From 78438b8a2916836dbed117ab072d00ab265f4211 Mon Sep 17 00:00:00 2001
From: Kirill Shcherbatov <kshcherbatov@tarantool.org>
Date: Sun, 25 Nov 2018 13:10:38 +0300
Subject: [PATCH] json: introduce json_path_cmp, json_path_validate
Introduced json_path_validate routine to ensure user-defined
JSON path is valid. This will be required to raise an error if
an incorrect user-defined jason-path is detected.
Introduced json_path_cmp routine to compare JSON paths that may
have different representation.
Note that:
- in case of paths that have same token-sequence prefix,
the path having more tokens is assumed to be greater
- both paths to compare must be valid
@locker: move json_path_validate to the object file (no need to
pollute the header with this cold routine).
Needed for #1012
---
src/lib/json/json.c | 42 ++++++++++++++++++++++++++++++++++++++
src/lib/json/json.h | 18 ++++++++++++++++
test/unit/json_path.c | 37 ++++++++++++++++++++++++++++++++-
test/unit/json_path.result | 13 +++++++++++-
4 files changed, 108 insertions(+), 2 deletions(-)
diff --git a/src/lib/json/json.c b/src/lib/json/json.c
index 430caf221d..c7909fde24 100644
--- a/src/lib/json/json.c
+++ b/src/lib/json/json.c
@@ -275,6 +275,48 @@ json_token_cmp(const struct json_token *a, const struct json_token *b)
return ret;
}
+int
+json_path_cmp(const char *a, int a_len, const char *b, int b_len,
+ int index_base)
+{
+ struct json_lexer lexer_a, lexer_b;
+ json_lexer_create(&lexer_a, a, a_len, index_base);
+ json_lexer_create(&lexer_b, b, b_len, index_base);
+ struct json_token token_a, token_b;
+ /* For the sake of json_token_cmp(). */
+ token_a.parent = NULL;
+ token_b.parent = NULL;
+ int rc_a, rc_b;
+ while ((rc_a = json_lexer_next_token(&lexer_a, &token_a)) == 0 &&
+ (rc_b = json_lexer_next_token(&lexer_b, &token_b)) == 0 &&
+ token_a.type != JSON_TOKEN_END &&
+ token_b.type != JSON_TOKEN_END) {
+ int rc = json_token_cmp(&token_a, &token_b);
+ if (rc != 0)
+ return rc;
+ }
+ /* Paths a and b must be valid. */
+ assert(rc_b == 0 && rc_b == 0);
+ /*
+ * The parser stopped because the end of one of the paths
+ * was reached. As JSON_TOKEN_END > JSON_TOKEN_{NUM, STR},
+ * the path having more tokens has lower key.type value.
+ */
+ return token_b.type - token_a.type;
+}
+
+int
+json_path_validate(const char *path, int path_len, int index_base)
+{
+ struct json_lexer lexer;
+ json_lexer_create(&lexer, path, path_len, index_base);
+ struct json_token token;
+ int rc;
+ while ((rc = json_lexer_next_token(&lexer, &token)) == 0 &&
+ token.type != JSON_TOKEN_END) {}
+ return rc;
+}
+
#define MH_SOURCE 1
#define mh_name _json
#define mh_key_t struct json_token *
diff --git a/src/lib/json/json.h b/src/lib/json/json.h
index ffdd7a9052..7ce10ce2ce 100644
--- a/src/lib/json/json.h
+++ b/src/lib/json/json.h
@@ -235,6 +235,24 @@ json_lexer_create(struct json_lexer *lexer, const char *src, int src_len,
int
json_lexer_next_token(struct json_lexer *lexer, struct json_token *token);
+/**
+ * Compare two JSON paths using Lexer class.
+ * - in case of paths that have same token-sequence prefix,
+ * the path having more tokens is assumed to be greater
+ * - both paths must be valid
+ * (may be tested with json_path_validate).
+ */
+int
+json_path_cmp(const char *a, int a_len, const char *b, int b_len,
+ int index_base);
+
+/**
+ * Check if the passed JSON path is valid.
+ * Return 0 for valid path and error position for invalid.
+ */
+int
+json_path_validate(const char *path, int path_len, int index_base);
+
/**
* Initialize a new empty JSON tree.
*
diff --git a/test/unit/json_path.c b/test/unit/json_path.c
index aa692e8a0a..e553ff23c5 100644
--- a/test/unit/json_path.c
+++ b/test/unit/json_path.c
@@ -404,15 +404,50 @@ test_tree()
footer();
}
+void
+test_path_cmp()
+{
+ const char *a = "Data[1][\"FIO\"].fname";
+ uint32_t a_len = strlen(a);
+ const struct path_and_errpos rc[] = {
+ {a, 0},
+ {"[\"Data\"][1].FIO[\"fname\"]", 0},
+ {"Data[1]", 1},
+ {"Data[1][\"FIO\"].fname[1]", -1},
+ {"Data[1][\"Info\"].fname[1]", -1},
+ };
+ header();
+ plan(lengthof(rc) + 2);
+ for (size_t i = 0; i < lengthof(rc); ++i) {
+ const char *path = rc[i].path;
+ int errpos = rc[i].errpos;
+ int rc = json_path_cmp(a, a_len, path, strlen(path),
+ INDEX_BASE);
+ if (rc > 0) rc = 1;
+ if (rc < 0) rc = -1;
+ is(rc, errpos, "path cmp result \"%s\" with \"%s\": "
+ "have %d, expected %d", a, path, rc, errpos);
+ }
+ const char *invalid = "Data[[1][\"FIO\"].fname";
+ int ret = json_path_validate(a, strlen(a), INDEX_BASE);
+ is(ret, 0, "path %s is valid", a);
+ ret = json_path_validate(invalid, strlen(invalid), INDEX_BASE);
+ is(ret, 6, "path %s error pos %d expected %d", invalid, ret, 6);
+
+ check_plan();
+ footer();
+}
+
int
main()
{
header();
- plan(3);
+ plan(4);
test_basic();
test_errors();
test_tree();
+ test_path_cmp();
int rc = check_plan();
footer();
diff --git a/test/unit/json_path.result b/test/unit/json_path.result
index ee23e70c1c..a174510997 100644
--- a/test/unit/json_path.result
+++ b/test/unit/json_path.result
@@ -1,5 +1,5 @@
*** main ***
-1..3
+1..4
*** test_basic ***
1..71
ok 1 - parse <[1]>
@@ -158,4 +158,15 @@ ok 2 - subtests
ok 54 - records iterated count 4 of 4
ok 3 - subtests
*** test_tree: done ***
+ *** test_path_cmp ***
+ 1..7
+ ok 1 - path cmp result "Data[1]["FIO"].fname" with "Data[1]["FIO"].fname": have 0, expected 0
+ ok 2 - path cmp result "Data[1]["FIO"].fname" with "["Data"][1].FIO["fname"]": have 0, expected 0
+ ok 3 - path cmp result "Data[1]["FIO"].fname" with "Data[1]": have 1, expected 1
+ ok 4 - path cmp result "Data[1]["FIO"].fname" with "Data[1]["FIO"].fname[1]": have -1, expected -1
+ ok 5 - path cmp result "Data[1]["FIO"].fname" with "Data[1]["Info"].fname[1]": have -1, expected -1
+ ok 6 - path Data[1]["FIO"].fname is valid
+ ok 7 - path Data[[1]["FIO"].fname error pos 6 expected 6
+ok 4 - subtests
+ *** test_path_cmp: done ***
*** main: done ***
--
GitLab