should be it

external/duckdb/benchmark/parquet/clickbench_write.benchmark

@@ -0,0 +1,23 @@
+# name: benchmark/parquet/clickbench_write.benchmark
+# description: Write ClickBench data to Parquet
+# group: [parquet]
+
+require httpfs
+
+require parquet
+
+name ClickBench Write Parquet
+group Clickbench
+
+cache clickbench.duckdb
+
+load benchmark/clickbench/queries/load.sql
+
+init
+set preserve_insertion_order=false;
+
+run
+COPY hits TO '${BENCHMARK_DIR}/hits.parquet';
+
+result I
+10000000

external/duckdb/benchmark/parquet/dictionary_read-long-100.benchmark

@@ -0,0 +1,7 @@
+# name: benchmark/parquet/dictionary_read-long-100.benchmark
+# description: Read dictionary-encoded data from Parquet, 100 unique long strings, 10% NULLs
+# group: [parquet]
+
+template benchmark/parquet/dictionary_read.benchmark.in
+UNIQUE_COUNT=100
+PREFIX=veryveryverylongstring_

external/duckdb/benchmark/parquet/dictionary_read-long-1000000.benchmark

@@ -0,0 +1,7 @@
+# name: benchmark/parquet/dictionary_read-long-1000000.benchmark
+# description: Read dictionary-encoded data from Parquet, 1000000 unique long strings, 10% NULLs
+# group: [parquet]
+
+template benchmark/parquet/dictionary_read.benchmark.in
+UNIQUE_COUNT=1000000
+PREFIX=veryveryverylongstring_

external/duckdb/benchmark/parquet/dictionary_read-short-100.benchmark

@@ -0,0 +1,7 @@
+# name: benchmark/parquet/dictionary_read-short-100.benchmark
+# description: Read dictionary-encoded data from Parquet, 100 unique short strings, 10% NULLs
+# group: [parquet]
+
+template benchmark/parquet/dictionary_read.benchmark.in
+UNIQUE_COUNT=100
+PREFIX=''

external/duckdb/benchmark/parquet/dictionary_read-short-1000000.benchmark

@@ -0,0 +1,7 @@
+# name: benchmark/parquet/dictionary_read-short-1000000.benchmark
+# description: Read dictionary-encoded data from Parquet, 1000000 unique short strings, 10% NULLs
+# group: [parquet]
+
+template benchmark/parquet/dictionary_read.benchmark.in
+UNIQUE_COUNT=1000000
+PREFIX=''

external/duckdb/benchmark/parquet/dictionary_read.benchmark.in

@@ -0,0 +1,15 @@
+# name: benchmark/parquet/dictionary_read-short-1000000.benchmark
+# description: Read dictionary-encoded data from Parquet, 1000000 unique long strings, 10% NULLs
+# group: [parquet]
+
+name Parquet Dictionary Read Benchmark
+group parquet
+
+load
+set variable total_rows = 100000000;
+set variable null_rows_factor = 0.1;
+set variable unique_values = ${UNIQUE_COUNT};
+copy (select * from (select '${PREFIX}' || r1.range::varchar v from range(0,getvariable('unique_values')) r1, range(0,((getvariable('total_rows') * (1-getvariable('null_rows_factor'))) // getvariable('unique_values'))::INTEGER) r2 union all select null from range((getvariable('total_rows') * getvariable('null_rows_factor'))::INTEGER) ) order by random()) to '${BENCHMARK_DIR}/dictionary_read.parquet' (dictionary_compression_ratio_threshold 0);
+
+run
+select count(v) from '${BENCHMARK_DIR}/dictionary_read.parquet';