pdfrs 0.1.1

use regex::{Captures, Regex};

/// Returns true for characters in Unicode blocks that are reliably present
/// in common system fonts (e.g. Arial Unicode.ttf). Characters in the
/// Phonetic Extensions (U+1D00–U+1D7F) and Latin Extended-C (U+2C60–U+2C7F)
/// blocks are often missing, so we avoid them in subscript/superscript
/// conversion and fall back to ASCII bracket notation instead.
fn is_well_supported(ch: char) -> bool {
    let cp = ch as u32;
    !((0x1D00..=0x1D7F).contains(&cp) || (0x2C60..=0x2C7F).contains(&cp))
}

fn to_superscript(text: &str) -> Option<String> {
    let mut out = String::new();
    for ch in text.chars() {
        let mapped = match ch {
            '0' => '⁰', '1' => '¹', '2' => '²', '3' => '³', '4' => '⁴',
            '5' => '⁵', '6' => '⁶', '7' => '⁷', '8' => '⁸', '9' => '⁹',
            '+' => '⁺', '-' => '⁻', '=' => '⁼', '(' => '⁽', ')' => '⁾',
            'a' => 'ᵃ', 'b' => 'ᵇ', 'c' => 'ᶜ', 'd' => 'ᵈ', 'e' => 'ᵉ',
            'f' => 'ᶠ', 'g' => 'ᵍ', 'h' => 'ʰ', 'i' => 'ⁱ', 'j' => 'ʲ',
            'k' => 'ᵏ', 'l' => 'ˡ', 'm' => 'ᵐ', 'n' => 'ⁿ', 'o' => 'ᵒ',
            'p' => 'ᵖ', 'r' => 'ʳ', 's' => 'ˢ', 't' => 'ᵗ', 'u' => 'ᵘ',
            'v' => 'ᵛ', 'w' => 'ʷ', 'x' => 'ˣ', 'y' => 'ʸ', 'z' => 'ᶻ',
            'A' => 'ᴬ', 'B' => 'ᴮ', 'D' => 'ᴰ', 'E' => 'ᴱ', 'G' => 'ᴳ',
            'H' => 'ᴴ', 'I' => 'ᴵ', 'J' => 'ᴶ', 'K' => 'ᴷ', 'L' => 'ᴸ',
            'M' => 'ᴹ', 'N' => 'ᴺ', 'O' => 'ᴼ', 'P' => 'ᴾ', 'R' => 'ᴿ',
            'T' => 'ᵀ', 'U' => 'ᵁ', 'V' => 'ⱽ', 'W' => 'ᵂ',
            _ => return None,
        };
        if !is_well_supported(mapped) {
            return None;
        }
        out.push(mapped);
    }
    Some(out)
}

fn to_subscript(text: &str) -> Option<String> {
    let mut out = String::new();
    for ch in text.chars() {
        let mapped = match ch {
            '0' => '₀', '1' => '₁', '2' => '₂', '3' => '₃', '4' => '₄',
            '5' => '₅', '6' => '₆', '7' => '₇', '8' => '₈', '9' => '₉',
            '+' => '₊', '-' => '₋', '=' => '₌', '(' => '₍', ')' => '₎',
            'a' => 'ₐ', 'e' => 'ₑ', 'h' => 'ₕ', 'i' => 'ᵢ', 'j' => 'ⱼ',
            'k' => 'ₖ', 'l' => 'ₗ', 'm' => 'ₘ', 'n' => 'ₙ', 'o' => 'ₒ',
            'p' => 'ₚ', 'r' => 'ᵣ', 's' => 'ₛ', 't' => 'ₜ', 'u' => 'ᵤ',
            'v' => 'ᵥ', 'x' => 'ₓ',
            _ => return None,
        };
        if !is_well_supported(mapped) {
            return None;
        }
        out.push(mapped);
    }
    Some(out)
}

fn render_operator_with_limits(symbol: &str, lower: &str, upper: &str) -> String {
    match (to_subscript(lower), to_superscript(upper)) {
        (Some(lo), Some(up)) => format!("{}{}{}", symbol, lo, up),
        _ => format!("{}[{}→{}]", symbol, lower, upper),
    }
}

/// Convert LaTeX-like math notation to readable text for PDF rendering.
/// Since Type1 fonts don't support full LaTeX glyph rendering, we convert
/// common math commands to their text/symbol equivalents.
pub(super) fn render_math_text(expr: &str) -> String {
    let mut s = expr.to_string();

    // Greek letters
    let greek = [
        ("\\alpha", "\u{03B1}"), ("\\beta", "\u{03B2}"), ("\\gamma", "\u{03B3}"),
        ("\\delta", "\u{03B4}"), ("\\epsilon", "\u{03B5}"), ("\\zeta", "\u{03B6}"),
        ("\\eta", "\u{03B7}"), ("\\theta", "\u{03B8}"), ("\\iota", "\u{03B9}"),
        ("\\kappa", "\u{03BA}"), ("\\lambda", "\u{03BB}"), ("\\mu", "\u{03BC}"),
        ("\\nu", "\u{03BD}"), ("\\xi", "\u{03BE}"), ("\\pi", "\u{03C0}"),
        ("\\rho", "\u{03C1}"), ("\\sigma", "\u{03C3}"), ("\\tau", "\u{03C4}"),
        ("\\upsilon", "\u{03C5}"), ("\\phi", "\u{03C6}"), ("\\chi", "\u{03C7}"),
        ("\\psi", "\u{03C8}"), ("\\omega", "\u{03C9}"),
        ("\\Alpha", "A"), ("\\Beta", "B"), ("\\Gamma", "\u{0393}"),
        ("\\Delta", "\u{0394}"), ("\\Theta", "\u{0398}"), ("\\Lambda", "\u{039B}"),
        ("\\Xi", "\u{039E}"), ("\\Pi", "\u{03A0}"), ("\\Sigma", "\u{03A3}"),
        ("\\Phi", "\u{03A6}"), ("\\Psi", "\u{03A8}"), ("\\Omega", "\u{03A9}"),
    ];

    // Math operators and symbols
    let operators = [
        ("\\infty", "∞"), ("\\infinity", "∞"),
        ("\\pm", "±"), ("\\mp", "∓"),
        ("\\times", "×"), ("\\cdot", "·"),
        ("\\div", "÷"), ("\\neq", "≠"), ("\\ne", "≠"),
        ("\\leq", "≤"), ("\\le", "≤"),
        ("\\geq", "≥"), ("\\ge", "≥"),
        ("\\approx", "≈"), ("\\sim", "∼"),
        ("\\equiv", "≡"), ("\\propto", "∝"),
        ("\\rightarrow", "→"), ("\\leftarrow", "←"),
        ("\\to", "→"),
        ("\\Rightarrow", "⇒"), ("\\Leftarrow", "⇐"),
        ("\\leftrightarrow", "↔"),
        ("\\forall", "∀"), ("\\exists", "∃"),
        ("\\notin", "∉"), ("\\in", "∈"),
        ("\\subseteq", "⊆"), ("\\supseteq", "⊇"),
        ("\\subsetneq", "⊊"), ("\\supsetneq", "⊋"),
        ("\\subset", "⊂"), ("\\supset", "⊃"),
        ("\\cup", "∪"), ("\\cap", "∩"),
        ("\\wedge", "∧"), ("\\land", "∧"),
        ("\\vee", "∨"), ("\\lor", "∨"),
        ("\\neg", "¬"), ("\\lnot", "¬"),
        ("\\iff", "⇔"), ("\\implies", "⇒"),
        ("\\therefore", "∴"), ("\\because", "∵"),
        ("\\emptyset", "∅"),
        ("\\nabla", "∇"), ("\\partial", "∂"),
        ("\\ldots", "..."), ("\\cdots", "..."), ("\\dots", "..."),
        ("\\quad", "  "), ("\\qquad", "    "),
        ("\\,", " "), ("\\;", " "), ("\\!", ""),
        ("\\left", ""), ("\\right", ""),
        ("\\big", ""), ("\\Big", ""), ("\\bigg", ""), ("\\Bigg", ""),
    ];

    // Apply Greek letter replacements (longer patterns first to avoid partial matches)
    for (cmd, replacement) in &greek {
        s = s.replace(cmd, replacement);
    }

    // Common LaTeX variants and blackboard-bold sets
    s = s.replace("\\not\\in", "∉");
    s = s.replace("\\mathbb{R}", "ℝ");
    s = s.replace("\\mathbb{N}", "ℕ");
    s = s.replace("\\mathbb{Z}", "ℤ");
    s = s.replace("\\mathbb{Q}", "ℚ");
    s = s.replace("\\mathbb{C}", "ℂ");
    s = s.replace("\\mathbb{P}", "ℙ");
    s = s.replace("\\mathbb{H}", "ℍ");

    // Handle \frac{a}{b} -> (a)/(b)
    let frac_re = Regex::new(r"\\frac\{([^}]*)\}\{([^}]*)\}").unwrap();
    while frac_re.is_match(&s) {
        s = frac_re.replace_all(&s, "($1)/($2)").to_string();
    }

    // Handle \sqrt[n]{x} -> √[n](x) (do this first, before simple sqrt)
    let nroot_re = Regex::new(r"\\sqrt\[([^\]]*)\]\{([^}]*)\}").unwrap();
    while nroot_re.is_match(&s) {
        s = nroot_re.replace_all(&s, "√[$1]($2)").to_string();
    }

    // Handle \sqrt{x} -> √(x) (apply iteratively for nested sqrt)
    let sqrt_re = Regex::new(r"\\sqrt\{([^}]*)\}").unwrap();
    while sqrt_re.is_match(&s) {
        s = sqrt_re.replace_all(&s, "√($1)").to_string();
    }

    // Handle \sum, \prod, \int with limits
    let sum_re = Regex::new(r"\\sum_\{([^}]*)\}\^\{([^}]*)\}").unwrap();
    s = sum_re.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∑", &caps[1], &caps[2])
    }).to_string();
    let sum_re_simple = Regex::new(r"\\sum_([^\s\^_{}]+)\^([^\s\^_{}]+)").unwrap();
    s = sum_re_simple.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∑", &caps[1], &caps[2])
    }).to_string();
    s = s.replace("\\sum", "∑");

    let prod_re = Regex::new(r"\\prod_\{([^}]*)\}\^\{([^}]*)\}").unwrap();
    s = prod_re.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∏", &caps[1], &caps[2])
    }).to_string();
    let prod_re_simple = Regex::new(r"\\prod_([^\s\^_{}]+)\^([^\s\^_{}]+)").unwrap();
    s = prod_re_simple.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∏", &caps[1], &caps[2])
    }).to_string();
    s = s.replace("\\prod", "∏");

    let int_re = Regex::new(r"\\int_\{([^}]*)\}\^\{([^}]*)\}").unwrap();
    s = int_re.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∫", &caps[1], &caps[2])
    }).to_string();
    let int_re_simple = Regex::new(r"\\int_([^\s\^_{}]+)\^([^\s\^_{}]+)").unwrap();
    s = int_re_simple.replace_all(&s, |caps: &Captures| {
        render_operator_with_limits("∫", &caps[1], &caps[2])
    }).to_string();
    s = s.replace("\\int", "∫");

    let lim_re = Regex::new(r"\\lim_\{([^}]*)\}").unwrap();
    s = lim_re.replace_all(&s, "lim($1)").to_string();
    let lim_re_simple = Regex::new(r"\\lim_([^\s\^_{}]+)").unwrap();
    s = lim_re_simple.replace_all(&s, "lim($1)").to_string();
    s = s.replace("\\lim", "lim");

    // Handle superscript ^{x} -> ˣ and subscript _{x} -> ₓ (fallback to ASCII markers)
    let sup_re = Regex::new(r"\^\{([^}]*)\}").unwrap();
    s = sup_re
        .replace_all(&s, |caps: &Captures| {
            to_superscript(&caps[1]).unwrap_or_else(|| format!("^({})", &caps[1]))
        })
        .to_string();
    let sup_simple_re = Regex::new(r"\^([A-Za-z0-9+\-*/=])").unwrap();
    s = sup_simple_re
        .replace_all(&s, |caps: &Captures| {
            to_superscript(&caps[1]).unwrap_or_else(|| format!("^({})", &caps[1]))
        })
        .to_string();

    let sub_re = Regex::new(r"_\{([^}]*)\}").unwrap();
    s = sub_re
        .replace_all(&s, |caps: &Captures| {
            to_subscript(&caps[1]).unwrap_or_else(|| format!("_({})", &caps[1]))
        })
        .to_string();
    let sub_simple_re = Regex::new(r"_([A-Za-z0-9])").unwrap();
    s = sub_simple_re
        .replace_all(&s, |caps: &Captures| {
            to_subscript(&caps[1]).unwrap_or_else(|| format!("_({})", &caps[1]))
        })
        .to_string();

    // Handle \text{...} -> ...
    let text_re = Regex::new(r"\\text\{([^}]*)\}").unwrap();
    s = text_re.replace_all(&s, "$1").to_string();

    // Handle \mathbf{...}, \mathrm{...}, \mathit{...} -> content
    let mathfmt_re = Regex::new(r"\\math[a-z]+\{([^}]*)\}").unwrap();
    s = mathfmt_re.replace_all(&s, "$1").to_string();

    // Handle \hat{x}, \bar{x}, \vec{x}, \tilde{x}
    let hat_re = Regex::new(r"\\hat\{([^}]*)\}").unwrap();
    s = hat_re.replace_all(&s, "$1^").to_string();
    let bar_re = Regex::new(r"\\bar\{([^}]*)\}").unwrap();
    s = bar_re.replace_all(&s, "$1_bar").to_string();
    let vec_re = Regex::new(r"\\vec\{([^}]*)\}").unwrap();
    s = vec_re.replace_all(&s, "vec($1)").to_string();

    // Handle \log, \ln, \sin, \cos, \tan, \exp
    for func in &[
        "log", "ln", "sin", "cos", "tan", "cot", "sec", "csc",
        "sinh", "cosh", "tanh", "exp", "min", "max", "det", "dim",
    ] {
        let cmd = format!("\\{}", func);
        s = s.replace(&cmd, func);
    }

    // Apply generic operator replacements late to avoid prefix collisions
    // such as \in replacing the start of \int.
    for (cmd, replacement) in &operators {
        s = s.replace(cmd, replacement);
    }

    // Strip remaining braces
    s = s.replace('{', "").replace('}', "");

    // Clean up multiple spaces
    let multi_space = Regex::new(r"  +").unwrap();
    s = multi_space.replace_all(&s, " ").to_string();

    s.trim().to_string()
}

/// Escape a PDF string literal (for ASCII-only text)
pub(super) fn escape_pdf_string(text: &str) -> String {
    text.replace('\\', "\\\\")
        .replace('(', "\\(")
        .replace(')', "\\)")
        .replace('\r', "\\r")
        .replace('\n', "\\n")
        .replace('\t', "\\t")
}

/// Normalize text for Base-14 fonts (Helvetica/Courier family).
///
/// This is a compatibility fallback mode that can be enabled when a viewer
/// cannot render UTF-16 text with Base-14 fonts.
fn normalize_for_base14_font(text: &str) -> String {
    let mut out = String::new();

    for ch in text.chars() {
        match ch {
            // Fast path
            c if c.is_ascii() => out.push(c),

            // Math operators / relations
            '∞' => out.push_str("infinity"),
            '∑' => out.push_str("sum"),
            '∏' => out.push_str("prod"),
            '∫' => out.push_str("int"),
            '∂' => out.push_str("partial"),
            '∇' => out.push_str("nabla"),
            '√' => out.push_str("sqrt"),
            '≈' => out.push_str("~="),
            '≠' => out.push_str("!="),
            '≤' => out.push_str("<="),
            '≥' => out.push_str(">="),
            '±' => out.push_str("+/-"),
            '×' => out.push('*'),
            '÷' => out.push('/'),
            '∈' => out.push_str(" in "),
            '∉' => out.push_str(" not-in "),
            '∩' => out.push_str(" cap "),
            '∪' => out.push_str(" cup "),
            '⊂' => out.push_str(" subset "),
            '⊃' => out.push_str(" superset "),
            '⊆' => out.push_str(" subseteq "),
            '⊇' => out.push_str(" superseteq "),
            '⊊' => out.push_str(" subsetneq "),
            '⊋' => out.push_str(" supersetneq "),
            '∀' => out.push_str("forall"),
            '∃' => out.push_str("exists"),
            '∧' => out.push_str(" and "),
            '∨' => out.push_str(" or "),
            '¬' => out.push_str(" not "),
            '∴' => out.push_str(" therefore "),
            '∵' => out.push_str(" because "),

            // Greek letters commonly produced by render_math_text
            'α' => out.push_str("alpha"),
            'β' => out.push_str("beta"),
            'γ' => out.push_str("gamma"),
            'δ' => out.push_str("delta"),
            'ε' => out.push_str("epsilon"),
            'θ' => out.push_str("theta"),
            'λ' => out.push_str("lambda"),
            'μ' => out.push_str("mu"),
            'π' => out.push_str("pi"),
            'σ' => out.push_str("sigma"),
            'φ' => out.push_str("phi"),
            'ω' => out.push_str("omega"),
            'Γ' => out.push_str("Gamma"),
            'Δ' => out.push_str("Delta"),
            'Θ' => out.push_str("Theta"),
            'Λ' => out.push_str("Lambda"),
            'Π' => out.push_str("Pi"),
            'Σ' => out.push_str("Sigma"),
            'Φ' => out.push_str("Phi"),
            'Ω' => out.push_str("Omega"),
            'ℝ' => out.push('R'),
            'ℕ' => out.push('N'),
            'ℤ' => out.push('Z'),
            'ℚ' => out.push('Q'),
            'ℂ' => out.push('C'),
            'ℙ' => out.push('P'),
            'ℍ' => out.push('H'),

            // Currency
            '€' => out.push_str("EUR"),
            '£' => out.push_str("GBP"),
            '¥' => out.push_str("JPY"),
            '₹' => out.push_str("INR"),
            '₽' => out.push_str("RUB"),
            '₩' => out.push_str("KRW"),
            '₿' => out.push_str("BTC"),

            // Arrows
            '←' => out.push_str("<-"),
            '→' => out.push_str("->"),
            '↔' => out.push_str("<->"),
            '⇐' => out.push_str("<="),
            '⇒' => out.push_str("=>"),
            '⇔' => out.push_str("<=>"),

            // Superscripts / subscripts seen in examples
            '²' => out.push_str("^2"),
            '³' => out.push_str("^3"),
            '₀' => out.push_str("_0"),
            '₁' => out.push_str("_1"),
            '₂' => out.push_str("_2"),
            '₃' => out.push_str("_3"),
            '₄' => out.push_str("_4"),
            '₅' => out.push_str("_5"),
            '₆' => out.push_str("_6"),
            '₇' => out.push_str("_7"),
            '₈' => out.push_str("_8"),
            '₉' => out.push_str("_9"),

            // Fallback: keep visibility instead of rendering blank glyphs
            other => out.push_str(&format!("[U+{:04X}]", other as u32)),
        }
    }

    out
}

pub(super) fn use_base14_normalization() -> bool {
    std::env::var("PDFRS_BASE14_NORMALIZE")
        .map(|v| v == "1" || v.eq_ignore_ascii_case("true"))
        .unwrap_or(false)
}

/// Encode text for PDF - uses UTF-16BE hex encoding for unicode, literal string for ASCII
pub(super) fn encode_pdf_text(text: &str) -> String {
    let normalized = if use_base14_normalization() {
        normalize_for_base14_font(text)
    } else {
        text.to_string()
    };

    // Check if text contains any non-ASCII characters
    let has_unicode = normalized.chars().any(|c| !c.is_ascii());

    if !has_unicode {
        // Pure ASCII - use literal string format
        format!("({})", escape_pdf_string(&normalized))
    } else {
        // Contains unicode - use UTF-16BE hex encoding with BOM
        let mut utf16be_bytes = Vec::new();

        // Add BOM (Big Endian)
        utf16be_bytes.push(0xFE);
        utf16be_bytes.push(0xFF);

        // Encode each character as UTF-16BE
        for c in normalized.chars() {
            let mut code = c as u32;
            if code < 0x10000 {
                // BMP character - single UTF-16 code unit
                utf16be_bytes.push((code >> 8) as u8);
                utf16be_bytes.push((code & 0xFF) as u8);
            } else {
                // Surrogate pair for characters beyond BMP
                code -= 0x10000;
                let high_surrogate = 0xD800 + ((code >> 10) & 0x3FF);
                let low_surrogate = 0xDC00 + (code & 0x3FF);
                utf16be_bytes.push((high_surrogate >> 8) as u8);
                utf16be_bytes.push((high_surrogate & 0xFF) as u8);
                utf16be_bytes.push((low_surrogate >> 8) as u8);
                utf16be_bytes.push((low_surrogate & 0xFF) as u8);
            }
        }

        // Format as hex string
        let hex_string: String = utf16be_bytes
            .iter()
            .map(|b| format!("{:02X}", b))
            .collect();

        format!("<{}>", hex_string)
    }
}