docs: update usage text with the new stats json cache file

src/cmd/schema.rs

@@ -21,9 +21,9 @@ a `stdin.csv` file will created with stdin's contents as well.
 Note that `stdin.csv` will be overwritten if it already exists.
 
 Schema generation can be a compute-intensive process, especially for large CSV files.
-To speed up generation, the `schema` command will reuse a `stats.csv.bin.sz` file if it
+To speed up generation, the `schema` command will reuse a `stats.csv.data.json` file if it
 exists and is current (i.e. stats generated with --cardinality and --infer-dates options).
-Otherwise, it will run the `stats` command to generate the `stats.csv.bin.sz` file first,
+Otherwise, it will run the `stats` command to generate the `stats.csv.data.json` file first,
 and then use that to generate the schema file.
 
 For examples, see https://github.com/jqnatividad/qsv/blob/master/tests/test_schema.rs.

src/cmd/tojsonl.rs

@@ -9,7 +9,7 @@ It will infer a column as boolean if its cardinality is 2, and the first charact
 the values are one of the following case-insensitive combinations:
   t/f; t/null; 1/0; 1/null; y/n & y/null are treated as true/false.
 
-The `tojsonl` command will reuse a `stats.csv.bin.sz` file if it exists and is current
+The `tojsonl` command will reuse a `stats.csv.data.json` file if it exists and is current
 (i.e. stats generated with --cardinality and --infer-dates options) and will skip
 recomputing stats.