duckdb 0.6.2-dev1978.0 → 0.6.2-dev2015.0

package/src/duckdb/extension/json/yyjson/include/yyjson.hpp

@@ -22,9 +22,8 @@
 #include <limits.h>
 #include <string.h>
 #include <float.h>
-#include "duckdb/common/fast_mem.hpp"
-
 
+#include "duckdb/common/fast_mem.hpp"
 
 /*==============================================================================
  * Compile-time Options
@@ -229,7 +228,8 @@
 
 /** likely for compiler */
 #ifndef yyjson_likely
-#   if yyjson_has_builtin(__builtin_expect) || (YYJSON_GCC_VER >= 4 && YYJSON_GCC_VER != 5)
+#   if yyjson_has_builtin(__builtin_expect) || \
+    (YYJSON_GCC_VER >= 4 && YYJSON_GCC_VER != 5)
 #       define yyjson_likely(expr) __builtin_expect(!!(expr), 1)
 #   else
 #       define yyjson_likely(expr) (expr)
@@ -238,7 +238,8 @@
 
 /** unlikely for compiler */
 #ifndef yyjson_unlikely
-#   if yyjson_has_builtin(__builtin_expect) || (YYJSON_GCC_VER >= 4 && YYJSON_GCC_VER != 5)
+#   if yyjson_has_builtin(__builtin_expect) || \
+    (YYJSON_GCC_VER >= 4 && YYJSON_GCC_VER != 5)
 #       define yyjson_unlikely(expr) __builtin_expect(!!(expr), 0)
 #   else
 #       define yyjson_unlikely(expr) (expr)
@@ -620,7 +621,7 @@ typedef uint32_t yyjson_read_code;
 /** Success, no error. */
 static const yyjson_read_code YYJSON_READ_SUCCESS                       = 0;
 
-/** Invalid parameter, such as NULL string or invalid file path. */
+/** Invalid parameter, such as NULL input string or 0 input length. */
 static const yyjson_read_code YYJSON_READ_ERROR_INVALID_PARAMETER       = 1;
 
 /** Memory allocation failure occurs. */
@@ -816,6 +817,9 @@ yyjson_api_inline size_t yyjson_read_max_memory_usage(size_t len,
  @param flg The JSON read options.
     Multiple options can be combined with `|` operator. 0 means no options.
     Suppors `YYJSON_READ_NUMBER_AS_RAW` and `YYJSON_READ_ALLOW_INF_AND_NAN`.
+ @param alc The memory allocator used for long number.
+    It is only used when the built-in floating point reader is disabled.
+    Pass NULL to use the libc's default allocator.
  @param err A pointer to receive error information.
     Pass NULL if you don't need error information.
  @return If successful, a pointer to the character after the last character
@@ -824,6 +828,7 @@ yyjson_api_inline size_t yyjson_read_max_memory_usage(size_t len,
 yyjson_api const char *yyjson_read_number(const char *dat,
                                           yyjson_val *val,
                                           yyjson_read_flag flg,
+                                          const yyjson_alc *alc,
                                           yyjson_read_err *err);
 
 /**
@@ -839,6 +844,9 @@ yyjson_api const char *yyjson_read_number(const char *dat,
  @param flg The JSON read options.
     Multiple options can be combined with `|` operator. 0 means no options.
     Suppors `YYJSON_READ_NUMBER_AS_RAW` and `YYJSON_READ_ALLOW_INF_AND_NAN`.
+ @param alc The memory allocator used for long number.
+    It is only used when the built-in floating point reader is disabled.
+    Pass NULL to use the libc's default allocator.
  @param err A pointer to receive error information.
     Pass NULL if you don't need error information.
  @return If successful, a pointer to the character after the last character
@@ -847,8 +855,9 @@ yyjson_api const char *yyjson_read_number(const char *dat,
 yyjson_api_inline const char *yyjson_mut_read_number(const char *dat,
                                                      yyjson_mut_val *val,
                                                      yyjson_read_flag flg,
+                                                     const yyjson_alc *alc,
                                                      yyjson_read_err *err) {
-    return yyjson_read_number(dat, (yyjson_val *)val, flg, err);
+    return yyjson_read_number(dat, (yyjson_val *)val, flg, alc, err);
 }
 
 
@@ -4230,14 +4239,14 @@ yyjson_api_inline char *unsafe_yyjson_mut_strncpy(yyjson_mut_doc *doc,
     char *mem;
     const yyjson_alc *alc = &doc->alc;
     yyjson_str_pool *pool = &doc->str_pool;
-    
+
     if (!str) return NULL;
     if (yyjson_unlikely((size_t)(pool->end - pool->cur) <= len)) {
         if (yyjson_unlikely(!unsafe_yyjson_str_pool_grow(pool, alc, len + 1))) {
             return NULL;
         }
     }
-    
+
     mem = pool->cur;
     pool->cur = mem + len + 1;
     memcpy((void *)mem, (const void *)str, len);
@@ -4255,7 +4264,7 @@ yyjson_api_inline yyjson_mut_val *unsafe_yyjson_mut_val(yyjson_mut_doc *doc,
             return NULL;
         }
     }
-    
+
     val = pool->cur;
     pool->cur += count;
     return val;

package/src/duckdb/extension/json/yyjson/yyjson.cpp

@@ -3041,7 +3041,7 @@ static_noinline bool read_number_raw(u8 **ptr,
         if (unlikely(ext)) {
             if (read_inf_or_nan(*hdr == '-', &cur, pre, val)) return_raw();
         }
-        return_err(cur - 1, "no digit after minus sign");
+        return_err(cur, "no digit after minus sign");
     }
     
     /* read integral part */
@@ -3069,7 +3069,7 @@ static_noinline bool read_number_raw(u8 **ptr,
     if (digi_is_exp(*cur)) {
         cur += 1 + digi_is_sign(cur[1]);
         if (!digi_is_digit(*cur++)) {
-            return_err(cur - 1, "no digit after exponent sign");
+            return_err(cur, "no digit after exponent sign");
         }
         while (digi_is_digit(*cur)) cur++;
     }
@@ -3481,7 +3481,7 @@ static_inline bool read_number(u8 **ptr,
                     return true;
                 }
             }
-            return_err(cur - 1, "no digit after minus sign");
+            return_err(cur, "no digit after minus sign");
         }
         /* begin with 0 */
         if (likely(!digi_is_digit_or_fp(*++cur))) return_i64(0);
@@ -3504,7 +3504,7 @@ static_inline bool read_number(u8 **ptr,
         if (unlikely(digi_is_exp(*cur))) { /* 0 with any exponent is still 0 */
             cur += (usize)1 + digi_is_sign(cur[1]);
             if (unlikely(!digi_is_digit(*cur))) {
-                return_err(cur - 1, "no digit after exponent sign");
+                return_err(cur, "no digit after exponent sign");
             }
             while (digi_is_digit(*++cur));
         }
@@ -3658,7 +3658,7 @@ digi_exp_more:
     exp_sign = (*++cur == '-');
     cur += digi_is_sign(*cur);
     if (unlikely(!digi_is_digit(*cur))) {
-        return_err(cur - 1, "no digit after exponent sign");
+        return_err(cur, "no digit after exponent sign");
     }
     while (*cur == '0') cur++;
     
@@ -4056,7 +4056,7 @@ static_noinline bool read_number(u8 **ptr,
                 return true;
             }
         }
-        return_err(cur - 1, "no digit after minus sign");
+        return_err(cur, "no digit after minus sign");
     }
     if (*cur == '0') {
         cur++;
@@ -4105,17 +4105,19 @@ read_double:
         /* skip fraction part */
         dot = cur;
         cur++;
-        if (!digi_is_digit(*cur++)) {
-            return_err(cur - 1, "no digit after decimal point");
+        if (!digi_is_digit(*cur)) {
+            return_err(cur, "no digit after decimal point");
         }
+        cur++;
         while (digi_is_digit(*cur)) cur++;
     }
     if (digi_is_exp(*cur)) {
         /* skip exponent part */
         cur += 1 + digi_is_sign(cur[1]);
-        if (!digi_is_digit(*cur++)) {
-            return_err(cur - 1, "no digit after exponent sign");
+        if (!digi_is_digit(*cur)) {
+            return_err(cur, "no digit after exponent sign");
         }
+        cur++;
         while (digi_is_digit(*cur)) cur++;
     }
     
@@ -4896,6 +4898,7 @@ arr_val_begin:
         cur++;
         if (likely(ctn_len == 0)) goto arr_end;
         if (has_flag(ALLOW_TRAILING_COMMAS)) goto arr_end;
+        cur--;
         goto fail_trailing_comma;
     }
     if (char_is_space(*cur)) {
@@ -4973,6 +4976,7 @@ obj_key_begin:
         cur++;
         if (likely(ctn_len == 0)) goto obj_end;
         if (has_flag(ALLOW_TRAILING_COMMAS)) goto obj_end;
+        cur--;
         goto fail_trailing_comma;
     }
     if (char_is_space(*cur)) {
@@ -5095,7 +5099,10 @@ obj_end:
 doc_end:
     /* check invalid contents after json document */
     if (unlikely(cur < end) && !has_flag(STOP_WHEN_DONE)) {
-        if (has_flag(ALLOW_COMMENTS)) skip_spaces_and_comments(&cur);
+        if (has_flag(ALLOW_COMMENTS)) {
+            skip_spaces_and_comments(&cur);
+            if (byte_match_2(cur, "/*")) goto fail_comment;
+        }
         else while (char_is_space(*cur)) cur++;
         if (unlikely(cur < end)) goto fail_garbage;
     }
@@ -5296,6 +5303,7 @@ arr_val_begin:
         cur++;
         if (likely(ctn_len == 0)) goto arr_end;
         if (has_flag(ALLOW_TRAILING_COMMAS)) goto arr_end;
+        cur--;
         goto fail_trailing_comma;
     }
     if (char_is_space(*cur)) {
@@ -5390,6 +5398,7 @@ obj_key_begin:
         cur++;
         if (likely(ctn_len == 0)) goto obj_end;
         if (has_flag(ALLOW_TRAILING_COMMAS)) goto obj_end;
+        cur--;
         goto fail_trailing_comma;
     }
     if (char_is_space(*cur)) {
@@ -5521,7 +5530,10 @@ obj_end:
 doc_end:
     /* check invalid contents after json document */
     if (unlikely(cur < end) && !has_flag(STOP_WHEN_DONE)) {
-        if (has_flag(ALLOW_COMMENTS)) skip_spaces_and_comments(&cur);
+        if (has_flag(ALLOW_COMMENTS)) {
+            skip_spaces_and_comments(&cur);
+            if (byte_match_2(cur, "/*")) goto fail_comment;
+        }
         else while (char_is_space(*cur)) cur++;
         if (unlikely(cur < end)) goto fail_garbage;
     }
@@ -5777,6 +5789,7 @@ yyjson_doc *yyjson_read_file(const char *path,
 const char *yyjson_read_number(const char *dat,
                                yyjson_val *val,
                                yyjson_read_flag flg,
+                               const yyjson_alc *alc,
                                yyjson_read_err *err) {
 #define return_err(_pos, _code, _msg) do { \
     err->pos = _pos > hdr ? (usize)(_pos - hdr) : 0; \
@@ -5792,8 +5805,13 @@ const char *yyjson_read_number(const char *dat,
     u8 **pre; /* previous raw end pointer */
     const char *msg;
     yyjson_read_err dummy_err;
-    if (!err) err = &dummy_err;
     
+#if !YYJSON_HAS_IEEE_754 || YYJSON_DISABLE_FAST_FP_CONV
+    u8 buf[128];
+    usize dat_len;
+#endif
+    
+    if (!err) err = &dummy_err;
     if (unlikely(!dat)) {
         return_err(cur, INVALID_PARAMETER, "input data is NULL");
     }
@@ -5801,6 +5819,23 @@ const char *yyjson_read_number(const char *dat,
         return_err(cur, INVALID_PARAMETER, "output value is NULL");
     }
     
+#if !YYJSON_HAS_IEEE_754 || YYJSON_DISABLE_FAST_FP_CONV
+    if (!alc) alc = &YYJSON_DEFAULT_ALC;
+    dat_len = strlen(dat);
+    if (dat_len < sizeof(buf)) {
+        memcpy(buf, dat, dat_len + 1);
+        hdr = buf;
+        cur = hdr;
+    } else {
+        hdr = (u8 *)alc->malloc(alc->ctx, dat_len + 1);
+        cur = hdr;
+        if (unlikely(!hdr)) {
+            return_err(cur, MEMORY_ALLOCATION, "memory allocation failed");
+        }
+        memcpy(hdr, dat, dat_len + 1);
+    }
+#endif
+    
 #if YYJSON_DISABLE_NON_STANDARD
     ext = false;
 #else
@@ -5811,10 +5846,20 @@ const char *yyjson_read_number(const char *dat,
     raw_end = NULL;
     pre = raw ? &raw_end : NULL;
     
+#if !YYJSON_HAS_IEEE_754 || YYJSON_DISABLE_FAST_FP_CONV
+    if (!read_number(&cur, pre, ext, val, &msg)) {
+        if (dat_len >= sizeof(buf)) alc->free(alc->ctx, hdr);
+        return_err(cur, INVALID_NUMBER, msg);
+    }
+    if (dat_len >= sizeof(buf)) alc->free(alc->ctx, hdr);
+    if (raw) val->uni.str = dat;
+    return dat + (cur - hdr);
+#else
     if (!read_number(&cur, pre, ext, val, &msg)) {
         return_err(cur, INVALID_NUMBER, msg);
     }
     return (const char *)cur;
+#endif
     
 #undef return_err
 }

package/src/duckdb/src/function/table/copy_csv.cpp

@@ -1,15 +1,16 @@
-#include "duckdb/function/table/read_csv.hpp"
-#include "duckdb/execution/operator/persistent/parallel_csv_reader.hpp"
-#include "duckdb/common/serializer/buffered_serializer.hpp"
-#include "duckdb/function/copy_function.hpp"
-#include "duckdb/parser/parsed_data/copy_info.hpp"
 #include "duckdb/common/bind_helpers.hpp"
-#include "duckdb/common/string_util.hpp"
 #include "duckdb/common/file_system.hpp"
+#include "duckdb/common/serializer/buffered_serializer.hpp"
+#include "duckdb/common/string_util.hpp"
 #include "duckdb/common/types/string_type.hpp"
 #include "duckdb/common/vector_operations/vector_operations.hpp"
+#include "duckdb/execution/operator/persistent/parallel_csv_reader.hpp"
+#include "duckdb/function/copy_function.hpp"
 #include "duckdb/function/scalar/string_functions.hpp"
+#include "duckdb/function/table/read_csv.hpp"
 #include "duckdb/main/config.hpp"
+#include "duckdb/parser/parsed_data/copy_info.hpp"
+
 #include <limits>
 
 namespace duckdb {
@@ -164,12 +165,14 @@ static bool RequiresQuotes(WriteCSVData &csv_data, const char *str, idx_t len) {
 		}
 
 		// check for delimiter
-		if (ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.delimiter.c_str(),
+		if (options.delimiter.length() != 0 &&
+		    ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.delimiter.c_str(),
 		                      options.delimiter.size()) != DConstants::INVALID_INDEX) {
 			return true;
 		}
 		// check for quote
-		if (ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.quote.c_str(),
+		if (options.quote.length() != 0 &&
+		    ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.quote.c_str(),
 		                      options.quote.size()) != DConstants::INVALID_INDEX) {
 			return true;
 		}
@@ -196,13 +199,15 @@ static void WriteQuotedString(Serializer &serializer, WriteCSVData &csv_data, co
 					break;
 				}
 			}
-		} else if (options.quote.length() != 0) {
+		} else {
 			// complex CSV
 			// check for quote or escape separately
-			if (ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.quote.c_str(),
+			if (options.quote.length() != 0 &&
+			    ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.quote.c_str(),
 			                      options.quote.size()) != DConstants::INVALID_INDEX) {
 				requires_escape = true;
-			} else if (ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.escape.c_str(),
+			} else if (options.escape.length() != 0 &&
+			           ContainsFun::Find((const unsigned char *)str, len, (const unsigned char *)options.escape.c_str(),
 			                             options.escape.size()) != DConstants::INVALID_INDEX) {
 				requires_escape = true;
 			}

package/src/duckdb/src/function/table/version/pragma_version.cpp

@@ -1,8 +1,8 @@
 #ifndef DUCKDB_VERSION
-#define DUCKDB_VERSION "0.6.2-dev1978"
+#define DUCKDB_VERSION "0.6.2-dev2015"
 #endif
 #ifndef DUCKDB_SOURCE_ID
-#define DUCKDB_SOURCE_ID "82a82efbbc"
+#define DUCKDB_SOURCE_ID "cbf43430b3"
 #endif
 #include "duckdb/function/table/system_functions.hpp"
 #include "duckdb/main/database.hpp"

package/src/duckdb/src/include/duckdb/function/scalar/strftime.hpp

@@ -138,8 +138,8 @@ public:
 
 	DUCKDB_API bool Parse(string_t str, ParseResult &result);
 
-	bool TryParseDate(string_t str, date_t &result, string &error_message);
-	bool TryParseTimestamp(string_t str, timestamp_t &result, string &error_message);
+	DUCKDB_API bool TryParseDate(string_t str, date_t &result, string &error_message);
+	DUCKDB_API bool TryParseTimestamp(string_t str, timestamp_t &result, string &error_message);
 
 	date_t ParseDate(string_t str);
 	timestamp_t ParseTimestamp(string_t str);

package/src/duckdb/src/include/duckdb/main/extension_functions.hpp

@@ -46,6 +46,7 @@ static constexpr ExtensionFunction EXTENSION_FUNCTIONS[] = {
     {"json_group_array", "json"},
     {"json_group_object", "json"},
     {"json_group_structure", "json"},
+    {"json_keys", "json"},
     {"json_merge_patch", "json"},
     {"json_object", "json"},
     {"json_quote", "json"},
@@ -62,7 +63,11 @@ static constexpr ExtensionFunction EXTENSION_FUNCTIONS[] = {
     {"postgres_attach", "postgres_scanner"},
     {"postgres_scan", "postgres_scanner"},
     {"postgres_scan_pushdown", "postgres_scanner"},
+    {"read_json", "json"},
+    {"read_json_auto", "json"},
     {"read_json_objects", "json"},
+    {"read_ndjson", "json"},
+    {"read_ndjson_auto", "json"},
     {"read_ndjson_objects", "json"},
     {"read_parquet", "parquet"},
     {"row_to_json", "json"},

package/src/duckdb/ub_extension_json_json_functions.cpp

@@ -4,6 +4,8 @@
 
 #include "extension/json/json_functions/json_extract.cpp"
 
+#include "extension/json/json_functions/json_keys.cpp"
+
 #include "extension/json/json_functions/json_merge_patch.cpp"
 
 #include "extension/json/json_functions/json_structure.cpp"
@@ -16,5 +18,7 @@
 
 #include "extension/json/json_functions/json_valid.cpp"
 
+#include "extension/json/json_functions/read_json.cpp"
+
 #include "extension/json/json_functions/read_json_objects.cpp"