3
0
Fork 0
mirror of https://github.com/Z3Prover/z3 synced 2026-07-06 07:16:13 +00:00
z3/src/test/seq_rewriter.cpp
Lev Nachmanson 2a8f66f22b
[snapshot-regression-fix] Keep symbolic re.range non-empty; fix soundness regression on range membership (#10017)
## Summary

Fixes a **soundness regression** in the sequence/regex rewriter: a
symbolic character range such as `(re.range x x)` was unsoundly
collapsed to `re.empty`, causing a satisfiable membership constraint to
be reported `unsat`.

This was surfaced by the `snapshot-regression` corpus in
`Z3Prover/bench`.

- **Originating discussion:**
https://github.com/Z3Prover/bench/discussions/2761
- **Benchmark:** `iss-5873/bug-2.smt2` (in `Z3Prover/bench`, under
`inputs/issues/iss-5873/`)
- **z3 under test at capture:** `z3-4.17.0-x64-glibc-2.39` (Nightly)

## Divergence

The recorded oracle expects `sat`; current z3 returns `unsat`:

```diff
--- bug-2.expected.out (expected)
+++ produced (current z3)
@@ -1,3 +1,4 @@
-sat
-((tmp_str0 "\u{0}"))
+unsat
+(error "line 12 column 10: check annotation that says sat")
+(error "line 14 column 22: model is not available")
 (:reason-unknown "")
```

The benchmark asserts (simplified):

```smt2
(assert (= (str.in_re (str.replace tmp_str0 tmp_str0 tmp_str0)
                      (re.range tmp_str0 tmp_str0))
           (str.contains tmp_str0 tmp_str0)))
```

`str.contains x x` is always true and `str.replace x x x = x`, so this
requires `str.in_re x (re.range x x)` to hold, which is satisfiable
exactly when `x` is a single character (`len(x) = 1`).

## Root cause

`seq_rewriter::mk_re_range` treated any bound that is not a concrete
single-character literal as making the whole range **empty**:

```cpp
if (str().is_string(lo, slo) && slo.length() == 1) clo = slo[0];
else if (str().is_unit(lo, lo1) && m_util.is_const_char(lo1, clo)) ;
else is_empty = true;   // unsound for a symbolic bound
```

For a symbolic bound this is unsound: `(re.range x x)` denotes `{x}`
whenever `x` is a single character, not `∅`. Collapsing it to `re.empty`
makes `str.in_re x (re.range x x)` false, contradicting the (true)
`str.contains x x`, so the solver derives an unsound `unsat`.

`git blame` attributes this unsound collapse to z3 commit `15f33f458d`
("Derive with ranges (#9965)"), which post-dates the oracle capture.

## Fix

Two surgical changes in `src/ast/rewriter/seq_rewriter.cpp`:

1. **`mk_re_range`** no longer assumes emptiness for symbolic bounds. It
concludes `re.empty` only when it can *prove* emptiness — a bound whose
length can never be 1, or two concrete bounds with `lo > hi`. When a
bound is symbolic it returns `BR_FAILED` and keeps the range. Concrete
single-character ranges keep their existing handling (`lo == hi →
str.to_re`, inverted → `re.empty`).

2. **`mk_str_in_regexp`** reduces membership in a range that has a
symbolic bound to the equivalent length/order constraints, which are
sound and complete under SMT-LIB `re.range` semantics:

`str.in_re e (re.range lo hi)` ⟶ `len(lo)=1 ∧ len(hi)=1 ∧ len(e)=1 ∧ lo
≤ e ∧ e ≤ hi`

(using `str.<=`). The derivative engine only unfolds ranges whose bounds
are concrete characters, so without this reduction a symbolic-bound
range would otherwise be left unsolved.

## Validation

Rebuilt z3 from this branch on the workflow runner (`./configure && make
-C build -j$(nproc)`) and re-ran the failing benchmark with the same
option the snapshot capture uses (`-T:20`):

```
$ z3 -T:20 inputs/issues/iss-5873/bug-2.smt2
sat
((tmp_str0 "A"))
(:reason-unknown "")
```

The verdict is now **`sat`** (was `unsat`) — the soundness regression is
resolved. A correctness battery over concrete and symbolic ranges all
returns the expected results, e.g.:

- `(str.in_re "b" (re.range "a" "c"))` → `sat`, `(str.in_re "d"
(re.range "a" "c"))` → `unsat`
- `(str.in_re x (re.range x x))` → `sat`; with `(= (str.len x) 2)` →
`unsat`
- `(str.in_re "b" (re.range x y))` → `sat`; with `(str.< y x)` → `unsat`
- `(str.in_re "" (re.range x y))` → `unsat`; `(str.in_re "ab" (re.range
"a" "c"))` → `unsat`

The pre-existing concrete-range derivative fast path is unchanged.

### Note on the model value (benign, unrelated to this fix)

The model value differs from the recorded oracle: current z3 prints
`((tmp_str0 "A"))` whereas the oracle recorded `((tmp_str0 "\u{0}"))`.
Both are valid single-character models (the formula has many). This
difference is **pre-existing and unrelated to this fix**: even a bare
`(assert (= (str.len x) 1))` yields `"A"` on current z3. It stems from
the seq/char theory's default character assignment for
otherwise-unconstrained characters (`theory_char.cpp` assigns fresh
characters starting from `'A'`), not from range handling. I deliberately
did **not** force the character to `\u{0}` — adding `x = "\u{0}"` would
be unsound over-constraining, and changing the global default character
is out of scope for this soundness fix and would perturb unrelated
models. The output is therefore semantically equivalent to the oracle
(same `sat` verdict and reason-unknown) but not byte-identical.

---
*Draft for human review. Diagnosed and fixed by the
`snapshot-regression-fixer` maintenance workflow.*




> Generated by [Fix a Z3 snapshot-regression
divergence](https://github.com/Z3Prover/bench/actions/runs/28502614658)
· 890.7 AIC · ⌖ 46.8 AIC · ⊞ 9K ·
[◷](https://github.com/search?q=repo%3AZ3Prover%2Fz3+%22gh-aw-workflow-id%3A+snapshot-regression-fixer%22&type=pullrequests)

<!-- gh-aw-agentic-workflow: Fix a Z3 snapshot-regression divergence,
engine: copilot, version: 1.0.63, model: claude-opus-4.8, id:
28502614658, workflow_id: snapshot-regression-fixer, run:
https://github.com/Z3Prover/bench/actions/runs/28502614658 -->

<!-- gh-aw-workflow-id: snapshot-regression-fixer -->
<!-- gh-aw-workflow-call-id: Z3Prover/bench/snapshot-regression-fixer
-->

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
2026-07-02 14:00:51 -07:00

259 lines
11 KiB
C++
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/*++
Copyright (c) 2024 Microsoft Corporation
Regression tests for seq_rewriter smart constructors for regex ranges.
Tests:
1. Empty range (lo > hi) → re.none
2. Singleton range (lo == hi) → str.to_re lo
3. Range ∩ Range → reduced range or re.none
4. Range Range → merged range for overlapping/adjacent
5. Complement of range → one or two ranges
6. Downstream operators absorb empty ranges correctly
15. Symbolic-bound range membership rewrite (structural)
16. Symbolic-bound range membership: concrete element, symbolic bounds (structural)
17. Solver: (str.in_re x (re.range x x)) sat when len(x)=1
18. Solver: (str.in_re x (re.range x x)) unsat when len(x)=2
19. Solver: inverted symbolic bounds make membership unsatisfiable
--*/
#include "ast/arith_decl_plugin.h"
#include "ast/ast_pp.h"
#include "ast/reg_decl_plugins.h"
#include "ast/rewriter/th_rewriter.h"
#include "ast/seq_decl_plugin.h"
#include "smt/smt_context.h"
#include <iostream>
// Build a single-char string literal expression.
static expr_ref mk_str(ast_manager& m, seq_util& su, unsigned c) {
return expr_ref(su.str.mk_string(zstring(c)), m);
}
void tst_seq_rewriter() {
ast_manager m;
reg_decl_plugins(m);
th_rewriter rw(m);
seq_util su(m);
sort* str_sort = su.str.mk_string_sort();
sort* re_sort = su.re.mk_re(str_sort);
auto range = [&](unsigned lo, unsigned hi) -> expr_ref {
return expr_ref(su.re.mk_range(mk_str(m, su, lo), mk_str(m, su, hi)), m);
};
// Arbitrary regex variable for downstream tests.
app_ref R(m.mk_fresh_const("R", re_sort), m);
// -----------------------------------------------------------------------
// 1. Empty range (lo > hi) → re.none
// -----------------------------------------------------------------------
{
expr_ref e = range('z', 'a');
rw(e);
std::cout << "empty range lo>hi: " << mk_pp(e, m) << "\n";
ENSURE(su.re.is_empty(e));
}
// -----------------------------------------------------------------------
// 2. Singleton range (lo == hi) → str.to_re lo
// -----------------------------------------------------------------------
{
expr_ref e = range('a', 'a');
rw(e);
std::cout << "singleton range: " << mk_pp(e, m) << "\n";
expr* inner = nullptr;
ENSURE(su.re.is_to_re(e, inner));
}
// -----------------------------------------------------------------------
// 3. Range intersection: overlapping → smaller range
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_inter(range('a', 'z'), range('f', 'k')), m);
rw(e);
std::cout << "range inter overlapping: " << mk_pp(e, m) << "\n";
unsigned lo = 0, hi = 0;
ENSURE(su.re.is_range(e, lo, hi) && lo == 'f' && hi == 'k');
}
// -----------------------------------------------------------------------
// 4. Range intersection: disjoint → re.none
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_inter(range('a', 'f'), range('k', 'z')), m);
rw(e);
std::cout << "range inter disjoint: " << mk_pp(e, m) << "\n";
ENSURE(su.re.is_empty(e));
}
// -----------------------------------------------------------------------
// 5. Range intersection: touching at boundary → singleton (str.to_re "f")
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_inter(range('a', 'f'), range('f', 'z')), m);
rw(e);
std::cout << "range inter touching: " << mk_pp(e, m) << "\n";
expr* inner = nullptr;
ENSURE(su.re.is_to_re(e, inner));
}
// -----------------------------------------------------------------------
// 6. Range union: overlapping → merged range
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_union(range('a', 'f'), range('e', 'k')), m);
rw(e);
std::cout << "range union overlapping: " << mk_pp(e, m) << "\n";
unsigned lo = 0, hi = 0;
ENSURE(su.re.is_range(e, lo, hi) && lo == 'a' && hi == 'k');
}
// -----------------------------------------------------------------------
// 7. Range union: adjacent → merged range
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_union(range('a', 'f'), range('g', 'k')), m);
rw(e);
std::cout << "range union adjacent: " << mk_pp(e, m) << "\n";
unsigned lo = 0, hi = 0;
ENSURE(su.re.is_range(e, lo, hi) && lo == 'a' && hi == 'k');
}
// -----------------------------------------------------------------------
// 8. Range union: disjoint → stays as union
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_union(range('a', 'c'), range('m', 'z')), m);
rw(e);
std::cout << "range union disjoint (stays as union): " << mk_pp(e, m) << "\n";
ENSURE(!su.re.is_range(e));
}
// -----------------------------------------------------------------------
// 11. Downstream: (re.* (re.range "z" "a")) → str.to_re ""
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_star(range('z', 'a')), m);
rw(e);
std::cout << "star of empty range: " << mk_pp(e, m) << "\n";
expr* inner = nullptr;
// star of empty → epsilon (str.to_re "")
ENSURE(su.re.is_to_re(e, inner) && su.str.is_empty(inner));
}
// -----------------------------------------------------------------------
// 12. Downstream: concat absorbs empty range → re.none
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_concat(R, su.re.mk_concat(range('z', 'a'), R)), m);
rw(e);
std::cout << "concat absorbs empty range: " << mk_pp(e, m) << "\n";
ENSURE(su.re.is_empty(e));
}
// -----------------------------------------------------------------------
// 13. Downstream: union absorbs empty range → R
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_union(R, range('z', 'a')), m);
rw(e);
std::cout << "union absorbs empty range: " << mk_pp(e, m) << "\n";
ENSURE(e.get() == R.get());
}
// -----------------------------------------------------------------------
// 14. Downstream: inter absorbs empty range → re.none
// -----------------------------------------------------------------------
{
expr_ref e(su.re.mk_inter(R, range('z', 'a')), m);
rw(e);
std::cout << "inter absorbs empty range: " << mk_pp(e, m) << "\n";
ENSURE(su.re.is_empty(e));
}
// -----------------------------------------------------------------------
// 15. Symbolic-bound range membership rewrite (structural).
// (str.in_re x (re.range x x)) with symbolic x should be unfolded
// by the rewriter into a conjunction of length and ordering
// constraints, not left stuck as an uninterpreted membership term.
// -----------------------------------------------------------------------
{
app_ref x(m.mk_fresh_const("x", str_sort), m);
expr_ref rng(su.re.mk_range(x, x), m);
expr_ref e(su.re.mk_in_re(x, rng), m);
rw(e);
std::cout << "symbolic range (x in [x,x]): " << mk_pp(e, m) << "\n";
ENSURE(m.is_and(e));
}
// -----------------------------------------------------------------------
// 16. Symbolic-bound range membership: concrete element, symbolic bounds.
// (str.in_re "b" (re.range lo hi)) should also be unfolded to a
// conjunction when lo/hi are free variables.
// -----------------------------------------------------------------------
{
app_ref lo(m.mk_fresh_const("lo", str_sort), m);
app_ref hi(m.mk_fresh_const("hi", str_sort), m);
expr_ref b_str(su.str.mk_string(zstring('b')), m);
expr_ref rng(su.re.mk_range(lo, hi), m);
expr_ref e(su.re.mk_in_re(b_str, rng), m);
rw(e);
std::cout << "symbolic range (\"b\" in [lo,hi]): " << mk_pp(e, m) << "\n";
ENSURE(m.is_and(e));
}
// -----------------------------------------------------------------------
// Solver-level tests: the unfolded conjunction must be decidable.
// -----------------------------------------------------------------------
{
arith_util a_util(m);
// 17. sat: (str.in_re x (re.range x x)) ∧ len(x)=1
{
smt_params sp;
smt::context ctx(m, sp);
app_ref x(m.mk_fresh_const("x", str_sort), m);
ctx.assert_expr(su.re.mk_in_re(x, su.re.mk_range(x, x)));
ctx.assert_expr(m.mk_eq(su.str.mk_length(x), a_util.mk_int(1)));
lbool res = ctx.check();
std::cout << "symbolic range solver sat (len=1): " << res << "\n";
ENSURE(res == l_true);
}
// 18. unsat: (str.in_re x (re.range x x)) ∧ len(x)=2
// The unfolded membership requires len(x)=1, which contradicts len(x)=2.
{
smt_params sp;
smt::context ctx(m, sp);
app_ref x(m.mk_fresh_const("x", str_sort), m);
ctx.assert_expr(su.re.mk_in_re(x, su.re.mk_range(x, x)));
ctx.assert_expr(m.mk_eq(su.str.mk_length(x), a_util.mk_int(2)));
lbool res = ctx.check();
std::cout << "symbolic range solver unsat (len=2): " << res << "\n";
ENSURE(res == l_false);
}
// 19. unsat: inverted symbolic bounds make membership false.
// (str.in_re "b" (re.range lo hi)) ∧ lo="z" ∧ hi="a"
// The unfolded conjunction requires lo <=_lex "b" <=_lex hi, but
// "z" > "b" > "a" so the ordering constraints are unsatisfiable.
{
smt_params sp;
smt::context ctx(m, sp);
app_ref lo(m.mk_fresh_const("lo", str_sort), m);
app_ref hi(m.mk_fresh_const("hi", str_sort), m);
expr_ref b_str(su.str.mk_string(zstring('b')), m);
ctx.assert_expr(su.re.mk_in_re(b_str, su.re.mk_range(lo, hi)));
ctx.assert_expr(m.mk_eq(lo, su.str.mk_string(zstring('z'))));
ctx.assert_expr(m.mk_eq(hi, su.str.mk_string(zstring('a'))));
lbool res = ctx.check();
std::cout << "symbolic range solver inverted bounds unsat: " << res << "\n";
ENSURE(res == l_false);
}
}
std::cout << "tst_seq_rewriter: all tests passed\n";
}