Applies to VAST Cluster 5.5.
Functions › Search and split › Pattern matching. like_escape(s, pattern, escape) is the function form of s LIKE pattern ESCAPE escape: true if the whole string matches a LIKE pattern in which the escape character makes the next % or _ literal. not_like_escape returns the opposite, like NOT LIKE ... ESCAPE.
Syntax
like_escape(s, pattern, escape)
not_like_escape(s, pattern, escape)
| Argument | Type | Description |
|---|---|---|
s |
VARCHAR |
The string to test. |
pattern |
VARCHAR |
A LIKE pattern: % matches any run of characters, _ exactly one. |
escape |
VARCHAR |
Empty, or one ASCII character. In the pattern, it makes the next %, _ or escape character literal. |
Returns: BOOLEAN.
Related: s LIKE pattern ESCAPE escape and s NOT LIKE pattern ESCAPE escape are the same tests; for a case-insensitive match, use ilike_escape.
Set up the examples
The examples connect with the ADBC driver, as in VAST Query Engine (ADBC), and work in a schema called docs_sql_fn_string_like_escape. Set QE_ENDPOINT, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and VASTDB_BUCKET (SQL reference), and run the blocks in order in one Python session. Below each block is the output you should see.
This block connects, defines the run() and plan() helpers, and creates a table of five file paths. Rows 1 and 3 have a literal _ or % in their names; rows 2 and 4 are the same names with another character in that place. The last row has no values:
import os
import re
import adbc_driver_manager
import adbc_driver_manager.dbapi as dbapi
import adbc_driver_vastdb
BUCKET = os.environ["VASTDB_BUCKET"]
SCHEMA = f'"{BUCKET}/docs_sql_fn_string_like_escape"'
conn = dbapi.connect(
driver=adbc_driver_vastdb.get_driver_path(),
db_kwargs={
"vast.db.endpoint": os.environ["QE_ENDPOINT"],
"vast.db.access_key": os.environ["AWS_ACCESS_KEY_ID"],
"vast.db.secret_key": os.environ["AWS_SECRET_ACCESS_KEY"],
},
autocommit=True,
)
def server_message(e):
"""The server's message in a driver error, without the request ID or the echoed statement."""
text = re.sub(r"\[req=0x[0-9a-f]+\]: ", "", str(e))
found = (re.search(r'message: "(.*?)", request_id', text, re.S)
or re.search(r"description: (.*)\)", text, re.S))
message = found.group(1) if found else text.strip().splitlines()[-1]
message = re.sub(r"^Plan query: '.*?' failed with: ", "", message, flags=re.S)
message = re.split(r"\\n|\n", message)[0].replace('\\"', '"')
return message.replace(BUCKET, "<bucket>")
def run(sql):
"""Run one statement. Print the rows it returns, or the error."""
try:
with conn.cursor() as cur:
cur.execute(sql)
rows = cur.fetchall() if cur.description else []
except (adbc_driver_manager.Error, OSError) as e:
print(f"{type(e).__name__}: {server_message(e)}")
else:
print(rows)
def plan(sql):
"""Print the plan without its ~N Rows lines: the table's row count, which can still be 0 just after an insert."""
with conn.cursor() as cur:
cur.execute("EXPLAIN " + sql)
steps = cur.fetchall()[0][1]
print("\n".join(line for line in steps.splitlines() if "Rows" not in line))
TABLE = f"{SCHEMA}.files"
run(f"CREATE SCHEMA IF NOT EXISTS {SCHEMA}")
run(f"CREATE TABLE IF NOT EXISTS {TABLE} (id BIGINT, path VARCHAR)")
run(f"INSERT INTO {TABLE} VALUES "
"(1, '/data/run_1.txt'), "
"(2, '/data/runx1.txt'), "
"(3, '/tmp/load_50%.csv'), "
"(4, '/tmp/load_50x.csv'), "
"(5, NULL)")
[]
[]
[(5,)]
What it does
like_escape(s, pattern, escape) gives the same result as s LIKE pattern ESCAPE escape, and not_like_escape the same as s NOT LIKE pattern ESCAPE escape. In the pattern:
In the pattern (escape $) |
Matches | Example | Result |
|---|---|---|---|
%, _ |
any run of characters; exactly one character | like_escape('ab', 'a%', '$') |
true |
$%, $_ |
a literal % or _ |
like_escape('a%b', 'a$%b', '$') |
true |
$$ |
a literal $ |
like_escape('a$b', 'a$$b', '$') |
true |
$ before another character |
that character | like_escape('ab', 'a$b', '$') |
true |
$ at the end of the pattern |
nothing | like_escape('ab', 'ab$', '$') |
false |
- The match is case-sensitive and covers the whole string, as for LIKE.
like_escape('AB', 'ab', '$')is false. - An empty escape means no escape character:
%and_are always wildcards. - Choose an escape character that isn't in your data, such as
$or!. A backslash works too:like_escape('a_b', 'a\_b', '\')is true.
NULLs and edge cases
| Input | Result |
|---|---|
NULL as any of the three arguments |
NULL |
like_escape('', '', '$'), like_escape('', '%', '$') |
true |
like_escape('ab', 'ab', '') (no escape character) |
true |
like_escape('axb', 'a$%b', '$') |
false |
like_escape('a', 'a$', '$') (pattern ends with the escape character) |
false |
like_escape('a_b', 'a\_b', '\') |
true |
like_escape('é', '_', '$') (precomposed) |
true |
like_escape('e\u0301', '_', '$') (e and a combining accent: two code points) |
false |
like_escape('a%', 'a%%', '%') (% as the escape character) |
true |
not_like_escape('a%b', 'a$%b', '$'), not_like_escape('axb', 'a$%b', '$') |
false, true |
An escape of two characters, or a non-ASCII one such as é |
Error: Invalid escape string. Escape string must be empty or one character. |
The same cases, as the cluster returns them. They use the table and helpers from Set up the examples:
run("SELECT like_escape(NULL, 'a', '$'), like_escape('a', NULL, '$'), like_escape('a', 'a', NULL), "
"like_escape('', '', '$'), like_escape('', '%', '$'), like_escape('ab', 'ab', '')")
run("SELECT like_escape('axb', 'a$%b', '$'), like_escape('a', 'a$', '$'), like_escape('a_b', 'a\\_b', '\\'), "
"like_escape('\u00e9', '_', '$'), like_escape('e\u0301', '_', '$'), like_escape('a%', 'a%%', '%')")
run("SELECT not_like_escape('a%b', 'a$%b', '$'), not_like_escape('axb', 'a$%b', '$')")
run("SELECT like_escape('ab', 'a_', '$$')")
run("SELECT like_escape('ab', 'ab', '\u00e9')")
[(None, None, None, True, True, True)]
[(False, False, True, True, False, True)]
[(False, True)]
OSError: Invalid escape string. Escape string must be empty or one character.
OSError: Invalid escape string. Escape string must be empty or one character.
Examples
Without escaping, _ and % are wildcards, so '%run_1%' matches rows 1 and 2. With $ as the escape character, only the literal _ matches. The LIKE ... ESCAPE form gives the same result:
run(f"SELECT id, path LIKE '%run_1%', like_escape(path, '%run$_1%', '$'), path LIKE '%run$_1%' ESCAPE '$' "
f"FROM {TABLE} ORDER BY id LIMIT 10")
[(1, True, True, True), (2, True, False, False), (3, False, False, False), (4, False, False, False), (5, None, None, None)]
Keep the files whose name has a literal %, and those that don't:
run(f"SELECT id, path FROM {TABLE} WHERE like_escape(path, '%$%%', '$') ORDER BY id LIMIT 10")
run(f"SELECT id, path FROM {TABLE} WHERE not_like_escape(path, '%$%%', '$') ORDER BY id LIMIT 10")
[(3, '/tmp/load_50%.csv')]
[(1, '/data/run_1.txt'), (2, '/data/runx1.txt'), (4, '/tmp/load_50x.csv')]
How it runs in VAST
The Query Engine runs on the cluster's CNodes, and a query on a table starts with a scan in VAST DataBase (VAST_SCAN). VAST DataBase can evaluate a comparison of a column with a constant, but not a pattern match, so like_escape() is always computed by the Query Engine:
- In the
SELECTlist, the scan reads the column and the Query Engine computeslike_escape()for each row it returns. - In
WHERE, the scan sends the column for every row, and aFILTERstep in the Query Engine keeps the rows that match. No rows are skipped in storage.
s LIKE pattern ESCAPE escape is planned as like_escape(s, pattern, escape). Unlike LIKE without ESCAPE, a prefix pattern isn't rewritten to prefix(). The plans of the function and of the LIKE ... ESCAPE form:
plan(f"SELECT id FROM {TABLE} WHERE like_escape(path, '%$%%', '$')")
plan(f"SELECT id FROM {TABLE} WHERE path LIKE '/tmp/%' ESCAPE '$'")
┌───────────────────────────┐
│ PROJECTION │
│ ──────────────────── │
│ #1 │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ FILTER │
│ ──────────────────── │
│ like_escape(path, '%$%%', │
│ '$') │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ VAST_SCAN │
│ ──────────────────── │
│ Function: VAST_SCAN │
│ │
│ Projections: │
│ path │
│ id │
│ │
└───────────────────────────┘
┌───────────────────────────┐
│ PROJECTION │
│ ──────────────────── │
│ #1 │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ FILTER │
│ ──────────────────── │
│ like_escape(path, '/tmp/%'│
│ , '$') │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ VAST_SCAN │
│ ──────────────────── │
│ Function: VAST_SCAN │
│ │
│ Projections: │
│ path │
│ id │
│ │
└───────────────────────────┘
Make it faster
- Leave out
ESCAPEwhen the pattern doesn't need it.LIKE '/tmp/%'is rewritten toprefix(), and a pattern without wildcards to an equality test that is pushed down to VAST DataBase; withESCAPEorlike_escape(), the pattern always runs aslike_escape(). - Use
contains()for a literal substring. To find a literal%or_anywhere, contains(path, '%')needs no escape character at all. - Combine it with a pushed-down condition. In
id > 1 AND like_escape(path, '%$%%', '$'),id > 1is still evaluated in VAST DataBase, solike_escape()runs only on the rows that pass it.
plan(f"SELECT id FROM {TABLE} WHERE path LIKE '/tmp/%'")
run(f"SELECT id, contains(path, '%'), like_escape(path, '%$%%', '$') FROM {TABLE} ORDER BY id LIMIT 10")
┌───────────────────────────┐
│ PROJECTION │
│ ──────────────────── │
│ #1 │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ FILTER │
│ ──────────────────── │
│ prefix(path, '/tmp/') │
│ │
└─────────────┬─────────────┘
┌─────────────┴─────────────┐
│ VAST_SCAN │
│ ──────────────────── │
│ Function: VAST_SCAN │
│ │
│ Projections: │
│ path │
│ id │
│ │
└───────────────────────────┘
[(1, False, False), (2, False, False), (3, True, True), (4, False, False), (5, None, None)]
Limitations and errors
- The escape is empty or one ASCII character. Two characters, or one non-ASCII character such as
é, fail withInvalid escape string. Escape string must be empty or one character.An escape taken from a column fails the same way if any row holds a longer value. - A pattern can't end with the escape character when it is
%.like_escape('ab', 'a%', '%')fails withLike pattern must not end with escape character!. With another escape character,like_escape('a', 'a$', '$')is false and doesn't fail. - All three arguments are strings. A number as the escape fails with
No function matches the given name and argument types 'like_escape(STRING_LITERAL, STRING_LITERAL, INTEGER_LITERAL)'; a number, date, boolean,BLOBor list as the string fails the same way.
run("SELECT like_escape('ab', 'a%', '%')")
run("SELECT like_escape('ab', 'ab', 1)")
OSError: Like pattern must not end with escape character!
InternalError: Binder Error: No function matches the given name and argument types 'like_escape(STRING_LITERAL, STRING_LITERAL, INTEGER_LITERAL)'. You might need to add explicit type casts.
Similar functions
| Function | Same as | Case |
|---|---|---|
like_escape(s, p, e) |
s LIKE p ESCAPE e |
sensitive |
not_like_escape(s, p, e) |
s NOT LIKE p ESCAPE e |
sensitive |
ilike_escape (s, p, e) |
s ILIKE p ESCAPE e |
ignored |
| LIKE | like_escape with no escape character |
sensitive |
| contains | LIKE '%text%' with every character literal |
sensitive |
run("SELECT 'a%b' LIKE 'a$%b' ESCAPE '$', like_escape('a%b', 'a$%b', '$'), not_like_escape('a%b', 'a$%b', '$'), "
"ilike_escape('A%B', 'a$%b', '$'), 'a%b' LIKE '%$%%' ESCAPE '$', contains('a%b', '%')")
[(True, True, False, True, True, True)]
Clean up:
run(f"DROP TABLE IF EXISTS {TABLE}")
run(f"DROP SCHEMA IF EXISTS {SCHEMA}")
conn.close()
[]
[]
See also
- Search and split: all the search and pattern functions.
- Operators and predicates.
- Where filters run and pushdowns.
- EXPLAIN: how to read a plan.