path: add path.glob

lib/path.js

@@ -31,6 +31,8 @@ const {
   StringPrototypeToLowerCase,
 } = primordials;
 
+const { glob } = internalBinding('path');
+
 const {
   CHAR_UPPERCASE_A,
   CHAR_LOWERCASE_A,
@@ -1064,6 +1066,7 @@ const win32 = {
 
     return ret;
   },
+  glob,
 
   sep: '\\',
   delimiter: ';',
@@ -1530,6 +1533,7 @@ const posix = {
 
     return ret;
   },
+  glob,
 
   sep: '/',
   delimiter: ':',

node.gyp

@@ -106,6 +106,7 @@
       'src/node_metadata.cc',
       'src/node_options.cc',
       'src/node_os.cc',
+      'src/node_path.cc',
       'src/node_perf.cc',
       'src/node_platform.cc',
       'src/node_postmortem_metadata.cc',
@@ -225,6 +226,7 @@
       'src/node_object_wrap.h',
       'src/node_options.h',
       'src/node_options-inl.h',
+      'src/node_path.h',
       'src/node_perf.h',
       'src/node_perf_common.h',
       'src/node_platform.h',

src/node_binding.cc

@@ -52,6 +52,7 @@
   V(mksnapshot)                                                                \
   V(options)                                                                   \
   V(os)                                                                        \
+  V(path)                                                                      \
   V(performance)                                                               \
   V(permission)                                                                \
   V(pipe_wrap)                                                                 \

src/node_path.cc

@@ -0,0 +1,120 @@
+#include "node_path.h"
+#include "env-inl.h"
+#include "node_errors.h"
+#include "node_external_reference.h"
+
+namespace node {
+
+namespace path {
+using v8::Context;
+using v8::FunctionCallbackInfo;
+using v8::Local;
+using v8::Object;
+using v8::Value;
+
+// extracted from
+// https://github.com/torvalds/linux/blob/cdc9718d5e590d6905361800b938b93f2b66818e/lib/glob.c
+bool glob(char const* pat, char const* str) {
+  /*
+   * Backtrack to previous * on mismatch and retry starting one
+   * character later in the string.  Because * matches all characters
+   * (no exception for /), it can be easily proved that there's
+   * never a need to backtrack multiple levels.
+   */
+  char const* back_pat = nullptr;
+  char const* back_str = nullptr;
+
+  /*
+   * Loop over each token (character or class) in pat, matching
+   * it against the remaining unmatched tail of str.  Return false
+   * on mismatch, or true after matching the trailing nul bytes.
+   */
+  for (;;) {
+    unsigned char c = *str++;
+    unsigned char d = *pat++;
+
+    switch (d) {
+      case '?': /* Wildcard: anything but nul */
+        if (c == '\0') return false;
+        break;
+      case '*':           /* Any-length wildcard */
+        if (*pat == '\0') /* Optimize trailing * case */
+          return true;
+        back_pat = pat;
+        back_str = --str; /* Allow zero-length match */
+        break;
+      case '[': { /* Character class */
+        bool match = false, inverted = (*pat == '!');
+        char const* cls = pat + inverted;
+        unsigned char a = *cls++;
+
+        /*
+         * Iterate over each span in the character class.
+         * A span is either a single character a, or a
+         * range a-b.  The first span may begin with ']'.
+         */
+        do {
+          unsigned char b = a;
+
+          if (a == '\0') /* Malformed */
+            goto literal;
+
+          if (cls[0] == '-' && cls[1] != ']') {
+            b = cls[1];
+
+            if (b == '\0') goto literal;
+
+            cls += 2;
+            /* Any special action if a > b? */
+          }
+          match |= (a <= c && c <= b);
+        } while ((a = *cls++) != ']');
+
+        if (match == inverted) goto backtrack;
+        pat = cls;
+      } break;
+      case '\\':
+        d = *pat++;
+        [[fallthrough]];
+      default: /* Literal character */
+      literal:
+        if (c == d) {
+          if (d == '\0') return true;
+          break;
+        }
+      backtrack:
+        if (c == '\0' || !back_pat) return false; /* No point continuing */
+        /* Try again from last *, one character later in str. */
+        pat = back_pat;
+        str = ++back_str;
+        break;
+    }
+  }
+}
+void glob(const FunctionCallbackInfo<Value>& args) {
+  Environment* env = Environment::GetCurrent(args);
+  CHECK_GE(args.Length(), 2);
+  CHECK(args[0]->IsString());
+  CHECK(args[1]->IsString());
+
+  std::string pattern = Utf8Value(env->isolate(), args[0]).ToString();
+  std::string str = Utf8Value(env->isolate(), args[1]).ToString();
+  args.GetReturnValue().Set(glob(pattern.c_str(), str.c_str()));
+}
+
+void Initialize(Local<Object> target,
+                Local<Value> unused,
+                Local<Context> context,
+                void* priv) {
+  SetMethod(context, target, "glob", glob);
+}
+
+void RegisterExternalReferences(ExternalReferenceRegistry* registry) {
+  registry->Register(glob);
+}
+}  // namespace path
+
+}  // namespace node
+
+NODE_BINDING_CONTEXT_AWARE_INTERNAL(path, node::path::Initialize)
+NODE_BINDING_EXTERNAL_REFERENCE(path, node::path::RegisterExternalReferences)

src/node_path.h

@@ -0,0 +1,17 @@
+#ifndef SRC_NODE_PATH_H_
+#define SRC_NODE_PATH_H_
+
+#if defined(NODE_WANT_INTERNALS) && NODE_WANT_INTERNALS
+
+#include "base_object.h"
+#include "node_snapshotable.h"
+#include "v8.h"
+
+namespace node {
+
+namespace path {}  // namespace path
+}  // namespace node
+
+#endif  // defined(NODE_WANT_INTERNALS) && NODE_WANT_INTERNALS
+
+#endif  // SRC_NODE_PATH_H_

test/parallel/test-path-glob.mjs

@@ -0,0 +1,88 @@
+import '../common/index.mjs';
+import { describe, it } from 'node:test';
+import * as assert from 'node:assert';
+import * as path from 'node:path';
+
+
+// https://github.com/torvalds/linux/blob/cdc9718d5e590d6905361800b938b93f2b66818e/lib/globtest.c
+const patterns = [
+  { expected: true, pattern: 'a', name: 'a' },
+  { expected: false, pattern: 'a', name: 'b' },
+  { expected: false, pattern: 'a', name: 'aa' },
+  { expected: false, pattern: 'a', name: '' },
+  { expected: true, pattern: '', name: '' },
+  { expected: false, pattern: '', name: 'a' },
+  /* Simple character class tests */
+  { expected: true, pattern: '[a]', name: 'a' },
+  { expected: false, pattern: '[a]', name: 'b' },
+  { expected: false, pattern: '[!a]', name: 'a' },
+  { expected: true, pattern: '[!a]', name: 'b' },
+  { expected: true, pattern: '[ab]', name: 'a' },
+  { expected: true, pattern: '[ab]', name: 'b' },
+  { expected: false, pattern: '[ab]', name: 'c' },
+  { expected: true, pattern: '[!ab]', name: 'c' },
+  { expected: true, pattern: '[a-c]', name: 'b' },
+  { expected: false, pattern: '[a-c]', name: 'd' },
+  /* Corner cases in character class parsing */
+  { expected: true, pattern: '[a-c-e-g]', name: '-' },
+  { expected: false, pattern: '[a-c-e-g]', name: 'd' },
+  { expected: true, pattern: '[a-c-e-g]', name: 'f' },
+  { expected: true, pattern: '[]a-ceg-ik[]', name: 'a' },
+  { expected: true, pattern: '[]a-ceg-ik[]', name: ']' },
+  { expected: true, pattern: '[]a-ceg-ik[]', name: '[' },
+  { expected: true, pattern: '[]a-ceg-ik[]', name: 'h' },
+  { expected: false, pattern: '[]a-ceg-ik[]', name: 'f' },
+  { expected: false, pattern: '[!]a-ceg-ik[]', name: 'h' },
+  { expected: false, pattern: '[!]a-ceg-ik[]', name: ']' },
+  { expected: true, pattern: '[!]a-ceg-ik[]', name: 'f' },
+  /* Simple wild cards */
+  { expected: true, pattern: '?', name: 'a' },
+  { expected: false, pattern: '?', name: 'aa' },
+  { expected: false, pattern: '??', name: 'a' },
+  { expected: true, pattern: '?x?', name: 'axb' },
+  { expected: false, pattern: '?x?', name: 'abx' },
+  { expected: false, pattern: '?x?', name: 'xab' },
+  /* Asterisk wild cards (backtracking) */
+  { expected: false, pattern: '*??', name: 'a' },
+  { expected: true, pattern: '*??', name: 'ab' },
+  { expected: true, pattern: '*??', name: 'abc' },
+  { expected: true, pattern: '*??', name: 'abcd' },
+  { expected: false, pattern: '??*', name: 'a' },
+  { expected: true, pattern: '??*', name: 'ab' },
+  { expected: true, pattern: '??*', name: 'abc' },
+  { expected: true, pattern: '??*', name: 'abcd' },
+  { expected: false, pattern: '?*?', name: 'a' },
+  { expected: true, pattern: '?*?', name: 'ab' },
+  { expected: true, pattern: '?*?', name: 'abc' },
+  { expected: true, pattern: '?*?', name: 'abcd' },
+  { expected: true, pattern: '*b', name: 'b' },
+  { expected: true, pattern: '*b', name: 'ab' },
+  { expected: false, pattern: '*b', name: 'ba' },
+  { expected: true, pattern: '*b', name: 'bb' },
+  { expected: true, pattern: '*b', name: 'abb' },
+  { expected: true, pattern: '*b', name: 'bab' },
+  { expected: true, pattern: '*bc', name: 'abbc' },
+  { expected: true, pattern: '*bc', name: 'bc' },
+  { expected: true, pattern: '*bc', name: 'bbc' },
+  { expected: true, pattern: '*bc', name: 'bcbc' },
+  /* Multiple asterisks (complex backtracking) */
+  { expected: true, pattern: '*ac*', name: 'abacadaeafag' },
+  { expected: true, pattern: '*ac*ae*ag*', name: 'abacadaeafag' },
+  { expected: true, pattern: '*a*b*[bc]*[ef]*g*', name: 'abacadaeafag' },
+  { expected: false, pattern: '*a*b*[ef]*[cd]*g*', name: 'abacadaeafag' },
+  { expected: true, pattern: '*abcd*', name: 'abcabcabcabcdefg' },
+  { expected: true, pattern: '*ab*cd*', name: 'abcabcabcabcdefg' },
+  { expected: true, pattern: '*abcd*abcdef*', name: 'abcabcdabcdeabcdefg' },
+  { expected: false, pattern: '*abcd*', name: 'abcabcabcabcefg' },
+  { expected: false, pattern: '*ab*cd*', name: 'abcabcabcabcefg' },
+];
+
+describe('path.glob', () => {
+  for (const { expected, pattern, name } of patterns) {
+    it(`pattern "${pattern}" should ${expected ? '' : 'not '}match "${name}"`, () => {
+      assert.strictEqual(path.glob(pattern, name), expected);
+    });
+  }
+
+  // TODO: error handling, invalid parameters
+});