Fix dpnp.insert ignoring out-of-bounds negative indices in multi-element obj (#3041)

`dpnp.insert` has two index paths: the singleton path
(`_insert_singleton_index`) already validated bounds, but the
multi-element array path (`_insert_array_indices`) normalized negative
indices with `indices[indices < 0] += n` without any bounds check. As a
result, an out-of-bounds negative index mixed with in-bounds ones — e.g.
`dpnp.insert([0, 1, 2], [-6, 0], [9, 8])` — silently produced a wrong
result instead of raising.

This PR adds the same bounds validation NumPy introduced: if any index
is `< -n` or `> n`, `IndexError` is raised with the standard `index {i}
is out of bounds for axis {axis} with size {n}` message. In-bounds and
out-of-bounds indices mixed in a single call now consistently raise. 1d7c6fa

Author: github-actions[bot]

_modules/dpnp/dpnp_iface_manipulation.html

@@ -1287,6 +1287,39 @@ <h1>Source code for dpnp.dpnp_iface_manipulation</h1><div class="highlight"><pre
     <span class="p">)</span>
 
 
+<span class="k">def</span><span class="w"> </span><span class="nf">_check_index_bounds</span><span class="p">(</span><span class="n">obj</span><span class="p">,</span> <span class="n">indices</span><span class="p">,</span> <span class="n">n</span><span class="p">,</span> <span class="n">axis</span><span class="p">):</span>
+<span class="w">    </span><span class="sd">"""</span>
+<span class="sd">    Raise ``IndexError`` if any index in `obj` is out of bounds for `axis`.</span>
+
+<span class="sd">    Mirrors the size-1 path in ``_insert_singleton_index``: when `obj` lives on</span>
+<span class="sd">    the host (a Python sequence/scalar or NumPy array) the bounds are validated</span>
+<span class="sd">    with NumPy, avoiding a device sync entirely. A slice cannot be out of bounds</span>
+<span class="sd">    (``obj.indices(n)`` is clamped to ``[0, n]``), so it is skipped. Only a</span>
+<span class="sd">    device array `obj` needs a single host transfer to read its extremes.</span>
+
+<span class="sd">    """</span>
+
+    <span class="k">if</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">obj</span><span class="p">,</span> <span class="nb">slice</span><span class="p">)</span> <span class="ow">or</span> <span class="n">indices</span><span class="o">.</span><span class="n">size</span> <span class="o">==</span> <span class="mi">0</span><span class="p">:</span>
+        <span class="k">return</span>
+
+    <span class="k">if</span> <span class="n">dpnp</span><span class="o">.</span><span class="n">is_supported_array_type</span><span class="p">(</span><span class="n">obj</span><span class="p">):</span>
+        <span class="n">min_idx</span><span class="p">,</span> <span class="n">max_idx</span> <span class="o">=</span> <span class="n">dpnp</span><span class="o">.</span><span class="n">stack</span><span class="p">([</span><span class="n">indices</span><span class="o">.</span><span class="n">min</span><span class="p">(),</span> <span class="n">indices</span><span class="o">.</span><span class="n">max</span><span class="p">()])</span><span class="o">.</span><span class="n">asnumpy</span><span class="p">()</span>
+    <span class="k">else</span><span class="p">:</span>
+        <span class="n">host_obj</span> <span class="o">=</span> <span class="n">numpy</span><span class="o">.</span><span class="n">asarray</span><span class="p">(</span><span class="n">obj</span><span class="p">)</span>
+        <span class="k">if</span> <span class="n">host_obj</span><span class="o">.</span><span class="n">dtype</span> <span class="o">==</span> <span class="n">dpnp</span><span class="o">.</span><span class="n">bool</span><span class="p">:</span>
+            <span class="c1"># a boolean mask selects positions, which (for an oversized mask)</span>
+            <span class="c1"># can fall out of bounds, so validate the flatnonzero result</span>
+            <span class="n">host_obj</span> <span class="o">=</span> <span class="n">numpy</span><span class="o">.</span><span class="n">flatnonzero</span><span class="p">(</span><span class="n">host_obj</span><span class="p">)</span>
+        <span class="n">min_idx</span><span class="p">,</span> <span class="n">max_idx</span> <span class="o">=</span> <span class="n">host_obj</span><span class="o">.</span><span class="n">min</span><span class="p">(),</span> <span class="n">host_obj</span><span class="o">.</span><span class="n">max</span><span class="p">()</span>
+
+    <span class="n">min_idx</span><span class="p">,</span> <span class="n">max_idx</span> <span class="o">=</span> <span class="nb">int</span><span class="p">(</span><span class="n">min_idx</span><span class="p">),</span> <span class="nb">int</span><span class="p">(</span><span class="n">max_idx</span><span class="p">)</span>
+    <span class="k">if</span> <span class="n">min_idx</span> <span class="o">&lt;</span> <span class="o">-</span><span class="n">n</span> <span class="ow">or</span> <span class="n">max_idx</span> <span class="o">&gt;</span> <span class="n">n</span><span class="p">:</span>
+        <span class="n">oob</span> <span class="o">=</span> <span class="n">min_idx</span> <span class="k">if</span> <span class="n">min_idx</span> <span class="o">&lt;</span> <span class="o">-</span><span class="n">n</span> <span class="k">else</span> <span class="n">max_idx</span>
+        <span class="k">raise</span> <span class="ne">IndexError</span><span class="p">(</span>
+            <span class="sa">f</span><span class="s2">"index </span><span class="si">{</span><span class="n">oob</span><span class="si">}</span><span class="s2"> is out of bounds for axis </span><span class="si">{</span><span class="n">axis</span><span class="si">}</span><span class="s2"> with size </span><span class="si">{</span><span class="n">n</span><span class="si">}</span><span class="s2">"</span>
+        <span class="p">)</span>
+
+
 <span class="k">def</span><span class="w"> </span><span class="nf">_insert_array_indices</span><span class="p">(</span><span class="n">parameters</span><span class="p">,</span> <span class="n">indices</span><span class="p">,</span> <span class="n">values</span><span class="p">,</span> <span class="n">obj</span><span class="p">):</span>
 <span class="w">    </span><span class="sd">"""</span>
 <span class="sd">    Utility function for ``dpnp.insert`` when indices is an array with</span>
@@ -3556,6 +3589,13 @@ <h1>Source code for dpnp.dpnp_iface_manipulation</h1><div class="highlight"><pre
 <span class="sd">        does not occur in-place: a new array is returned. If</span>
 <span class="sd">        `axis` is ``None``, `out` is a flattened array.</span>
 
+<span class="sd">    Warnings</span>
+<span class="sd">    --------</span>
+<span class="sd">    This function might synchronize in order to validate that the indices are</span>
+<span class="sd">    within bounds. This may harm performance in some applications. To avoid</span>
+<span class="sd">    synchronization, pass `obj` as a Python scalar or sequence, or as a NumPy</span>
+<span class="sd">    array.</span>
+
 <span class="sd">    See Also</span>
 <span class="sd">    --------</span>
 <span class="sd">    :obj:`dpnp.append` : Append elements at the end of an array.</span>
@@ -3643,8 +3683,10 @@ <h1>Source code for dpnp.dpnp_iface_manipulation</h1><div class="highlight"><pre
             <span class="p">)</span>
 
     <span class="k">if</span> <span class="n">indices</span><span class="o">.</span><span class="n">size</span> <span class="o">==</span> <span class="mi">1</span><span class="p">:</span>
+        <span class="c1"># the size-1 path validates the bounds itself while reading the index</span>
         <span class="k">return</span> <span class="n">_insert_singleton_index</span><span class="p">(</span><span class="n">params</span><span class="p">,</span> <span class="n">indices</span><span class="p">,</span> <span class="n">values</span><span class="p">,</span> <span class="n">obj</span><span class="p">)</span>
 
+    <span class="n">_check_index_bounds</span><span class="p">(</span><span class="n">obj</span><span class="p">,</span> <span class="n">indices</span><span class="p">,</span> <span class="n">params</span><span class="o">.</span><span class="n">n</span><span class="p">,</span> <span class="n">params</span><span class="o">.</span><span class="n">axis</span><span class="p">)</span>
     <span class="k">return</span> <span class="n">_insert_array_indices</span><span class="p">(</span><span class="n">params</span><span class="p">,</span> <span class="n">indices</span><span class="p">,</span> <span class="n">values</span><span class="p">,</span> <span class="n">obj</span><span class="p">)</span></div>
 
 

reference/comparison.html

@@ -1055,7 +1055,7 @@
 <h1>Comparison Table NumPy/ DPNP/ CuPy<a class="headerlink" href="#comparison-table-numpy-dpnp-cupy" title="Link to this heading">¶</a></h1>
 <p>Here is a list of NumPy and CuPy APIs and its corresponding DPNP implementations.</p>
 <p><code class="docutils literal notranslate"><span class="pre">-</span></code> in DPNP column means that DPNP implementation is not provided yet.</p>
-<p><strong>NumPy(v2.5.2) / DPNP(v0.21.0dev7+6.gc578ddcbbf8) / CuPy(v14.2.0) APIs</strong></p>
+<p><strong>NumPy(v2.5.2) / DPNP(v0.21.0dev7+7.g1d7c6fa4e43) / CuPy(v14.2.0) APIs</strong></p>
 <section id="summary">
 <h2>Summary<a class="headerlink" href="#summary" title="Link to this heading">¶</a></h2>
 <div class="table-wrapper docutils container">

reference/generated/dpnp.insert.html

@@ -1086,6 +1086,13 @@ <h1>dpnp.insert<a class="headerlink" href="#dpnp-insert" title="Link to this hea
 <dd class="field-odd"><p><a class="reference internal" href="dpnp.ndarray.html#dpnp.ndarray" title="dpnp.ndarray">dpnp.ndarray</a></p>
 </dd>
 </dl>
+<div class="admonition warning">
+<p class="admonition-title">Warning</p>
+<p>This function might synchronize in order to validate that the indices are
+within bounds. This may harm performance in some applications. To avoid
+synchronization, pass <cite>obj</cite> as a Python scalar or sequence, or as a NumPy
+array.</p>
+</div>
 <div class="admonition seealso">
 <p class="admonition-title">See also</p>
 <dl class="simple">