branch: elpa/pg
commit 24ed24db001d37dbcb49ab54f46885a1e361f735
Author: Eric Marsden <eric.mars...@risk-engineering.org>
Commit: Eric Marsden <eric.mars...@risk-engineering.org>
Add serialization support for the _varchar data type
---
CHANGELOG.md | 2 ++
pg.el | 24 ++++++++++++++++--------
2 files changed, 18 insertions(+), 8 deletions(-)
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 6441b0bbc1e..cec95a03bb6 100755
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,8 @@
`text[]` (known as `_text` in the `pg_type` system table), `int4[]`,
`float4[]`, `float8[]` can now
be passed as function arguments.
+- Add serialization support for the `_varchar` data type.
+
- Add workarounds and detection code for the CedarDB PostgreSQL variant.
- New error classes `pg-duplicate-table` and `pg-duplicate-column`, subclasses
of
diff --git a/pg.el b/pg.el
index 3df1f200ce5..0a4520f5174 100644
--- a/pg.el
+++ b/pg.el
@@ -2385,6 +2385,8 @@ PostgreSQL and Emacs. CON should no longer be used."
("_json" "199")
("_bool" "1000")
("_char" "1002")
+ ("_bpchar" "1014")
+ ("_varchar" "1015")
("_int8" "1016")
("_int2" "1005")
("_int4" "1007")
@@ -2524,18 +2526,22 @@ the PostgreSQL connection CON."
(decode-coding-string str encoding)
str))
+;; STR could be either a single character (char datatype) or a character
sequence for the PostgreSQL
+;; character(n) datatype specifier, which is fixed-length and blank-padded.
Note that we return
+;; either a single character or a string.
(defun pg-char-parser (str encoding)
- (if encoding
- (aref (pg-text-parser str encoding) 0)
- (aref str 0)))
+ (let ((len (length str)))
+ (cond ((zerop len)
+ (signal 'pg-protocol-error (list "Unexpected zero-length char
data")))
+ ((eql 1 len)
+ (if encoding
+ (aref (pg-text-parser str encoding) 0)
+ (aref str 0)))
+ (t
+ (pg-text-parser str encoding)))))
(pg-register-parser "char" #'pg-char-parser)
(pg-register-parser "bpchar" #'pg-char-parser)
-
-(pg-register-parser "char2" #'pg-text-parser)
-(pg-register-parser "char4" #'pg-text-parser)
-(pg-register-parser "char8" #'pg-text-parser)
-(pg-register-parser "char16" #'pg-text-parser)
(pg-register-parser "name" #'pg-text-parser)
(pg-register-parser "text" #'pg-text-parser)
(pg-register-parser "varchar" #'pg-text-parser)
@@ -2740,6 +2746,7 @@ Uses text encoding ENCODING."
(apply #'vector (mapcar (lambda (x) (pg-text-parser x encoding))
items)))))))
(pg-register-parser "_text" #'pg-textarray-parser)
+(pg-register-parser "_varchar" #'pg-textarray-parser)
;; Anonymouse records in PostgreSQL (oid = 2249) are little used in practice,
and difficult to parse
;; because we receive no information concerning the types of the different
record "columns".
@@ -2980,6 +2987,7 @@ Return nil if the extension could not be set up."
(pg-register-textual-serializer "varbit" #'pg--serialize-boolvec)
+;; Here we assume that the value is a single character: if serializing a
string, use the "text" datatype specifier.
(pg-register-serializer "char"
(lambda (v _encoding)
(unless (<= 0 v 255)
