- Categories:
String & binary functions (Matching/Comparison)
EDITDISTANCE¶
Computes the Levenshtein distance between two input strings. It is the number of single-character insertions, deletions, or substitutions needed to convert one string to another.
Note
Unlike some other metrics (e.g. Damerau-Levenshtein distance), character transpositions are not considered.
Syntax¶
EDITDISTANCE( <string_expr1> , <string_expr2> [, <max_distance> ] )
Arguments¶
Required:
string_expr1
, .string_expr2
The input strings.
Optional:
max_distance
Integer expression that specifies the maximum distance to compute.
When the distance between the strings exceeds this number, the function stops computing the distance and just returns the maximum distance.
Specifying this argument has the same effect as calling
LEAST( EDITDISTANCE( string_expr1, string_expr2 ), max_distance )
.If you specify a negative number (i.e.
-n
), the function uses0
as the maximum distance and returns0
.
Usage notes¶
The execution time of the EDITDISTANCE function is proportional to the product of the lengths of the input strings.
For better performance, Snowflake recommends using input strings not longer than 4096 characters.
You can also use the optional
max_distance
argument to set an upper bound for the distance computed.
Collation details¶
No impact. In languages where the alphabet contains digraphs or trigraphs (such as “Dz” and “Dzs” in Hungarian), each character in each digraph and trigraph is treated as an independent character, not as part of a single multi-character letter.
The result is based solely on the characters in the strings, not on the collation specifications of the strings.
Examples¶
The following example computes the distance between the strings in the columns s
and t
in the table ed
.
The last two columns use the max_distance
argument to specify the maximum distance to compute:
When
max_distance
is3
, the function returns3
if the distance between the strings is greater than or equal to 3 (as shown below).If
max_distance
is a negative number (e.g.-1
, as shown below), the function uses0
as the maximum distance and returns0
.
SELECT s, t, EDITDISTANCE(s, t), EDITDISTANCE(t, s), EDITDISTANCE(s, t, 3), EDITDISTANCE(s, t, -1) FROM ed;
----------------+-----------------+--------------------+--------------------+-----------------------+------------------------+
S | T | EDITDISTANCE(S, T) | EDITDISTANCE(T, S) | EDITDISTANCE(S, T, 3) | EDITDISTANCE(S, T, -1) |
----------------+-----------------+--------------------+--------------------+-----------------------+------------------------|
| | 0 | 0 | 0 | 0 |
Gute nacht | Ich weis nicht | 8 | 8 | 3 | 0 |
Ich weiß nicht | Ich wei? nicht | 1 | 1 | 1 | 0 |
Ich weiß nicht | Ich weiss nicht | 2 | 2 | 2 | 0 |
Ich weiß nicht | [NULL] | [NULL] | [NULL] | [NULL] | [NULL] |
Snowflake | Oracle | 7 | 7 | 3 | 0 |
święta | swieta | 2 | 2 | 2 | 0 |
[NULL] | | [NULL] | [NULL] | [NULL] | [NULL] |
[NULL] | [NULL] | [NULL] | [NULL] | [NULL] | [NULL] |
----------------+-----------------+--------------------+--------------------+-----------------------+------------------------+
The next example returns FALSE
if the distance between two strings is at least 2. Because max_distance
is
specified as 2
, the function stops calculating the distance once the distance is determined to be at least 2. (The actual
distance between the input strings is 6.)
SELECT EDITDISTANCE('future', 'past', 2) < 2;
+---------------------------------------+
| EDITDISTANCE('FUTURE', 'PAST', 2) < 2 |
|---------------------------------------|
| False |
+---------------------------------------+