circle-ir 3.136.0 → 3.137.0

package/dist/analysis/passes/language-sources-pass.js

@@ -228,6 +228,11 @@ export class LanguageSourcesPass {
             for (const finding of findGoMongoNosqlInjectionFindings(code, graph.ir.meta.file)) {
                 ctx.addFinding(finding);
             }
+            // Sprint 87 (#189): Go ldap_injection — go-ldap NewSearchRequest with
+            // a tainted filter argument (slot 7).
+            for (const finding of findGoLdapInjectionFindings(code, graph.ir.meta.file)) {
+                ctx.addFinding(finding);
+            }
         }
         // -- Python: safe-handler sanitizer detectors (cognium-dev #114 Sprint 31) --
         if (language === 'python') {
@@ -320,6 +325,16 @@ export class LanguageSourcesPass {
             for (const finding of findRustEvalCrateCodeInjectionFindings(code, graph.ir.meta.file)) {
                 ctx.addFinding(finding);
             }
+            // Sprint 87 (#189): Rust ldap_injection — ldap3 LdapConn::search with
+            // tainted filter argument (slot 3).
+            for (const finding of findRustLdapInjectionFindings(code, graph.ir.meta.file)) {
+                ctx.addFinding(finding);
+            }
+            // Sprint 87 (#189): Rust log_injection — log crate macros
+            // (info!/warn!/error!/debug!/trace!) with tainted format args.
+            for (const finding of findRustLogInjectionFindings(code, graph.ir.meta.file)) {
+                ctx.addFinding(finding);
+            }
         }
         // -- JavaScript/TypeScript: Sprint 73 (#216 Pattern A + B) — ETE
         // sanitizer-chain recognition (JSON.parse / bcrypt.hash / csv
@@ -363,6 +378,11 @@ export class LanguageSourcesPass {
             for (const finding of findJsUtilFormatFormatStringFindings(code, graph.ir.meta.file)) {
                 ctx.addFinding(finding);
             }
+            // Sprint 87 (#189): JS/TS ldap_injection — ldapjs/ldapts
+            // client.search(base, { filter: <tainted>, ... }).
+            for (const finding of findJsLdapInjectionFindings(code, graph.ir.meta.file)) {
+                ctx.addFinding(finding);
+            }
         }
         // -- Java: Sprint 73 (#216 Pattern A) — Jackson readValue / Gson
         // fromJson recognized as ETE terminator (does not affect
@@ -6448,4 +6468,667 @@ export function findPythonHeaderCrlfInjectionFindings(code, file) {
     }
     return findings;
 }
+/**
+ * Sprint 87 detector A (#189) — JavaScript / TypeScript LDAP injection
+ * via ldapjs / ldapts `client.search(base, { filter: <tainted>, ... })`.
+ *
+ * ldapjs and ldapts share the same call shape: `client.search(base, opts,
+ * cb)` where `opts.filter` is the LDAP filter string. When an attacker
+ * controls the filter content, they can break out of the intended filter
+ * (e.g. `*)(uid=*` injection) to enumerate the directory. CWE-90.
+ *
+ * The detector tracks taint propagation from Express request extractors
+ * (`req.query.X`, `req.body.X`, `req.params.X`, `req.headers.X`,
+ * `req.cookies.X`) through both `let|const|var` assignments and
+ * template-literal / `+`-concat construction of the filter string.
+ */
+export function findJsLdapInjectionFindings(code, file) {
+    const findings = [];
+    if (typeof code !== 'string' || code.length === 0)
+        return findings;
+    // Gate: must reference an ldapjs/ldapts-shaped `.search(...)` call AND
+    // an Express-shaped request extractor.
+    if (!/\.\s*search\s*\(/.test(code))
+        return findings;
+    if (!/\breq\s*\.\s*(?:query|body|params|headers|cookies)\b/.test(code)) {
+        return findings;
+    }
+    // Soft library gate — only fire when ldapjs or ldapts is in the file.
+    if (!/\b(?:ldapjs|ldapts|require\(['"]ldapjs['"]\)|from\s+['"]ldapts['"])\b/.test(code)) {
+        return findings;
+    }
+    const lines = code.split('\n');
+    const reqExtractRe = /\breq\s*\.\s*(?:query|body|params|headers|cookies)\b/;
+    const taintedVars = new Set();
+    // 3-pass whole-file taint propagation across `(const|let|var) x = expr`.
+    for (let pass = 0; pass < 3; pass++) {
+        const before = taintedVars.size;
+        for (let i = 0; i < lines.length; i++) {
+            const t = lines[i].trim();
+            if (t.startsWith('//'))
+                continue;
+            const m = t.match(/^(?:const|let|var)\s+(\w+)(?:\s*:\s*[^=]+)?\s*=\s*(.+?);?$/);
+            if (!m)
+                continue;
+            const lhs = m[1];
+            const rhs = m[2];
+            if (taintedVars.has(lhs))
+                continue;
+            if (reqExtractRe.test(rhs)) {
+                taintedVars.add(lhs);
+                continue;
+            }
+            for (const v of taintedVars) {
+                if (new RegExp(`\\b${v}\\b`).test(rhs)) {
+                    taintedVars.add(lhs);
+                    break;
+                }
+            }
+        }
+        if (taintedVars.size === before)
+            break;
+    }
+    if (taintedVars.size === 0)
+        return findings;
+    // Find lines containing `filter: <expr>` (object property) where the
+    // expression resolves to a tainted symbol, plus the shorthand
+    // `{ filter, ... }` / `{ ..., filter }` form (ES6 property shorthand
+    // means `{ filter }` is equivalent to `{ filter: filter }`).
+    const filterPropRe = /\bfilter\s*:\s*([^,}\n]+)/;
+    const filterShorthandRe = /(?:^|[\{,])\s*filter\s*[,}]/;
+    const seen = new Set();
+    for (let i = 0; i < lines.length; i++) {
+        const line = lines[i];
+        const fm = line.match(filterPropRe);
+        let expr = null;
+        if (fm) {
+            expr = fm[1].trim();
+        }
+        else if (filterShorthandRe.test(line)) {
+            // Shorthand: value is the local symbol named "filter".
+            expr = 'filter';
+        }
+        else {
+            continue;
+        }
+        // Tainted if any tainted var appears in the expression.
+        let tainted = false;
+        for (const v of taintedVars) {
+            if (new RegExp(`\\b${v}\\b`).test(expr)) {
+                tainted = true;
+                break;
+            }
+        }
+        if (!tainted)
+            continue;
+        if (seen.has(i + 1))
+            continue;
+        seen.add(i + 1);
+        findings.push({
+            id: `ldap_injection-${file}-${i + 1}-js-ldap-filter`,
+            pass: 'language-sources',
+            category: 'security',
+            rule_id: 'ldap_injection',
+            cwe: 'CWE-90',
+            severity: 'critical',
+            level: 'error',
+            message: 'LDAP injection: ldapjs/ldapts `client.search(..., { filter: ' +
+                '<tainted>, ... })` lets the attacker break out of the filter ' +
+                'expression (e.g. `*)(uid=*`) and enumerate the directory. ' +
+                'Escape the user input with a tight allowlist (`[A-Za-z0-9_-]+`) ' +
+                'or use a structured filter builder.',
+            file,
+            line: i + 1,
+            snippet: line.trim(),
+        });
+    }
+    return findings;
+}
+/**
+ * Sprint 87 detector B (#189) — Go LDAP injection via go-ldap/ldap.v3
+ * `ldap.NewSearchRequest(base, scope, deref, sizeLimit, timeLimit,
+ * typesOnly, <tainted-filter>, attributes, controls)`.
+ *
+ * The 7th positional argument is the LDAP filter. The detector tracks
+ * tainted strings derived from `r.URL.Query().Get(...)` / `r.Form.Get`
+ * / `r.PostForm.Get` / `r.FormValue` / `r.PostFormValue` through
+ * `:=`/`=` assignments and `fmt.Sprintf` calls, then fires when the
+ * filter slot resolves to a tainted symbol. CWE-90.
+ */
+export function findGoLdapInjectionFindings(code, file) {
+    const findings = [];
+    if (typeof code !== 'string' || code.length === 0)
+        return findings;
+    if (!/\bldap\s*\.\s*NewSearchRequest\s*\(/.test(code))
+        return findings;
+    if (!/\br\s*\.\s*(?:URL\s*\.\s*Query\s*\(\s*\)|Form|PostForm|FormValue|PostFormValue|Header)/.test(code)) {
+        return findings;
+    }
+    const lines = code.split('\n');
+    const reqExtractRe = /\br\s*\.\s*(?:URL\s*\.\s*Query\s*\(\s*\)\s*\.\s*Get|Form\s*\.\s*Get|PostForm\s*\.\s*Get|FormValue|PostFormValue|Header\s*\.\s*Get)\s*\(/;
+    const taintedVars = new Set();
+    for (let pass = 0; pass < 3; pass++) {
+        const before = taintedVars.size;
+        for (let i = 0; i < lines.length; i++) {
+            const t = lines[i].trim();
+            if (t.startsWith('//'))
+                continue;
+            // `name := expr` or `name = expr`
+            const m = t.match(/^(\w+)\s*(?::=|=)\s*(.+?)$/);
+            if (!m)
+                continue;
+            const lhs = m[1];
+            const rhs = m[2];
+            if (taintedVars.has(lhs))
+                continue;
+            if (reqExtractRe.test(rhs)) {
+                taintedVars.add(lhs);
+                continue;
+            }
+            for (const v of taintedVars) {
+                if (new RegExp(`\\b${v}\\b`).test(rhs)) {
+                    taintedVars.add(lhs);
+                    break;
+                }
+            }
+        }
+        if (taintedVars.size === before)
+            break;
+    }
+    if (taintedVars.size === 0)
+        return findings;
+    // Multiline-tolerant scan: NewSearchRequest call may span several
+    // lines. Find each opener and collect its full argument span.
+    const seen = new Set();
+    for (let i = 0; i < lines.length; i++) {
+        const openIdx = lines[i].indexOf('NewSearchRequest(');
+        if (openIdx === -1)
+            continue;
+        // Aggregate text from `i` until we balance the parens.
+        let depth = 0;
+        let buf = '';
+        let endLine = i;
+        let started = false;
+        for (let j = i; j < Math.min(i + 25, lines.length); j++) {
+            for (const ch of lines[j]) {
+                if (ch === '(') {
+                    depth++;
+                    started = true;
+                }
+                else if (ch === ')') {
+                    depth--;
+                }
+                buf += ch;
+                if (started && depth === 0) {
+                    endLine = j;
+                    break;
+                }
+            }
+            if (started && depth === 0)
+                break;
+            buf += '\n';
+        }
+        // buf now contains "NewSearchRequest(...)". Strip prefix.
+        const argsStart = buf.indexOf('(');
+        if (argsStart === -1)
+            continue;
+        const args = buf.substring(argsStart + 1, buf.length - 1);
+        // Top-level split on commas (string-literal aware).
+        const parts = [];
+        {
+            let d = 0;
+            let b = '';
+            let inStr = null;
+            for (let k = 0; k < args.length; k++) {
+                const ch = args[k];
+                if (inStr) {
+                    if (ch === '\\') {
+                        b += ch + (args[k + 1] ?? '');
+                        k++;
+                        continue;
+                    }
+                    if (ch === inStr)
+                        inStr = null;
+                    b += ch;
+                    continue;
+                }
+                if (ch === '"' || ch === '`') {
+                    inStr = ch;
+                    b += ch;
+                    continue;
+                }
+                if (ch === '(' || ch === '[' || ch === '{')
+                    d++;
+                else if (ch === ')' || ch === ']' || ch === '}')
+                    d--;
+                if (ch === ',' && d === 0) {
+                    parts.push(b);
+                    b = '';
+                    continue;
+                }
+                b += ch;
+            }
+            if (b.trim().length > 0)
+                parts.push(b);
+        }
+        if (parts.length < 7)
+            continue;
+        const filterExpr = parts[6].trim();
+        let tainted = false;
+        for (const v of taintedVars) {
+            if (new RegExp(`\\b${v}\\b`).test(filterExpr)) {
+                tainted = true;
+                break;
+            }
+        }
+        if (!tainted)
+            continue;
+        if (seen.has(i + 1))
+            continue;
+        seen.add(i + 1);
+        findings.push({
+            id: `ldap_injection-${file}-${i + 1}-go-newsearchrequest`,
+            pass: 'language-sources',
+            category: 'security',
+            rule_id: 'ldap_injection',
+            cwe: 'CWE-90',
+            severity: 'critical',
+            level: 'error',
+            message: 'LDAP injection: go-ldap `ldap.NewSearchRequest(..., <tainted ' +
+                'filter>, ...)` lets the attacker break out of the filter ' +
+                'expression and enumerate the directory. Escape the user input ' +
+                'with `ldap.EscapeFilter(...)` or use a structured filter ' +
+                'builder.',
+            file,
+            line: i + 1,
+            snippet: (lines[i] + (endLine > i ? ' …' : '')).trim(),
+        });
+    }
+    return findings;
+}
+/**
+ * Sprint 87 detector C (#189) — Rust LDAP injection via ldap3
+ * `LdapConn::search(base, scope, &<tainted-filter>, attrs)` (and the
+ * async `Ldap::search(...)` mirror).
+ *
+ * The 3rd positional argument is the LDAP filter. Tainted strings come
+ * from actix-web `web::Query<HashMap<String, String>>` / `web::Path` /
+ * `web::Form` / `web::Json` parameter types (extractor handlers) and
+ * propagate through `let` bindings and `format!(...)` macros. CWE-90.
+ */
+export function findRustLdapInjectionFindings(code, file) {
+    const findings = [];
+    if (typeof code !== 'string' || code.length === 0)
+        return findings;
+    if (!/\.\s*search\s*\(/.test(code))
+        return findings;
+    if (!/\b(?:ldap3|LdapConn|Ldap)\b/.test(code))
+        return findings;
+    const lines = code.split('\n');
+    const extractorTypeRe = /:\s*(?:String|Bytes|bytes::Bytes|axum::body::Bytes|web::Query\b|web::Path\b|web::Form\b|web::Json\b|HttpRequest\b|actix_web::HttpRequest\b)/;
+    const fns = [];
+    let cur = null;
+    for (let i = 0; i < lines.length; i++) {
+        const t = lines[i];
+        if (/^\s*(?:pub\s+)?(?:async\s+)?fn\s+\w+\s*\(/.test(t)) {
+            if (cur) {
+                cur.end = i - 1;
+                fns.push(cur);
+            }
+            cur = { start: i, end: lines.length - 1, tainted: new Set() };
+            // Assemble multi-line header.
+            let headerJoined = '';
+            for (let j = i; j < Math.min(i + 12, lines.length); j++) {
+                headerJoined += lines[j];
+                if (/\{\s*$/.test(lines[j]))
+                    break;
+            }
+            const open = headerJoined.indexOf('(');
+            const close = headerJoined.lastIndexOf(')');
+            if (open !== -1 && close > open) {
+                const params = headerJoined.substring(open + 1, close);
+                let depth = 0;
+                let buf = '';
+                const parts = [];
+                for (const ch of params) {
+                    if (ch === '<' || ch === '(')
+                        depth++;
+                    else if (ch === '>' || ch === ')')
+                        depth--;
+                    if (ch === ',' && depth === 0) {
+                        parts.push(buf);
+                        buf = '';
+                        continue;
+                    }
+                    buf += ch;
+                }
+                if (buf.trim().length > 0)
+                    parts.push(buf);
+                for (const p of parts) {
+                    const pm = p.match(/(?:mut\s+)?(\w+)\s*:/);
+                    if (!pm)
+                        continue;
+                    if (extractorTypeRe.test(p))
+                        cur.tainted.add(pm[1]);
+                }
+            }
+        }
+    }
+    if (cur)
+        fns.push(cur);
+    // Propagate taint through `let` bindings.
+    for (const fn of fns) {
+        for (let pass = 0; pass < 3; pass++) {
+            const before = fn.tainted.size;
+            for (let i = fn.start; i <= fn.end; i++) {
+                const t = lines[i].trim();
+                const m = t.match(/^let\s+(?:mut\s+)?(\w+)\s*(?::\s*[^=]+)?=\s*(.+?);?$/);
+                if (!m)
+                    continue;
+                const lhs = m[1];
+                const rhs = m[2];
+                if (fn.tainted.has(lhs))
+                    continue;
+                for (const v of fn.tainted) {
+                    if (new RegExp(`\\b${v}\\b`).test(rhs)) {
+                        fn.tainted.add(lhs);
+                        break;
+                    }
+                }
+            }
+            if (fn.tainted.size === before)
+                break;
+        }
+    }
+    // Multiline-tolerant `.search(` call walker.
+    const seen = new Set();
+    for (const fn of fns) {
+        if (fn.tainted.size === 0)
+            continue;
+        for (let i = fn.start; i <= fn.end; i++) {
+            const searchIdx = lines[i].indexOf('.search(');
+            if (searchIdx === -1)
+                continue;
+            // Assemble args span.
+            let depth = 0;
+            let buf = '';
+            let started = false;
+            for (let j = i; j < Math.min(i + 15, lines.length); j++) {
+                const startCol = j === i ? searchIdx : 0;
+                for (let k = startCol; k < lines[j].length; k++) {
+                    const ch = lines[j][k];
+                    if (ch === '(') {
+                        depth++;
+                        started = true;
+                    }
+                    else if (ch === ')') {
+                        depth--;
+                    }
+                    buf += ch;
+                    if (started && depth === 0)
+                        break;
+                }
+                if (started && depth === 0)
+                    break;
+                buf += '\n';
+            }
+            const argsStart = buf.indexOf('(');
+            if (argsStart === -1)
+                continue;
+            const args = buf.substring(argsStart + 1, buf.length - 1);
+            // String-literal aware top-level comma split.
+            const parts = [];
+            {
+                let d = 0;
+                let b = '';
+                let inStr = null;
+                for (let k = 0; k < args.length; k++) {
+                    const ch = args[k];
+                    if (inStr) {
+                        if (ch === '\\') {
+                            b += ch + (args[k + 1] ?? '');
+                            k++;
+                            continue;
+                        }
+                        if (ch === inStr)
+                            inStr = null;
+                        b += ch;
+                        continue;
+                    }
+                    if (ch === '"') {
+                        inStr = ch;
+                        b += ch;
+                        continue;
+                    }
+                    if (ch === '(' || ch === '[' || ch === '{')
+                        d++;
+                    else if (ch === ')' || ch === ']' || ch === '}')
+                        d--;
+                    if (ch === ',' && d === 0) {
+                        parts.push(b);
+                        b = '';
+                        continue;
+                    }
+                    b += ch;
+                }
+                if (b.trim().length > 0)
+                    parts.push(b);
+            }
+            // ldap3 search signature: search(base, scope, filter, attrs)
+            if (parts.length < 3)
+                continue;
+            const filterExpr = parts[2].trim();
+            let tainted = false;
+            for (const v of fn.tainted) {
+                if (new RegExp(`\\b${v}\\b`).test(filterExpr)) {
+                    tainted = true;
+                    break;
+                }
+            }
+            if (!tainted)
+                continue;
+            if (seen.has(i + 1))
+                continue;
+            seen.add(i + 1);
+            findings.push({
+                id: `ldap_injection-${file}-${i + 1}-rust-ldap3-search`,
+                pass: 'language-sources',
+                category: 'security',
+                rule_id: 'ldap_injection',
+                cwe: 'CWE-90',
+                severity: 'critical',
+                level: 'error',
+                message: 'LDAP injection: ldap3 `LdapConn::search(..., &<tainted ' +
+                    'filter>, ...)` lets the attacker break out of the filter ' +
+                    'expression and enumerate the directory. Escape the user input ' +
+                    'or use a structured filter builder.',
+                file,
+                line: i + 1,
+                snippet: lines[i].trim(),
+            });
+        }
+    }
+    return findings;
+}
+/**
+ * Sprint 87 detector D (#189) — Rust log injection via the `log` crate
+ * macros (`info!`, `warn!`, `error!`, `debug!`, `trace!`) where a
+ * tainted value is interpolated into the format args.
+ *
+ * Unsanitized CRLF in log lines can split log entries, forge
+ * authentication events, or escape into log-aggregation pipelines that
+ * parse newlines as record boundaries. CWE-117.
+ */
+export function findRustLogInjectionFindings(code, file) {
+    const findings = [];
+    if (typeof code !== 'string' || code.length === 0)
+        return findings;
+    if (!/\b(?:info|warn|error|debug|trace)\s*!\s*\(/.test(code))
+        return findings;
+    if (!/\b(?:log|tracing)\b/.test(code))
+        return findings;
+    const lines = code.split('\n');
+    const extractorTypeRe = /:\s*(?:String|Bytes|bytes::Bytes|axum::body::Bytes|web::Query\b|web::Path\b|web::Form\b|web::Json\b|HttpRequest\b|actix_web::HttpRequest\b)/;
+    const fns = [];
+    let cur = null;
+    for (let i = 0; i < lines.length; i++) {
+        const t = lines[i];
+        if (/^\s*(?:pub\s+)?(?:async\s+)?fn\s+\w+\s*\(/.test(t)) {
+            if (cur) {
+                cur.end = i - 1;
+                fns.push(cur);
+            }
+            cur = { start: i, end: lines.length - 1, tainted: new Set() };
+            let headerJoined = '';
+            for (let j = i; j < Math.min(i + 12, lines.length); j++) {
+                headerJoined += lines[j];
+                if (/\{\s*$/.test(lines[j]))
+                    break;
+            }
+            const open = headerJoined.indexOf('(');
+            const close = headerJoined.lastIndexOf(')');
+            if (open !== -1 && close > open) {
+                const params = headerJoined.substring(open + 1, close);
+                let depth = 0;
+                let buf = '';
+                const parts = [];
+                for (const ch of params) {
+                    if (ch === '<' || ch === '(')
+                        depth++;
+                    else if (ch === '>' || ch === ')')
+                        depth--;
+                    if (ch === ',' && depth === 0) {
+                        parts.push(buf);
+                        buf = '';
+                        continue;
+                    }
+                    buf += ch;
+                }
+                if (buf.trim().length > 0)
+                    parts.push(buf);
+                for (const p of parts) {
+                    const pm = p.match(/(?:mut\s+)?(\w+)\s*:/);
+                    if (!pm)
+                        continue;
+                    if (extractorTypeRe.test(p))
+                        cur.tainted.add(pm[1]);
+                }
+            }
+        }
+    }
+    if (cur)
+        fns.push(cur);
+    for (const fn of fns) {
+        for (let pass = 0; pass < 3; pass++) {
+            const before = fn.tainted.size;
+            for (let i = fn.start; i <= fn.end; i++) {
+                const t = lines[i].trim();
+                const m = t.match(/^let\s+(?:mut\s+)?(\w+)\s*(?::\s*[^=]+)?=\s*(.+?);?$/);
+                if (!m)
+                    continue;
+                const lhs = m[1];
+                const rhs = m[2];
+                if (fn.tainted.has(lhs))
+                    continue;
+                for (const v of fn.tainted) {
+                    if (new RegExp(`\\b${v}\\b`).test(rhs)) {
+                        fn.tainted.add(lhs);
+                        break;
+                    }
+                }
+            }
+            if (fn.tainted.size === before)
+                break;
+        }
+    }
+    const macroRe = /\b(info|warn|error|debug|trace)\s*!\s*\(\s*([^;]+?)\)\s*;?\s*$/;
+    const seen = new Set();
+    for (const fn of fns) {
+        if (fn.tainted.size === 0)
+            continue;
+        for (let i = fn.start; i <= fn.end; i++) {
+            const line = lines[i];
+            const m = line.match(macroRe);
+            if (!m)
+                continue;
+            const macroName = m[1];
+            const argSpan = m[2];
+            // Split on top-level commas to separate the fmt-string from args.
+            const parts = [];
+            {
+                let d = 0;
+                let b = '';
+                let inStr = null;
+                for (let k = 0; k < argSpan.length; k++) {
+                    const ch = argSpan[k];
+                    if (inStr) {
+                        if (ch === '\\') {
+                            b += ch + (argSpan[k + 1] ?? '');
+                            k++;
+                            continue;
+                        }
+                        if (ch === inStr)
+                            inStr = null;
+                        b += ch;
+                        continue;
+                    }
+                    if (ch === '"') {
+                        inStr = ch;
+                        b += ch;
+                        continue;
+                    }
+                    if (ch === '(' || ch === '[' || ch === '{')
+                        d++;
+                    else if (ch === ')' || ch === ']' || ch === '}')
+                        d--;
+                    if (ch === ',' && d === 0) {
+                        parts.push(b);
+                        b = '';
+                        continue;
+                    }
+                    b += ch;
+                }
+                if (b.trim().length > 0)
+                    parts.push(b);
+            }
+            if (parts.length < 2)
+                continue;
+            // Any of the format args (parts[1..]) being tainted = fire.
+            let tainted = false;
+            for (let p = 1; p < parts.length; p++) {
+                for (const v of fn.tainted) {
+                    if (new RegExp(`\\b${v}\\b`).test(parts[p])) {
+                        tainted = true;
+                        break;
+                    }
+                }
+                if (tainted)
+                    break;
+            }
+            if (!tainted)
+                continue;
+            const key = `${i + 1}:${macroName}`;
+            if (seen.has(key))
+                continue;
+            seen.add(key);
+            findings.push({
+                id: `log_injection-${file}-${i + 1}-rust-${macroName}`,
+                pass: 'language-sources',
+                category: 'security',
+                rule_id: 'log_injection',
+                cwe: 'CWE-117',
+                severity: 'medium',
+                level: 'warning',
+                message: `Log injection: Rust \`${macroName}!(...)\` interpolates a ` +
+                    'tainted value into the log line. Unsanitized CRLF lets an ' +
+                    'attacker forge log entries or split records. Strip control ' +
+                    'characters or use a structured logging API.',
+                file,
+                line: i + 1,
+                snippet: line.trim(),
+            });
+        }
+    }
+    return findings;
+}
 //# sourceMappingURL=language-sources-pass.js.map