feat: alias read_parquet, read_njson, read_csv to their scan counterp…

…art (#2185)

Added three new functions: `read_parquet`, `read_njson` and `read_csv`.
They do the same as their scan counterparts.
Closes: #2169

Co-authored-by: universalmind303 <cory.grinstead@gmail.com>

crates/sqlbuiltins/src/functions/mod.rs

@@ -34,7 +34,7 @@ use self::iceberg::{IcebergDataFiles, IcebergScan, IcebergSnapshots};
 use self::lance::LanceScan;
 use self::mongo::ReadMongoDb;
 use self::mysql::ReadMysql;
-use self::object_store::{CSV_SCAN, JSON_SCAN, PARQUET_SCAN};
+use self::object_store::{CSV_SCAN, JSON_SCAN, PARQUET_SCAN, READ_CSV, READ_JSON, READ_PARQUET};
 use self::postgres::ReadPostgres;
 use self::snowflake::ReadSnowflake;
 use self::virtual_listing::{ListColumns, ListSchemas, ListTables};
@@ -109,8 +109,11 @@ impl BuiltinTableFuncs {
             Arc::new(ReadSnowflake),
             // Object store
             Arc::new(PARQUET_SCAN),
+            Arc::new(READ_PARQUET),
             Arc::new(CSV_SCAN),
+            Arc::new(READ_CSV),
             Arc::new(JSON_SCAN),
+            Arc::new(READ_JSON),
             // Data lakes
             Arc::new(DeltaScan),
             Arc::new(IcebergScan),

crates/sqlbuiltins/src/functions/object_store.rs

@@ -31,10 +31,13 @@ use protogen::metastore::types::catalog::RuntimePreference;
 use protogen::metastore::types::options::{CredentialsOptions, StorageOptions};
 
 pub const PARQUET_SCAN: ObjScanTableFunc = ObjScanTableFunc(FileType::PARQUET, "parquet_scan");
+pub const READ_PARQUET: ObjScanTableFunc = ObjScanTableFunc(FileType::PARQUET, "read_parquet");
 
 pub const CSV_SCAN: ObjScanTableFunc = ObjScanTableFunc(FileType::CSV, "csv_scan");
+pub const READ_CSV: ObjScanTableFunc = ObjScanTableFunc(FileType::CSV, "read_csv");
 
 pub const JSON_SCAN: ObjScanTableFunc = ObjScanTableFunc(FileType::JSON, "ndjson_scan");
+pub const READ_JSON: ObjScanTableFunc = ObjScanTableFunc(FileType::JSON, "read_ndjson");
 
 #[derive(Debug, Clone)]
 pub struct ObjScanTableFunc(FileType, &'static str);

testdata/sqllogictests/functions/read_csv.slt

@@ -0,0 +1,64 @@
+# Tests `read_csv`
+
+# Absolute path
+query I
+select count(*) from read_csv('file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv')
+----
+102
+
+# Absolute path (compressed)
+
+query I
+select count(*) from read_csv('file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv.gz')
+----
+102
+
+# Compressed (with function argument)
+
+query I
+select count(*) from read_csv(
+  'file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv.gz',
+  compression => 'gzip'
+);
+----
+102
+
+# To prove this actually picks up the compression from the argument, giving a
+# wrong compression type should fail.
+statement error stream/file format not recognized
+select count(*) from read_csv(
+  'file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv.gz',
+  compression => 'xz'
+);
+
+# Relative path
+query I
+select count(*) from read_csv('../../testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv')
+----
+102
+
+# Remote path
+query I
+select count(*) from read_csv('https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv');
+----
+102
+
+# Multiple URLs
+
+query I
+select count(*) from read_csv([
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv',
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.csv'
+]);
+----
+204
+
+statement error at least one url expected
+select * from read_csv([]);
+
+# Glob patterns not supported on HTTP
+
+statement error Note that globbing is not supported for HTTP.
+select * from read_csv(
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/*.csv'
+);

testdata/sqllogictests/functions/read_json.slt

@@ -0,0 +1,52 @@
+# Tests `read_ndjson`
+
+# Absolute path
+query I
+select count(*) from read_ndjson('file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson')
+----
+102
+
+# Absolute path (compressed)
+query I
+select count(*) from read_ndjson('file://${PWD}/testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson.gz')
+----
+102
+
+# # Relative path
+query I
+select count(*) from read_ndjson('../../testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson')
+----
+102
+
+# Remote path
+query I
+select count(*) from read_ndjson('https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson');
+----
+102
+
+
+# Multiple URLs
+
+query I
+select count(*) from read_ndjson([
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson',
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/bikeshare_stations.ndjson'
+]);
+----
+204
+
+statement error at least one url expected
+select * from read_ndjson([]);
+
+# Glob patterns not supported on HTTP
+
+statement error Unexpected status code '404 Not Found' 
+select * from read_ndjson(
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/*.ndjson'
+);
+
+statement error Note that globbing is not supported for HTTP.
+select * from read_ndjson(
+  'https://raw.githubusercontent.com/GlareDB/glaredb/main/testdata/sqllogictests_datasources_common/data/*.ndjson'
+);
+

testdata/sqllogictests/functions/read_parquet.slt

@@ -0,0 +1,58 @@
+# Tests `read_parquet`
+
+# Absolute path
+query I
+select count(*) from read_parquet('file://${PWD}/testdata/parquet/userdata1.parquet')
+----
+1000
+
+# Relative path
+query I
+select count(*) from read_parquet('../../testdata/parquet/userdata1.parquet')
+----
+1000
+
+# Remote path
+query I
+select count(*) from read_parquet('https://github.com/GlareDB/glaredb/raw/main/testdata/parquet/userdata1.parquet');
+----
+1000
+
+# Huggingface (percent encoded paths)
+#
+# Note that this is a pretty big file, but the limit will be pushed down to the
+# exec, ensuring we don't need to load the whole thing.
+query T
+select length(head) > 1 from read_parquet(
+  'https://huggingface.co/datasets/allenai/soda/resolve/refs%2Fconvert%2Fparquet/default/test/0000.parquet'
+) limit 1;
+----
+t
+
+# Multiple URLs
+
+query I
+select count(*) from read_parquet([
+  'https://github.com/GlareDB/glaredb/raw/main/testdata/parquet/userdata1.parquet',
+  'https://github.com/GlareDB/glaredb/raw/main/testdata/parquet/userdata1.parquet'
+]);
+----
+2000
+
+statement error No such file or directory
+select * from read_parquet('./testdata/parquet/userdata1.paruqet');
+
+# Ambiguous name.
+# query I
+# select count(*)
+#   from read_parquet('../../testdata/parquet/userdata1.parquet') p
+#   inner join (values ('Sweden')) as c(country) on p.country = c.country
+# ----
+# 1000
+
+# query I
+# select count(*)
+#   from read_parquet('../../testdata/parquet/userdata1.parquet') p
+#   inner join (select 'Sweden') as c(country) on p.country = c.country
+# ----
+# 1000