clone(), initial.clone());
        let position = random_range(..data.len());
        data[position] = "replaced".to_string();
        let expected = TestData::to_input(data.clone());
        ustr.update(lang.clone(), expected.clone());
        assert_eq!(ustr.get(&lang).unwrap(), expected);
        assert_eq!(ustr.revision_count(), 1);
        assert_eq!(ustr.paragraphs.len(), (data.len() + 1), "{:?}", ustr);
        assert_eq!(ustr.get_revision(0, &lang).unwrap(), initial);
    }

    #[test]
    fn does_not_store_duplicate_data() {
        let lang = Language::from_639_1("en").unwrap();
        let mut data = Vec::new();
        for _ in 0..3 {
            data.push("same".to_string());
        }
        let initial = TestData::to_input(data.clone());
        let mut ustr = UniversalString::new(lang.clone(), initial.clone());
        assert_eq!(ustr.get(&lang).unwrap(), initial);
        assert_eq!(ustr.paragraphs.len(), 1, "{:?}", ustr);
    }

    #[test]
    fn can_text_be_removed() {
        let (lang, mut data) = TestData::english();
        let expected_paragraphs = data.len();
        let initial = TestData::to_input(data.clone());
        let mut ustr = UniversalString::new(lang.clone(), initial.clone());
        let position = random_range(..data.len());
        data.remove(position);
        let expected = TestData::to_input(data.clone());
        ustr.update(lang.clone(), expected.clone());
        assert_eq!(ustr.get(&lang).unwrap(), expected);
        assert_eq!(ustr.revision_count(), 1);
        assert_eq!(ustr.paragraphs.len(), expected_paragraphs, "{:?}", ustr);
        assert_eq!(ustr.get_revision(0, &lang).unwrap(), initial);
    }
}