badness 0.5.0

A language server, formatter, and linter for LaTeX
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221

//! Shared projector for the BibTeX differential parse oracle (`bib_parse_oracle.rs`,
//! `bib_parse_compat.rs`).
//!
//! Like the LaTeX `parse_skeleton`, this projects both badness's and texlab's bib CSTs
//! onto one coarse, common skeleton and measures structural concordance as a triage
//! signal, never a hard tree-equality gate. AGENTS.md: measure against, never match.
//!
//! Two facts shape the bib oracle differently from the LaTeX one:
//!
//! 1. **texlab's BibTeX parser has no error channel.** Unlike its LaTeX parser (which
//!    emits `ERROR` nodes), the bib parser is fully lenient — a missing `}` / `=` / key
//!    is silently skipped via `expect`, and there is no `ERROR` kind at all. So the
//!    LaTeX-style "texlab must not error" soundness gate is vacuous here. The hard gate
//!    instead checks an **entry-recognition floor** (see [`count_entries`]).
//! 2. **The two parsers genuinely disagree inside values.** texlab models accents
//!    (`ACCENT`), commands (`COMMAND`), and `~` (`NBSP`) inside braces/quotes; badness
//!    keeps value internals generic. That divergence is *expected*, so the skeleton
//!    deliberately stops at the field level: it keeps the entry type and field names —
//!    the structure both parsers should agree on — and drops value internals, cite-key
//!    text, junk, and comments.

#![allow(dead_code)] // each test binary uses only part of this module.

use badness::bib;
use texlab_syntax::bibtex;

/// One node in the common bib skeleton. A document projects to a `Vec<BibAtom>` forest.
#[derive(Debug, Clone, PartialEq, Eq)]
pub enum BibAtom {
    /// `@type{ key, fields }`: the entry type (lowercased, `@` stripped) and its field
    /// atoms. Cite-key text and value internals are intentionally not modeled.
    Entry(String, Vec<BibAtom>),
    /// `@string{ name = value }` — a macro definition (the name/value are dropped; only
    /// its presence and position matter for concordance).
    StringDef,
    /// `@preamble{ value }`.
    Preamble,
    /// `name = value` — the field name, lowercased. The value is intentionally dropped.
    Field(String),
}

// --- badness projector ----------------------------------------------------

/// Project badness's bib CST for `text` onto the skeleton forest.
pub fn project_badness(text: &str) -> Vec<BibAtom> {
    use bib::syntax::SyntaxKind::*;
    let root = bib::parse(text).syntax();
    let mut out = Vec::new();
    for node in root.children() {
        match node.kind() {
            ENTRY => out.push(BibAtom::Entry(
                badness_entry_type(&node),
                badness_fields(&node),
            )),
            STRING_ENTRY => out.push(BibAtom::StringDef),
            PREAMBLE_ENTRY => out.push(BibAtom::Preamble),
            // COMMENT_ENTRY and JUNK carry no structure both parsers agree on.
            _ => {}
        }
    }
    out
}

/// The entry-type word of a badness `ENTRY` node, lowercased.
fn badness_entry_type(entry: &bib::syntax::SyntaxNode) -> String {
    use bib::syntax::SyntaxKind::ENTRY_TYPE;
    entry
        .children()
        .find(|n| n.kind() == ENTRY_TYPE)
        .map(|n| n.text().to_string().trim().to_ascii_lowercase())
        .unwrap_or_default()
}

/// The `Field(name)` atoms of a badness `ENTRY` node.
fn badness_fields(entry: &bib::syntax::SyntaxNode) -> Vec<BibAtom> {
    use bib::syntax::SyntaxKind::{FIELD, FIELD_NAME};
    entry
        .children()
        .filter(|n| n.kind() == FIELD)
        .map(|field| {
            let name = field
                .children()
                .find(|n| n.kind() == FIELD_NAME)
                .map(|n| n.text().to_string().trim().to_ascii_lowercase())
                .unwrap_or_default();
            BibAtom::Field(name)
        })
        .collect()
}

// --- texlab projector -----------------------------------------------------

/// Project texlab's bib CST for `text` onto the skeleton forest.
pub fn project_texlab(text: &str) -> Vec<BibAtom> {
    use bibtex::SyntaxKind::*;
    let root = texlab_root(text);
    let mut out = Vec::new();
    for node in root.children() {
        match node.kind() {
            ENTRY => out.push(BibAtom::Entry(
                texlab_entry_type(&node),
                texlab_fields(&node),
            )),
            STRING => out.push(BibAtom::StringDef),
            PREAMBLE => out.push(BibAtom::Preamble),
            _ => {}
        }
    }
    out
}

fn texlab_root(text: &str) -> bibtex::SyntaxNode {
    let green = texlab_parser::parse_bibtex(text);
    bibtex::SyntaxNode::new_root(green)
}

/// The entry-type word of a texlab `ENTRY` node. texlab keeps the leading `@` on the
/// `TYPE` token (e.g. `@article`), so strip it and lowercase.
fn texlab_entry_type(entry: &bibtex::SyntaxNode) -> String {
    use bibtex::SyntaxKind::TYPE;
    entry
        .children_with_tokens()
        .filter_map(|e| e.into_token())
        .find(|t| t.kind() == TYPE)
        .map(|t| t.text().trim().trim_start_matches('@').to_ascii_lowercase())
        .unwrap_or_default()
}

/// The `Field(name)` atoms of a texlab `ENTRY` node. In texlab a field's name is the
/// first `NAME` token inside the `FIELD` node.
fn texlab_fields(entry: &bibtex::SyntaxNode) -> Vec<BibAtom> {
    use bibtex::SyntaxKind::{FIELD, NAME};
    entry
        .children()
        .filter(|n| n.kind() == FIELD)
        .map(|field| {
            let name = field
                .children_with_tokens()
                .filter_map(|e| e.into_token())
                .find(|t| t.kind() == NAME)
                .map(|t| t.text().trim().to_ascii_lowercase())
                .unwrap_or_default();
            BibAtom::Field(name)
        })
        .collect()
}

// --- entry-recognition floor (hard gate) ----------------------------------

/// Count the top-level structured entries (`Entry` / `StringDef` / `Preamble`) in a
/// skeleton forest. The hard gate requires texlab to recognize at least as many as
/// badness does for a badness-clean file: a one-way soundness floor that catches
/// badness hallucinating structure texlab does not see, without depending on an error
/// channel texlab's bib parser lacks.
pub fn count_entries(forest: &[BibAtom]) -> usize {
    forest
        .iter()
        .filter(|a| {
            matches!(
                a,
                BibAtom::Entry(..) | BibAtom::StringDef | BibAtom::Preamble
            )
        })
        .count()
}

// --- serialization & similarity -------------------------------------------

/// Render a skeleton forest as indented S-expression lines (one atom per line;
/// indentation encodes depth so structure is comparison-significant).
pub fn render_lines(forest: &[BibAtom]) -> Vec<String> {
    let mut out = Vec::new();
    for atom in forest {
        render_atom(atom, 0, &mut out);
    }
    out
}

fn render_atom(atom: &BibAtom, depth: usize, out: &mut Vec<String>) {
    let pad = "  ".repeat(depth);
    let (head, children): (String, &[BibAtom]) = match atom {
        BibAtom::Entry(ty, ch) => (format!("(entry {ty})"), ch),
        BibAtom::StringDef => ("(string)".to_string(), &[]),
        BibAtom::Preamble => ("(preamble)".to_string(), &[]),
        BibAtom::Field(name) => (format!("(field {name})"), &[]),
    };
    out.push(format!("{pad}{head}"));
    for child in children {
        render_atom(child, depth + 1, out);
    }
}

/// Length of the longest common subsequence of two line slices. Copied from
/// `parse_skeleton.rs`.
pub fn lcs_len(a: &[String], b: &[String]) -> usize {
    if a.is_empty() || b.is_empty() {
        return 0;
    }
    let mut prev = vec![0usize; b.len() + 1];
    for line_a in a {
        let mut cur = vec![0usize; b.len() + 1];
        for (j, line_b) in b.iter().enumerate() {
            cur[j + 1] = if line_a == line_b {
                prev[j] + 1
            } else {
                cur[j].max(prev[j + 1])
            };
        }
        prev = cur;
    }
    prev[b.len()]
}

/// Dice coefficient over skeleton lines: `2·LCS / (|a| + |b|)`, in `0.0..=1.0`.
pub fn dice(a: &[String], b: &[String]) -> f64 {
    let denom = a.len() + b.len();
    if denom == 0 {
        return 1.0;
    }
    2.0 * lcs_len(a, b) as f64 / denom as f64
}