From 892f5574b251773a3cb8f14b5e18902bf6871200 Mon Sep 17 00:00:00 2001 From: Florents Tselai Date: Mon, 13 Apr 2026 12:07:46 +0300 Subject: [PATCH v1 3/3] Add jsonpath $.join(sep) method --- doc/src/sgml/func/func-json.sgml | 18 +++ src/backend/utils/adt/jsonpath.c | 22 ++++ src/backend/utils/adt/jsonpath_exec.c | 111 ++++++++++++++++++ src/backend/utils/adt/jsonpath_gram.y | 7 +- src/backend/utils/adt/jsonpath_scan.l | 1 + src/include/utils/jsonpath.h | 1 + src/test/regress/expected/jsonb_jsonpath.out | 80 +++++++++++++ src/test/regress/expected/jsonpath.out | 32 +++++ .../regress/expected/sqljson_queryfuncs.out | 1 + src/test/regress/sql/jsonb_jsonpath.sql | 28 +++++ src/test/regress/sql/jsonpath.sql | 7 ++ src/test/regress/sql/sqljson_queryfuncs.sql | 1 + 12 files changed, 308 insertions(+), 1 deletion(-) diff --git a/doc/src/sgml/func/func-json.sgml b/doc/src/sgml/func/func-json.sgml index 7aad2d69a88..715bdf62578 100644 --- a/doc/src/sgml/func/func-json.sgml +++ b/doc/src/sgml/func/func-json.sgml @@ -2959,6 +2959,24 @@ ERROR: jsonpath member accessor can only be applied to an object + + + array . join(separator [, null_string]) + string + + + Concatenates the elements of an array into a single string using the + specified separator. If the optional null_string + argument is provided, it replaces JSON null values; + otherwise, null values are skipped. The input must + be an array consisting only of strings or null values. + + + jsonb_path_query('["a", null, "c"]', '$.join("-", "N/A")') + "a-N/A-c" + + + diff --git a/src/backend/utils/adt/jsonpath.c b/src/backend/utils/adt/jsonpath.c index d14fe450b6b..b501f197964 100644 --- a/src/backend/utils/adt/jsonpath.c +++ b/src/backend/utils/adt/jsonpath.c @@ -330,6 +330,7 @@ flattenJsonPathParseItem(StringInfo buf, int *result, struct Node *escontext, } break; case jpiStrSplit: + case jpiStrJoin: { /* Reserve space for left and right arg positions */ int32 left = reserveSpaceForItemPointer(buf); @@ -946,6 +947,20 @@ printJsonPathItem(StringInfo buf, JsonPathItem *v, bool inKey, printJsonPathItem(buf, &elem, false, false); appendStringInfoChar(buf, ')'); break; + case jpiStrJoin: + appendStringInfoString(buf, ".join("); + jspGetLeftArg(v, &elem); + printJsonPathItem(buf, &elem, false, false); + + /* Check if null_string was provided. */ + if (v->content.args.right != 0) + { + appendStringInfoString(buf, ", "); + jspGetRightArg(v, &elem); + printJsonPathItem(buf, &elem, false, false); + } + appendStringInfoChar(buf, ')'); + break; default: elog(ERROR, "unrecognized jsonpath item type: %d", v->type); } @@ -1049,6 +1064,8 @@ jspOperationName(JsonPathItemType type) return "translate"; case jpiStrSplit: return "split"; + case jpiStrJoin: + return "join"; default: elog(ERROR, "unrecognized jsonpath item type: %d", type); return NULL; @@ -1183,6 +1200,7 @@ jspInitByBuffer(JsonPathItem *v, char *base, int32 pos) case jpiStrTranslate: case jpiStrSplit: case jpiStrSplitPart: + case jpiStrJoin: read_int32(v->content.args.left, base, pos); read_int32(v->content.args.right, base, pos); break; @@ -1310,6 +1328,7 @@ jspGetNext(JsonPathItem *v, JsonPathItem *a) v->type == jpiStrInitcap || v->type == jpiStrSplitPart || v->type == jpiStrSplit || + v->type == jpiStrJoin || v->type == jpiStrTranslate); if (a) @@ -1341,6 +1360,7 @@ jspGetLeftArg(JsonPathItem *v, JsonPathItem *a) v->type == jpiStrReplace || v->type == jpiStrTranslate || v->type == jpiStrSplit || + v->type == jpiStrJoin || v->type == jpiStrSplitPart); jspInitByBuffer(a, v->base, v->content.args.left); @@ -1367,6 +1387,7 @@ jspGetRightArg(JsonPathItem *v, JsonPathItem *a) v->type == jpiStrReplace || v->type == jpiStrTranslate || v->type == jpiStrSplitPart || + v->type == jpiStrJoin || v->type == jpiStrSplit); jspInitByBuffer(a, v->base, v->content.args.right); @@ -1677,6 +1698,7 @@ jspIsMutableWalker(JsonPathItem *jpi, struct JsonPathMutableContext *cxt) case jpiStrSplitPart: case jpiStrSplit: case jpiStrTranslate: + case jpiStrJoin: status = jpdsNonDateTime; break; diff --git a/src/backend/utils/adt/jsonpath_exec.c b/src/backend/utils/adt/jsonpath_exec.c index 93f31b40b0b..cb8bb1aac25 100644 --- a/src/backend/utils/adt/jsonpath_exec.c +++ b/src/backend/utils/adt/jsonpath_exec.c @@ -1699,6 +1699,117 @@ executeItemOptUnwrapTarget(JsonPathExecContext *cxt, JsonPathItem *jsp, return executeStringInternalMethod(cxt, jsp, jb, found); } break; + case jpiStrJoin: + { + JsonPathItem next_elem; + JsonbValue jbv_res; + char *sep; + char *null_replace = NULL; + StringInfoData buf; + bool first = true; + bool hasNext; + + jspGetLeftArg(jsp, &elem); + sep = jspGetString(&elem, NULL); + + if (jsp->content.args.right != 0) + { + jspGetRightArg(jsp, &elem); + null_replace = jspGetString(&elem, NULL); + } + + /* Validate target is an array */ + if (JsonbType(jb) != jbvArray) + { + RETURN_ERROR(ereport(ERROR, + (errcode(ERRCODE_INVALID_PARAMETER_VALUE), + errmsg("jsonpath item method .join() can only be applied to an array")))); + } + + initStringInfo(&buf); + + /* Process the array elements */ + if (jb->type == jbvBinary) + { + /* Serialized Binary Blob */ + JsonbIterator *it; + JsonbValue v; + JsonbIteratorToken tok; + + it = JsonbIteratorInit(jb->val.binary.data); + while ((tok = JsonbIteratorNext(&it, &v, true)) != WJB_DONE) + { + if (tok != WJB_ELEM) + continue; + + if (v.type == jbvString) + { + if (!first) + appendStringInfoString(&buf, sep); + appendBinaryStringInfo(&buf, v.val.string.val, v.val.string.len); + first = false; + } + else if (v.type == jbvNull) + { + if (null_replace) + { + if (!first) + appendStringInfoString(&buf, sep); + appendStringInfoString(&buf, null_replace); + first = false; + } + } + else + { + RETURN_ERROR(ereport(ERROR, + (errcode(ERRCODE_INVALID_PARAMETER_VALUE), + errmsg("jsonpath .join() array elements must be strings or nulls")))); + } + } + } + else + { + /* Recursive Tree (jbvArray) */ + for (int i = 0; i < jb->val.array.nElems; i++) + { + JsonbValue *v = &jb->val.array.elems[i]; + + if (v->type == jbvString) + { + if (!first) + appendStringInfoString(&buf, sep); + appendBinaryStringInfo(&buf, v->val.string.val, v->val.string.len); + first = false; + } + else if (v->type == jbvNull) + { + if (null_replace) + { + if (!first) + appendStringInfoString(&buf, sep); + appendStringInfoString(&buf, null_replace); + first = false; + } + } + else + { + RETURN_ERROR(ereport(ERROR, + (errcode(ERRCODE_INVALID_PARAMETER_VALUE), + errmsg("jsonpath .join() array elements must be strings or nulls")))); + } + } + } + + jbv_res.type = jbvString; + jbv_res.val.string.val = buf.data; + jbv_res.val.string.len = buf.len; + + hasNext = jspGetNext(jsp, &next_elem); + if (!hasNext && !found) + return jperOk; + + return executeNextItem(cxt, jsp, hasNext ? &next_elem : NULL, &jbv_res, found); + } default: elog(ERROR, "unrecognized jsonpath item type: %d", jsp->type); diff --git a/src/backend/utils/adt/jsonpath_gram.y b/src/backend/utils/adt/jsonpath_gram.y index fe30e8a6147..eb2b952e311 100644 --- a/src/backend/utils/adt/jsonpath_gram.y +++ b/src/backend/utils/adt/jsonpath_gram.y @@ -87,7 +87,7 @@ static bool makeItemLikeRegex(JsonPathParseItem *expr, %token BIGINT_P BOOLEAN_P DATE_P DECIMAL_P INTEGER_P NUMBER_P %token STRINGFUNC_P TIME_P TIME_TZ_P TIMESTAMP_P TIMESTAMP_TZ_P %token STR_REPLACE_P STR_LOWER_P STR_UPPER_P STR_LTRIM_P STR_RTRIM_P STR_BTRIM_P - STR_INITCAP_P STR_SPLIT_P STR_SPLIT_PART_P STR_TRANSLATE_P + STR_INITCAP_P STR_SPLIT_P STR_SPLIT_PART_P STR_TRANSLATE_P STR_JOIN_P %type result @@ -290,6 +290,10 @@ accessor_op: { $$ = makeItemBinary(jpiStrSplit, $4, NULL); } | '.' STR_SPLIT_P '(' str_str_args ')' { $$ = makeItemBinary(jpiStrSplit, linitial($4), lsecond($4)); } + | '.' STR_JOIN_P '(' str_elem ')' + { $$ = makeItemBinary(jpiStrJoin, $4, NULL); } + | '.' STR_JOIN_P '(' str_str_args ')' + { $$ = makeItemBinary(jpiStrJoin, linitial($4), lsecond($4)); } | '.' STR_LTRIM_P '(' opt_str_arg ')' { $$ = makeItemUnary(jpiStrLtrim, $4); } | '.' STR_RTRIM_P '(' opt_str_arg ')' @@ -393,6 +397,7 @@ key_name: | STR_LTRIM_P | STR_RTRIM_P | STR_BTRIM_P + | STR_JOIN_P ; method: diff --git a/src/backend/utils/adt/jsonpath_scan.l b/src/backend/utils/adt/jsonpath_scan.l index b06d0abb64a..bd49941a8fc 100644 --- a/src/backend/utils/adt/jsonpath_scan.l +++ b/src/backend/utils/adt/jsonpath_scan.l @@ -406,6 +406,7 @@ static const JsonPathKeyword keywords[] = { {3, false, LAX_P, "lax"}, {4, false, DATE_P, "date"}, {4, false, FLAG_P, "flag"}, + {4, false, STR_JOIN_P, "join"}, {4, false, LAST_P, "last"}, {4, true, NULL_P, "null"}, {4, false, SIZE_P, "size"}, diff --git a/src/include/utils/jsonpath.h b/src/include/utils/jsonpath.h index c1c7812a36a..f61639e3767 100644 --- a/src/include/utils/jsonpath.h +++ b/src/include/utils/jsonpath.h @@ -115,6 +115,7 @@ typedef enum JsonPathItemType jpiTimeTz, /* .time_tz() item method */ jpiTimestamp, /* .timestamp() item method */ jpiTimestampTz, /* .timestamp_tz() item method */ + jpiStrJoin, /* .join() item method */ jpiStrReplace, /* .replace() item method */ jpiStrLower, /* .lower() item method */ jpiStrUpper, /* .upper() item method */ diff --git a/src/test/regress/expected/jsonb_jsonpath.out b/src/test/regress/expected/jsonb_jsonpath.out index 8fc530201a6..7e247ba7505 100644 --- a/src/test/regress/expected/jsonb_jsonpath.out +++ b/src/test/regress/expected/jsonb_jsonpath.out @@ -3135,6 +3135,80 @@ select jsonb_path_query('"a,b,c"', '$.split(",")[1]'); "b" (1 row) +-- Test .join() method +select jsonb_path_query('["a", "b", "c"]', '$.join("-")'); + jsonb_path_query +------------------ + "a-b-c" +(1 row) + +-- Join with null replacement +select jsonb_path_query('["a", "b", "c"]', '$.join("-", "N/A")'); + jsonb_path_query +------------------ + "a-b-c" +(1 row) + +-- Null handling: default (skip) +select jsonb_path_query('["a", null, "c"]', '$.join("-")'); + jsonb_path_query +------------------ + "a-c" +(1 row) + +-- Null handling: replacement +select jsonb_path_query('["a", null, "c"]', '$.join("-", "N/A")'); + jsonb_path_query +------------------ + "a-N/A-c" +(1 row) + +-- Empty array (should return empty string) +select jsonb_path_query('[]', '$.join("-")'); + jsonb_path_query +------------------ + "" +(1 row) + +-- Pipeline integration: .join().upper() +select jsonb_path_query('["hello", "world"]', '$.join(" ").upper()'); + jsonb_path_query +------------------ + "HELLO WORLD" +(1 row) + +-- Pipeline integration: .split().join() +select jsonb_path_query('"a,b,c"', '$.split(",").join("|")'); + jsonb_path_query +------------------ + "a|b|c" +(1 row) + +-- Error case: Non-string element (should trigger our ereport) +select jsonb_path_query('[1, "a"]', '$.join("-")'); +ERROR: jsonpath .join() array elements must be strings or nulls +-- Error case: Nested object +select jsonb_path_query('["a", {"b": 1}]', '$.join("-")'); +ERROR: jsonpath .join() array elements must be strings or nulls +-- Error case: Applied to a scalar +select jsonb_path_query('"not an array"', '$.join("-")'); +ERROR: jsonpath item method .join() can only be applied to an array +-- Lax mode: should still error under current conservative implementation +select jsonb_path_query('[1, "a"]', 'lax $.join("-")'); +ERROR: jsonpath .join() array elements must be strings or nulls +select jsonb_path_query('"not an array"', 'lax $.join("-")'); +ERROR: jsonpath item method .join() can only be applied to an array +-- Silent mode: should suppress errors and return no rows +select jsonb_path_query('[1, "a"]', '$.join("-")', silent => true); + jsonb_path_query +------------------ +(0 rows) + +select jsonb_path_query('"not an array"', '$.join("-")', silent => true); + jsonb_path_query +------------------ +(0 rows) + -- Test string methods play nicely together select jsonb_path_query('"hello world"', '$.replace("hello","bye").upper()'); jsonb_path_query @@ -3166,6 +3240,12 @@ select jsonb_path_query('" hElLo WorlD "', '$.btrim().lower().upper().lower(). true (1 row) +select jsonb_path_query('" A,b,C "', '$.btrim().lower().split(",").join("-").replace("a","x").upper() starts with "X-B"'); + jsonb_path_query +------------------ + true +(1 row) + -- Test .time() select jsonb_path_query('null', '$.time()'); ERROR: jsonpath item method .time() can only be applied to a string diff --git a/src/test/regress/expected/jsonpath.out b/src/test/regress/expected/jsonpath.out index a9ef9cefced..6b64603a0bf 100644 --- a/src/test/regress/expected/jsonpath.out +++ b/src/test/regress/expected/jsonpath.out @@ -519,6 +519,18 @@ select '$.split(",")'::jsonpath; $.split(",") (1 row) +select '$.join(",")'::jsonpath; + jsonpath +------------- + $.join(",") +(1 row) + +select '$.join(",", "N/A")'::jsonpath; + jsonpath +-------------------- + $.join(",", "N/A") +(1 row) + -- Parse errors select '$.replace("hello")'::jsonpath; ERROR: syntax error at or near ")" of jsonpath input @@ -655,6 +667,26 @@ select '$.btrim'::jsonpath; $."btrim" (1 row) +select '$.join()'::jsonpath; +ERROR: syntax error at or near ")" of jsonpath input +LINE 1: select '$.join()'::jsonpath; + ^ +select '$.join(",", "replacement", "extra")'::jsonpath; +ERROR: syntax error at or near "," of jsonpath input +LINE 1: select '$.join(",", "replacement", "extra")'::jsonpath; + ^ +select '$.join(42)'::jsonpath; +ERROR: syntax error at or near "42" of jsonpath input +LINE 1: select '$.join(42)'::jsonpath; + ^ +select '$.join(true)'::jsonpath; +ERROR: syntax error at end of jsonpath input +LINE 1: select '$.join(true)'::jsonpath; + ^ +select '$.join("x", 123)'::jsonpath; +ERROR: syntax error at or near "123" of jsonpath input +LINE 1: select '$.join("x", 123)'::jsonpath; + ^ select '$.time()'::jsonpath; jsonpath ---------- diff --git a/src/test/regress/expected/sqljson_queryfuncs.out b/src/test/regress/expected/sqljson_queryfuncs.out index c3fc890a9c4..2df3e34d0e0 100644 --- a/src/test/regress/expected/sqljson_queryfuncs.out +++ b/src/test/regress/expected/sqljson_queryfuncs.out @@ -1282,6 +1282,7 @@ CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.replace("hello", "bye") CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.translate("hello", "bye")')); CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.split_part(",", 2)')); CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.split(",")')); +CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.join(",")')); -- DEFAULT expression CREATE OR REPLACE FUNCTION ret_setint() RETURNS SETOF integer AS $$ diff --git a/src/test/regress/sql/jsonb_jsonpath.sql b/src/test/regress/sql/jsonb_jsonpath.sql index 784fbc577f2..86f7d2fe4b6 100644 --- a/src/test/regress/sql/jsonb_jsonpath.sql +++ b/src/test/regress/sql/jsonb_jsonpath.sql @@ -739,12 +739,40 @@ select jsonb_path_query('"a,,c"', '$.split(",", "")'); -- proving the output is a real, indexable JSON array select jsonb_path_query('"a,b,c"', '$.split(",")[1]'); +-- Test .join() method +select jsonb_path_query('["a", "b", "c"]', '$.join("-")'); +-- Join with null replacement +select jsonb_path_query('["a", "b", "c"]', '$.join("-", "N/A")'); +-- Null handling: default (skip) +select jsonb_path_query('["a", null, "c"]', '$.join("-")'); +-- Null handling: replacement +select jsonb_path_query('["a", null, "c"]', '$.join("-", "N/A")'); +-- Empty array (should return empty string) +select jsonb_path_query('[]', '$.join("-")'); +-- Pipeline integration: .join().upper() +select jsonb_path_query('["hello", "world"]', '$.join(" ").upper()'); +-- Pipeline integration: .split().join() +select jsonb_path_query('"a,b,c"', '$.split(",").join("|")'); +-- Error case: Non-string element (should trigger our ereport) +select jsonb_path_query('[1, "a"]', '$.join("-")'); +-- Error case: Nested object +select jsonb_path_query('["a", {"b": 1}]', '$.join("-")'); +-- Error case: Applied to a scalar +select jsonb_path_query('"not an array"', '$.join("-")'); +-- Lax mode: should still error under current conservative implementation +select jsonb_path_query('[1, "a"]', 'lax $.join("-")'); +select jsonb_path_query('"not an array"', 'lax $.join("-")'); +-- Silent mode: should suppress errors and return no rows +select jsonb_path_query('[1, "a"]', '$.join("-")', silent => true); +select jsonb_path_query('"not an array"', '$.join("-")', silent => true); + -- Test string methods play nicely together select jsonb_path_query('"hello world"', '$.replace("hello","bye").upper()'); select jsonb_path_query('"hElLo WorlD"', '$.lower().upper().lower().replace("hello","bye")'); select jsonb_path_query('"hElLo WorlD"', '$.upper().lower().upper().replace("HELLO", "BYE")'); select jsonb_path_query('"hElLo WorlD"', '$.lower().upper().lower().replace("hello","bye") starts with "bye"'); select jsonb_path_query('" hElLo WorlD "', '$.btrim().lower().upper().lower().replace("hello","bye") starts with "bye"'); +select jsonb_path_query('" A,b,C "', '$.btrim().lower().split(",").join("-").replace("a","x").upper() starts with "X-B"'); -- Test .time() select jsonb_path_query('null', '$.time()'); diff --git a/src/test/regress/sql/jsonpath.sql b/src/test/regress/sql/jsonpath.sql index ef27c8e7bb6..c524dc7a3d4 100644 --- a/src/test/regress/sql/jsonpath.sql +++ b/src/test/regress/sql/jsonpath.sql @@ -92,6 +92,8 @@ select '$.btrim("xyz")'::jsonpath; select '$.initcap()'::jsonpath; select '$.split_part("~@~", 2)'::jsonpath; select '$.split(",")'::jsonpath; +select '$.join(",")'::jsonpath; +select '$.join(",", "N/A")'::jsonpath; -- Parse errors select '$.replace("hello")'::jsonpath; @@ -125,6 +127,11 @@ select '$.split'::jsonpath; select '$.ltrim'::jsonpath; select '$.rtrim'::jsonpath; select '$.btrim'::jsonpath; +select '$.join()'::jsonpath; +select '$.join(",", "replacement", "extra")'::jsonpath; +select '$.join(42)'::jsonpath; +select '$.join(true)'::jsonpath; +select '$.join("x", 123)'::jsonpath; select '$.time()'::jsonpath; select '$.time(6)'::jsonpath; diff --git a/src/test/regress/sql/sqljson_queryfuncs.sql b/src/test/regress/sql/sqljson_queryfuncs.sql index 83ddfc8bc83..ba9d7b7ee47 100644 --- a/src/test/regress/sql/sqljson_queryfuncs.sql +++ b/src/test/regress/sql/sqljson_queryfuncs.sql @@ -412,6 +412,7 @@ CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.replace("hello", "bye") CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.translate("hello", "bye")')); CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.split_part(",", 2)')); CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.split(",")')); +CREATE INDEX ON test_jsonb_mutability (JSON_QUERY(js, '$.join(",")')); -- DEFAULT expression CREATE OR REPLACE FUNCTION ret_setint() RETURNS SETOF integer AS -- 2.53.0