HexaBridge
OracleSQL Server
HexaBridge - SQL Server (T-SQL) Compatibility for PostgreSQL
HexaBridge brings SQL Server's T-SQL functions, data types, and operators to Community PostgreSQL - no proprietary fork required. 100% routine compatibility, delivered at native speed through C extensions.
0
SQL Server routine compatibility
0
SQL Server-Compatible Types
0
Functions & Aggregates
0
Operators
Deep T-SQL Compatibility on Community PostgreSQL
Moving from SQL Server to PostgreSQL usually means rewriting every T-SQL function call, every DATEADD/DATEDIFF, every CHARINDEX, and re-implementing types like nvarchar, uniqueidentifier, and hierarchyid by hand. HexaBridge removes that work: it resolves SQL Server functions and type names natively in PostgreSQL's sys schema, with SQL Server semantics (boundary-crossing DATEDIFF, trailing-space-insensitive nvarchar comparison, LEN trimming, T-SQL + string concatenation).
HexaBridge covers 100% of the SQL Server routine surface while preserving SQL Server semantics. Logic-heavy functions are written in C and direct maps are inlined by the planner, so the compatibility layer adds no performance overhead. It runs on Community PostgreSQL as-is with a single CREATE EXTENSION. No proprietary database fork required.
The compatibility no AI or migration tool can generate
AI and migration tools convert your schema and application code, but they cannot recreate the source database's built-in routines on PostgreSQL. HexaBridge provides exactly that runtime compatibility, so HexaRocket can achieve 100% automatic schema conversion. Without this layer, every built-in function your application relies on becomes a manual rewrite that no tool can fully automate.
Runs on Community PostgreSQL - no commercial fork needed
Works on Community PostgreSQL - No Proprietary Fork Required
HexaBridge for SQL Server is a standard PostgreSQL extension that installs with a simple CREATE EXTENSION hexabridge_mssql; command.
It is available to customers who migrate with HexaRocket and to customers of HexaCluster-approved partners. Contact us to see how it accelerates your SQL Server-to-PostgreSQL migration.
SQL Server-Compatible Types
12 types including nvarchar, nchar, datetime2, uniqueidentifier, rowversion, smallmoney, sql_variant, and a full hierarchyid. Optional geometry/geography via PostGIS.
SQL Server-Compatible Functions
180+ functions and aggregates - DATEADD, DATEDIFF, CHARINDEX, ISNULL, FORMAT, TRY_CONVERT, COMPRESS/DECOMPRESS, OPENJSON, and more.
Native Performance
Advanced performance through native C extensions. Logic-heavy functions run in C and direct maps are inlined by the planner, so the compatibility layer adds no performance overhead.
How It Works
-- Install HexaBridge for SQL Server on Community PostgreSQL
CREATE EXTENSION hexabridge_mssql;
-- Resolve sys.* / dbo.* names like SQL Server
SET search_path TO sys, public;
-- Use T-SQL functions naturally
SELECT charindex('cd', 'abcdef', 2); -- 3
SELECT dateadd('month', 2, '2021-01-31'); -- 2021-03-31 00:00:00
SELECT datediff('year', '2020-12-31', '2021-01-01'); -- 1 (boundary crossing)
SELECT isnull(NULL, 'default'); -- default
-- Optional: geometry/geography (requires PostGIS)
-- CREATE EXTENSION hexabridge_mssql_spatial CASCADE;SQL Server-Compatible Data Types
HexaBridge provides 12 SQL Server data types that behave like their T-SQL counterparts - character-count length semantics, trailing-space-insensitive comparison, and SQL Server datetime precision. Optional geometry/geography are available via a separate PostGIS-backed submodule.
All types are created in the sys schema. With SET search_path TO sys, public;, they can be used by bare name - just like in SQL Server.
Optional spatial types geometry and geography map 1:1 onto PostGIS via the separate hexabridge_mssql_spatial submodule, which requires PostGIS. The core extension has no PostGIS dependency.
SQL Server-Compatible Functions
HexaBridge provides 180+ standalone T-SQL functions and aggregates that follow SQL Server semantics - boundary-crossing DATEDIFF, char-accurate CHARINDEX, LEN trailing-space trimming, ISNULL, .NET-style FORMAT, and more. Analytic window functions (ROW_NUMBER, RANK, DENSE_RANK, NTILE, LEAD, LAG, FIRST_VALUE, LAST_VALUE, CUME_DIST, PERCENT_RANK) are native PostgreSQL and work unchanged.
String / Character(15)
CHARINDEX
charindex(needle, haystack [, start])1-based, char-accurate (multibyte-safe) search with optional start; empty needle returns 0. C-implemented.CHARINDEX
charindex(needle, haystack [, start])1-based, char-accurate (multibyte-safe) search with optional start; empty needle returns 0. C-implemented.LEN
len(str)Character length excluding trailing spaces (SQL Server semantics).LEN
len(str)Character length excluding trailing spaces (SQL Server semantics).PATINDEX
patindex(pattern, str)1-based position of a LIKE pattern (%/_).PATINDEX
patindex(pattern, str)1-based position of a LIKE pattern (%/_).LEFT / RIGHT
left(str, n) / right(str, n)Leftmost / rightmost n characters.LEFT / RIGHT
left(str, n) / right(str, n)Leftmost / rightmost n characters.REPLICATE
replicate(str, n)Repeats a string n times.REPLICATE
replicate(str, n)Repeats a string n times.SPACE
space(n)Returns n spaces.SPACE
space(n)Returns n spaces.STUFF
stuff(str, start, length, replace)Deletes length chars at start and inserts replace.STUFF
stuff(str, start, length, replace)Deletes length chars at start and inserts replace.ISNUMERIC
isnumeric(str)1 if the string is a valid numeric literal, else 0.ISNUMERIC
isnumeric(str)1 if the string is a valid numeric literal, else 0.CONCAT
concat(v1, v2, ...)Concatenates, treating NULL as empty string.CONCAT
concat(v1, v2, ...)Concatenates, treating NULL as empty string.CONCAT_WS
concat_ws(sep, v1, v2, ...)Concatenates with a separator, skipping NULLs.CONCAT_WS
concat_ws(sep, v1, v2, ...)Concatenates with a separator, skipping NULLs.STRING_SPLIT
string_split(str, sep)Table-valued split into rows.STRING_SPLIT
string_split(str, sep)Table-valued split into rows.QUOTENAME
quotename(str [, quote_char])Delimits an identifier (default []).QUOTENAME
quotename(str [, quote_char])Delimits an identifier (default []).STR
str(float [, length [, decimals]])Right-justified numeric-to-string; * overflow fill.STR
str(float [, length [, decimals]])Right-justified numeric-to-string; * overflow fill.STRING_ESCAPE
string_escape(str, 'json')JSON-escapes a string.STRING_ESCAPE
string_escape(str, 'json')JSON-escapes a string.UNICODE
unicode(str)Unicode code point of the first character.UNICODE
unicode(str)Unicode code point of the first character.Date / Time(16)
DATEPART
datepart(part, date)Extracts a date part (year, month, day, hour, ..., dayofyear, week, weekday). C-implemented token parsing.DATEPART
datepart(part, date)Extracts a date part (year, month, day, hour, ..., dayofyear, week, weekday). C-implemented token parsing.DATEADD
dateadd(part, n, date)Adds n of a date part with SQL Server end-of-month snapping.DATEADD
dateadd(part, n, date)Adds n of a date part with SQL Server end-of-month snapping.DATEDIFF
datediff(part, start, end)Boundary-crossing count (not elapsed-unit subtraction). Overflow-checked.DATEDIFF
datediff(part, start, end)Boundary-crossing count (not elapsed-unit subtraction). Overflow-checked.DATEDIFF_BIG
datediff_big(part, start, end)Like DATEDIFF but returns bigint.DATEDIFF_BIG
datediff_big(part, start, end)Like DATEDIFF but returns bigint.DATENAME
datename(part, date)Textual name of a date part.DATENAME
datename(part, date)Textual name of a date part.YEAR / MONTH / DAY
year(date) / month(date) / day(date)Component extractors.YEAR / MONTH / DAY
year(date) / month(date) / day(date)Component extractors.EOMONTH
eomonth(date [, months])Last day of the month (optionally offset).EOMONTH
eomonth(date [, months])Last day of the month (optionally offset).GETDATE / GETUTCDATE
getdate() / getutcdate()Current statement timestamp (local / UTC). STABLE.GETDATE / GETUTCDATE
getdate() / getutcdate()Current statement timestamp (local / UTC). STABLE.SYSDATETIME family
sysdatetime() / sysutcdatetime() / sysdatetimeoffset()High-precision current timestamps.SYSDATETIME family
sysdatetime() / sysutcdatetime() / sysdatetimeoffset()High-precision current timestamps.DATEFROMPARTS family
datefromparts(y,m,d), datetimefromparts(...), smalldatetimefromparts(...), datetimeoffsetfromparts(...), timefromparts(...)Build datetime values from components.DATEFROMPARTS family
datefromparts(y,m,d), datetimefromparts(...), smalldatetimefromparts(...), datetimeoffsetfromparts(...), timefromparts(...)Build datetime values from components.DATETRUNC
datetrunc(part, date)Truncate to a date part.DATETRUNC
datetrunc(part, date)Truncate to a date part.DATE_BUCKET
date_bucket(part, n, date [, origin])Bucket a date into fixed-width bins.DATE_BUCKET
date_bucket(part, n, date [, origin])Bucket a date into fixed-width bins.ISDATE
isdate(str)1 if the string parses as a date, else 0.ISDATE
isdate(str)1 if the string parses as a date, else 0.SWITCHOFFSET / TODATETIMEOFFSET
switchoffset(dto, tz) / todatetimeoffset(dt, tz)Adjust/attach a timezone offset. (tz-display semantics differ; functions present.)SWITCHOFFSET / TODATETIMEOFFSET
switchoffset(dto, tz) / todatetimeoffset(dt, tz)Adjust/attach a timezone offset. (tz-display semantics differ; functions present.)FORMAT
format(value, format [, culture]).NET-style date formatting.FORMAT
format(value, format [, culture]).NET-style date formatting.DATETIME_ROUND
datetime_round(ts)Exact SQL Server 1/300-second DATETIME tick rounding.DATETIME_ROUND
datetime_round(ts)Exact SQL Server 1/300-second DATETIME tick rounding.Numeric / Math(10)
ABS / SIGN / CEILING / FLOOR
abs(n) / sign(n) / ceiling(n) / floor(n)Type-preserving overloads (int/bigint/numeric/float).ABS / SIGN / CEILING / FLOOR
abs(n) / sign(n) / ceiling(n) / floor(n)Type-preserving overloads (int/bigint/numeric/float).ROUND
round(n, length [, function])Scale-preserving round; truncate flag (SQL Server semantics).ROUND
round(n, length [, function])Scale-preserving round; truncate flag (SQL Server semantics).SQUARE
square(n)n * n.SQUARE
square(n)n * n.SQRT / EXP / POWER
sqrt(n) / exp(n) / power(n, e)Standard math.SQRT / EXP / POWER
sqrt(n) / exp(n) / power(n, e)Standard math.LOG / LOG10
log(n [, base]) / log10(n)log(n) is natural log (T-SQL trap handled); log10 base-10.LOG / LOG10
log(n [, base]) / log10(n)log(n) is natural log (T-SQL trap handled); log10 base-10.PI
pi()3.14159...PI
pi()3.14159...DEGREES / RADIANS
degrees(n) / radians(n)Angle conversion.DEGREES / RADIANS
degrees(n) / radians(n)Angle conversion.Trigonometry
sin(n) / cos(n) / tan(n) / cot(n) / asin(n) / acos(n) / atan(n)Trigonometry (radians).Trigonometry
sin(n) / cos(n) / tan(n) / cot(n) / asin(n) / acos(n) / atan(n)Trigonometry (radians).ATN2
atn2(y, x)Two-argument arctangent.ATN2
atn2(y, x)Two-argument arctangent.RAND
rand([seed])Pseudo-random float in [0,1).RAND
rand([seed])Pseudo-random float in [0,1).Aggregate / Statistical(4)
STDEV / STDEVP
stdev(expr) / stdevp(expr)Sample / population standard deviation.STDEV / STDEVP
stdev(expr) / stdevp(expr)Sample / population standard deviation.VAR / VARP
var(expr) / varp(expr)Sample / population variance.VAR / VARP
var(expr) / varp(expr)Sample / population variance.COUNT_BIG
count_big(expr)COUNT returning bigint.COUNT_BIG
count_big(expr)COUNT returning bigint.CHECKSUM_AGG
checksum_agg(int)XOR-based group checksum (not value-identical to SQL Server).CHECKSUM_AGG
checksum_agg(int)XOR-based group checksum (not value-identical to SQL Server).Conversion(3)
TRY_CONVERT_*
try_convert_int/bigint/numeric/float/date/datetime(text)Typed safe conversion; returns NULL on bad input.TRY_CONVERT_*
try_convert_int/bigint/numeric/float/date/datetime(text)Typed safe conversion; returns NULL on bad input.CONVERT_DATETIME_STYLE
convert_datetime_style(value, style)SQL Server CONVERT date styles (1-126).CONVERT_DATETIME_STYLE
convert_datetime_style(value, style)SQL Server CONVERT date styles (1-126).FORMAT (numeric)
format(value, format [, culture]).NET-style numeric formatting (N/F/D/C/P/X and custom masks).FORMAT (numeric)
format(value, format [, culture]).NET-style numeric formatting (N/F/D/C/P/X and custom masks).JSON(4)
ISJSON
isjson(str)1 if valid JSON, else 0.ISJSON
isjson(str)1 if valid JSON, else 0.JSON_PATH_EXISTS
json_path_exists(json, path)1 if a JSON path resolves, else 0.JSON_PATH_EXISTS
json_path_exists(json, path)1 if a JSON path resolves, else 0.JSON_MODIFY
json_modify(json, path, value)Update/insert/delete at a JSON path.JSON_MODIFY
json_modify(json, path, value)Update/insert/delete at a JSON path.OPENJSON
openjson(json)Table-valued expansion (key/value/type), objects and arrays.OPENJSON
openjson(json)Table-valued expansion (key/value/type), objects and arrays.Metadata / Catalog(12)
OBJECT_ID / OBJECT_NAME
object_id(name) / object_name(id)Resolve object name <-> id.OBJECT_ID / OBJECT_NAME
object_id(name) / object_name(id)Resolve object name <-> id.OBJECT_SCHEMA_NAME / OBJECT_DEFINITION
object_schema_name(id) / object_definition(id)Schema name / source text.OBJECT_SCHEMA_NAME / OBJECT_DEFINITION
object_schema_name(id) / object_definition(id)Schema name / source text.OBJECTPROPERTY / COLUMNPROPERTY
objectproperty(id, prop) / columnproperty(id, col, prop)Object/column property flags (IsTable, AllowsNull, ...).OBJECTPROPERTY / COLUMNPROPERTY
objectproperty(id, prop) / columnproperty(id, col, prop)Object/column property flags (IsTable, AllowsNull, ...).COL_NAME / COL_LENGTH
col_name(id, n) / col_length(tab, col)Column name / length.COL_NAME / COL_LENGTH
col_name(id, n) / col_length(tab, col)Column name / length.SCHEMA_ID / SCHEMA_NAME
schema_id([name]) / schema_name([id])Schema lookups.SCHEMA_ID / SCHEMA_NAME
schema_id([name]) / schema_name([id])Schema lookups.TYPE_ID / TYPE_NAME
type_id(name) / type_name(id)Type lookups.TYPE_ID / TYPE_NAME
type_id(name) / type_name(id)Type lookups.DB_ID / DB_NAME
db_id([name]) / db_name([id])Database lookups.DB_ID / DB_NAME
db_id([name]) / db_name([id])Database lookups.SERVERPROPERTY / DATABASEPROPERTYEX
serverproperty(p) / databasepropertyex(db, p)Server / database properties.SERVERPROPERTY / DATABASEPROPERTYEX
serverproperty(p) / databasepropertyex(db, p)Server / database properties.APP_NAME
app_name()Application name for the session.APP_NAME
app_name()Application name for the session.INDEX_COL / INDEXPROPERTY
index_col(...) / indexproperty(...)Index metadata.INDEX_COL / INDEXPROPERTY
index_col(...) / indexproperty(...)Index metadata.FILE / FILEGROUP metadata
file_id(name) / file_name(id) / filegroup_id(name) / filegroup_name(id)File / filegroup metadata.FILE / FILEGROUP metadata
file_id(name) / file_name(id) / filegroup_id(name) / filegroup_name(id)File / filegroup metadata.STATS_DATE
stats_date(id, stat_id)Statistics last-updated time.STATS_DATE
stats_date(id, stat_id)Statistics last-updated time.Security(7)
SUSER_ID / SUSER_NAME / SUSER_SNAME / SUSER_SID
suser_id([login]) / suser_name([id]) / suser_sname([sid]) / suser_sid([login])Login identity lookups.SUSER_ID / SUSER_NAME / SUSER_SNAME / SUSER_SID
suser_id([login]) / suser_name([id]) / suser_sname([sid]) / suser_sid([login])Login identity lookups.USER_ID / USER_NAME
user_id([name]) / user_name([id])Database-user lookups.USER_ID / USER_NAME
user_id([name]) / user_name([id])Database-user lookups.IS_MEMBER / IS_ROLEMEMBER / IS_SRVROLEMEMBER
is_member(group) / is_rolemember(role) / is_srvrolemember(role)Role membership tests (mapped to PostgreSQL roles).IS_MEMBER / IS_ROLEMEMBER / IS_SRVROLEMEMBER
is_member(group) / is_rolemember(role) / is_srvrolemember(role)Role membership tests (mapped to PostgreSQL roles).HAS_PERMS_BY_NAME
has_perms_by_name(obj, class, perm)Permission check via has_table_privilege.HAS_PERMS_BY_NAME
has_perms_by_name(obj, class, perm)Permission check via has_table_privilege.ORIGINAL_LOGIN
original_login()Original session login.ORIGINAL_LOGIN
original_login()Original session login.LOGINPROPERTY
loginproperty(login, prop)Login property lookup.LOGINPROPERTY
loginproperty(login, prop)Login property lookup.PERMISSIONS
permissions([object_id [, column]])Deprecated in SQL Server 2008+ but valid in older releases; returns the documented permission bitmap over PostgreSQL privileges.PERMISSIONS
permissions([object_id [, column]])Deprecated in SQL Server 2008+ but valid in older releases; returns the documented permission bitmap over PostgreSQL privileges.System / Session(12)
ISNULL
isnull(check, replacement)2-argument NULL substitution.ISNULL
isnull(check, replacement)2-argument NULL substitution.IIF
iif(cond, t, f)Inline conditional.IIF
iif(cond, t, f)Inline conditional.CHOOSE
choose(index, v1, v2, ...)1-based pick; out-of-range returns NULL.CHOOSE
choose(index, v1, v2, ...)1-based pick; out-of-range returns NULL.NEWID / NEWSEQUENTIALID
newid() / newsequentialid()GUID generation.NEWID / NEWSEQUENTIALID
newid() / newsequentialid()GUID generation.SCOPE_IDENTITY / IDENT_CURRENT
scope_identity() / ident_current(table)Last identity values.SCOPE_IDENTITY / IDENT_CURRENT
scope_identity() / ident_current(table)Last identity values.CHECKSUM / BINARY_CHECKSUM
checksum(...) / binary_checksum(...)Deterministic 32-bit hash (not value-identical to SQL Server).CHECKSUM / BINARY_CHECKSUM
checksum(...) / binary_checksum(...)Deterministic 32-bit hash (not value-identical to SQL Server).HOST_NAME / HOST_ID
host_name() / host_id()Client host info.HOST_NAME / HOST_ID
host_name() / host_id()Client host info.CONTEXT_INFO / SET_CONTEXT_INFO
context_info() / set_context_info(bytea)Session context bytes.CONTEXT_INFO / SET_CONTEXT_INFO
context_info() / set_context_info(bytea)Session context bytes.SESSION_CONTEXT / SP_SET_SESSION_CONTEXT
session_context(key) / sp_set_session_context(key, value)Key/value session context.SESSION_CONTEXT / SP_SET_SESSION_CONTEXT
session_context(key) / sp_set_session_context(key, value)Key/value session context.FORMATMESSAGE
formatmessage(msg, args...)printf-style message formatting.FORMATMESSAGE
formatmessage(msg, args...)printf-style message formatting.PARSENAME
parsename('a.b.c.d', n)Extracts the nth name part from the right.PARSENAME
parsename('a.b.c.d', n)Extracts the nth name part from the right.COMPRESS / DECOMPRESS
compress(bytea|text) / decompress(bytea)GZIP compression, byte-interoperable with SQL Server.COMPRESS / DECOMPRESS
compress(bytea|text) / decompress(bytea)GZIP compression, byte-interoperable with SQL Server.Get HexaBridge for Your Migration
HexaBridge is available to customers who migrate their databases using HexaRocket, and to customers of HexaCluster-approved partners. We may open-source HexaBridge in the future - but you don't have to wait. Contact us today and we'll help you get started.
No proprietary PostgreSQL fork required. HexaBridge works as-is on Community PostgreSQL.