This is an automated email from the ASF dual-hosted git repository.
junrushao pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm-ffi.git
The following commit(s) were added to refs/heads/main by this push:
new bd12b26 feat: add find and substr methods to String (#383)
bd12b26 is described below
commit bd12b26ac36ae6e770d128710fba13108957ee52
Author: Guan-Ming (Wesley) Chiu <[email protected]>
AuthorDate: Wed Jan 7 00:50:35 2026 +0800
feat: add find and substr methods to String (#383)
## Why
String class lacks essential string manipulation methods that are
commonly expected in standard library equivalents, requiring users to
convert to std::string for basic operations like substring search and
extraction.
## How
- Add find() method with three overloads (String, const char*, const
char* with count) to search for substrings
- Add substr() method to extract substrings with bounds checking
---
include/tvm/ffi/string.h | 44 ++++++++++++++++++++++++++++++++++++++++++++
tests/cpp/test_string.cc | 37 +++++++++++++++++++++++++++++++++++++
tests/python/test_string.py | 18 ++++++++++++++++++
3 files changed, 99 insertions(+)
diff --git a/include/tvm/ffi/string.h b/include/tvm/ffi/string.h
index b18e690..b93c3a9 100644
--- a/include/tvm/ffi/string.h
+++ b/include/tvm/ffi/string.h
@@ -611,6 +611,50 @@ class String {
}
}
+ /*! \brief Value returned by find() when no match is found */
+ static constexpr size_t npos = static_cast<size_t>(-1);
+
+ /*!
+ * \brief Find the first occurrence of a substring
+ * \param str The substring to search for
+ * \param pos The position at which to start the search
+ * \return The position of the first character of the first match, or npos
if not found
+ */
+ size_t find(const String& str, size_t pos = 0) const { return
find(str.data(), pos, str.size()); }
+
+ /*!
+ * \brief Find the first occurrence of a substring
+ * \param str The substring to search for
+ * \param pos The position at which to start the search
+ * \return The position of the first character of the first match, or npos
if not found
+ */
+ size_t find(const char* str, size_t pos = 0) const { return find(str, pos,
std::strlen(str)); }
+
+ /*!
+ * \brief Find the first occurrence of a substring
+ * \param str The substring to search for
+ * \param pos The position at which to start the search
+ * \param count The length of the substring
+ * \return The position of the first character of the first match, or npos
if not found
+ */
+ size_t find(const char* str, size_t pos, size_t count) const {
+ return std::string_view(data(), size()).find(std::string_view(str, count),
pos);
+ }
+
+ /*!
+ * \brief Returns a substring [pos, pos+count)
+ * \param pos The position of the first character to include
+ * \param count The length of the substring (default: until end of string)
+ * \return A string containing the substring
+ */
+ String substr(size_t pos = 0, size_t count = size_t(-1)) const {
+ if (pos > size()) {
+ throw std::out_of_range("tvm::String substr index out of bounds");
+ }
+ size_t rcount = std::min(count, size() - pos);
+ return String(data() + pos, rcount);
+ }
+
/*!
* \brief Convert String to an std::string object
*
diff --git a/tests/cpp/test_string.cc b/tests/cpp/test_string.cc
index 072c5cd..afcb651 100644
--- a/tests/cpp/test_string.cc
+++ b/tests/cpp/test_string.cc
@@ -445,4 +445,41 @@ TEST(String, StdHash) {
EXPECT_EQ(std::hash<Bytes>()(s3), std::hash<Bytes>()(s4));
}
+TEST(String, Find) {
+ String s{"hello world"};
+ EXPECT_EQ(s.find("world"), 6);
+ EXPECT_EQ(s.find("hello"), 0);
+ EXPECT_EQ(s.find("o"), 4);
+ EXPECT_EQ(s.find("o", 5), 7);
+ EXPECT_EQ(s.find("notfound"), String::npos);
+ EXPECT_EQ(s.find(""), 0);
+ EXPECT_EQ(s.find("", 5), 5);
+ EXPECT_EQ(s.find("", 11), 11);
+ EXPECT_EQ(s.find("", 20), String::npos);
+
+ String pattern{"world"};
+ EXPECT_EQ(s.find(pattern), 6);
+
+ String empty{""};
+ EXPECT_EQ(empty.find("x"), String::npos);
+ EXPECT_EQ(empty.find(""), 0);
+}
+
+TEST(String, Substr) {
+ String s{"hello world"};
+ EXPECT_EQ(s.substr(0, 5), "hello");
+ EXPECT_EQ(s.substr(6, 5), "world");
+ EXPECT_EQ(s.substr(6), "world");
+ EXPECT_EQ(s.substr(0), "hello world");
+ EXPECT_EQ(s.substr(11), "");
+ EXPECT_EQ(s.substr(0, 0), "");
+
+ EXPECT_THROW(s.substr(12), std::out_of_range);
+ EXPECT_THROW(s.substr(100), std::out_of_range);
+
+ String empty{""};
+ EXPECT_EQ(empty.substr(0), "");
+ EXPECT_THROW(empty.substr(1), std::out_of_range);
+}
+
} // namespace
diff --git a/tests/python/test_string.py b/tests/python/test_string.py
index 5dd06f4..499270b 100644
--- a/tests/python/test_string.py
+++ b/tests/python/test_string.py
@@ -53,3 +53,21 @@ def test_bytes() -> None:
b5 = pickle.loads(pickle.dumps(b))
assert b5 == b"hello"
assert isinstance(b5, tvm_ffi.core.Bytes)
+
+
+def test_string_find_substr() -> None:
+ s = tvm_ffi.core.String("hello world")
+ assert s.find("world") == 6
+ assert s.find("hello") == 0
+ assert s.find("o") == 4
+ assert s.find("o", 5) == 7
+ assert s.find("notfound") == -1
+ assert s.find("") == 0
+ assert s.find("", 5) == 5
+ assert s.find("", 11) == 11
+ assert s.find("", 20) == -1
+
+ assert s[6:11] == "world"
+ assert s[0:5] == "hello"
+ assert s[6:] == "world"
+ assert s[:5] == "hello"
