ViniciusSouzaRoque commented on code in PR #12716:
URL: https://github.com/apache/arrow/pull/12716#discussion_r850331184
##########
cpp/src/gandiva/precompiled/string_ops.cc:
##########
@@ -2815,43 +2814,62 @@ const char* soundex_utf8(gdv_int64 ctx, const char* in,
gdv_int32 in_len,
}
// The soundex code is composed by one letter and three numbers
+ char* soundex = reinterpret_cast<char*>(gdv_fn_context_arena_malloc(ctx,
in_len));
char* ret = reinterpret_cast<char*>(gdv_fn_context_arena_malloc(ctx, 4));
- if (ret == nullptr) {
+
+ if (soundex == nullptr || ret == nullptr) {
gdv_fn_context_set_error_msg(ctx, "Could not allocate memory for output
string");
*out_len = 0;
return "";
}
int si = 1;
+ int ret_len = 1;
unsigned char c;
int start_idx = 0;
for (int i = 0; i < in_len; ++i) {
if (isalpha(in[i]) > 0) {
+ // Retain the first letter
ret[0] = toupper(in[i]);
start_idx = i + 1;
break;
}
}
- for (int i = start_idx, l = in_len; i < l; i++) {
+ soundex[0] = '\0';
+ // Replace consonants with digits and special letters with 0
+ for (int i = start_idx; i < in_len; i++) {
if (isalpha(in[i]) > 0) {
c = toupper(in[i]) - 65;
- if (mappings[c] != '0') {
- if (mappings[c] != ret[si - 1]) {
- ret[si] = mappings[c];
- si++;
- }
-
- if (si > 3) break;
+ if (mappings[c] != soundex[si - 1]) {
+ soundex[si] = mappings[c];
+ si++;
}
}
}
- if (si <= 3) {
- while (si <= 3) {
- ret[si] = '0';
- si++;
+ int i = 1;
+ // If the saved letter's digit is the same as the resulting first digit,
remove the
+ // digit.
+ if (soundex[i] == mappings[ret[0] - 65]) {
Review Comment:
> You have to handle the case where ret[0] is not initialised.. for e.g. the
input is '123456' - since there are no alphabets, ret[0] will not be initialised
>
> Similarly, you have to handle the case where soundex[1] is not set..
About the soundex[1];
If ret[0] is valid, we don't need handle with this case
##########
cpp/src/gandiva/precompiled/string_ops_test.cc:
##########
@@ -2310,12 +2310,49 @@ TEST(TestStringOps, TestFromHex) {
::testing::HasSubstr("Error parsing hex string, one or more bytes are
not valid."));
ctx.Reset();
}
+
TEST(TestStringOps, TestSoundex) {
gandiva::ExecutionContext ctx;
auto ctx_ptr = reinterpret_cast<int64_t>(&ctx);
int32_t out_len = 0;
const char* out;
+ out = soundex_utf8(ctx_ptr, "robert", 6, &out_len);
Review Comment:
Done
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
