Hiding unfinished cluster command

Author: Yomguithereal

CHANGELOG.md

@@ -4,6 +4,7 @@
 
 *Features*
 
+* Adding `xan bisect`.
 * Adding `xan flatten -N/--non-empty`.
 
 *Fixes*

README.md

@@ -637,7 +637,6 @@ To access the expression language's [cheatsheet](./docs/moonblade/cheatsheet.md)
 
 - [**tokenize**](./docs/cmd/tokenize.md): Tokenize a text column
 - [**vocab**](./docs/cmd/vocab.md): Build a vocabulary over tokenized documents
-- [**cluster**](./docs/cmd/cluster.md): Cluster CSV data to find near-duplicates
 
 *Matrix & network-related commands*
 

docs/cmd/bisect.md

@@ -2,20 +2,37 @@
 # xan bisect
 
 ```txt
-Search for rows where the value in <column> matches <value> using binary search.
-It is assumed that the INPUT IS SORTED according to the specified column.
+Search for rows where the value in <column> matches <value> using binary search,
+and flush all records after the target value.
+The default behavior is similar to a lower_bound bisection, but you can exclude
+records (equivalent to upper_bound) with the target value using the -E/--exclude
+flag. It is assumed that the INPUT IS SORTED according to the specified column.
 The ordering of the rows is assumed to be sorted according ascending lexicographic
 order per default, but you can specify numeric ordering using the -N or --numeric
 flag. You can also reverse the order using the -R/--reverse flag.
+Use the -S/--search flag to only flush records matching the target value instead
+of all records after it.
 
 Usage:
     xan bisect [options] [--] <column> <value> <input>
     xan bisect --help
 
 bisect options:
+    -E, --exclude            When set, the records with the target value will be
+                             excluded from the output. By default, they are
+                             included. Cannot be used with -S/--search.
+                             TODO: not equivalent to upper_bound
     -N, --numeric            Compare according to the numerical value of cells
                              instead of the default lexicographic order.
     -R, --reverse            Reverse sort order, i.e. descending order.
+    -S, --search             Perform a search on the target value instead of
+                             flushing all records after the value (included).
+                             Cannot be used with -E/--exclude nor -e/--end.
+    -e, --end <end-value>    When set, the records after the target value will be
+                             flushed until <end-value> is reached (included).
+                             By default, all records after the target value are
+                             flushed. Cannot be used with -S/--search.
+    -v, --verbose
 
 Common options:
     -h, --help               Display this message

docs/cmd/cluster.md

@@ -50,6 +50,8 @@ flatten options:
     -H, --highlight <pat>  Highlight in red parts of text cells matching given regex
                            pattern. Will not work with -R/--rainbow.
     -i, --ignore-case      If given, pattern given to -H/--highlight will be case-insensitive.
+    -N, --non-empty        For each row, only show non-empty values. This can be useful
+                           when the data is sparse.
 
 Common options:
     -h, --help             Display this message

docs/cmd/flatten.md

@@ -120,7 +120,6 @@ macro_rules! command_list {
 ## Lexicometry & fuzzy matching
     tokenize    Tokenize a text column
     vocab       Build a vocabulary over tokenized documents
-    cluster     Cluster CSV data to find near-duplicates
 
 ## Matrix & network-related commands
     matrix      Convert CSV data to matrix data