Skip to content

Commit

Permalink
Merge branch 'branch-24.10' into bug/dask-array-roundtrip
Browse files Browse the repository at this point in the history
  • Loading branch information
@rjzamora
rjzamora authored Sep 24, 2024
2 parents 57c416f + f8db575 commit fdfd93a
Show file tree
Hide file tree
Showing 5 changed files with 99 additions and 14 deletions.
6 changes: 3 additions & 3 deletions ci/release/update-version.sh
View file
Original file line number Diff line number Diff line change
Expand Up @@ -25,9 +25,9 @@ NEXT_PATCH=$(echo $NEXT_FULL_TAG | awk '{split($0, a, "."); print a[3]}')
NEXT_SHORT_TAG=${NEXT_MAJOR}.${NEXT_MINOR}

# Need to distutils-normalize the versions for some use cases
CURRENT_SHORT_TAG_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${CURRENT_SHORT_TAG}'))")
NEXT_SHORT_TAG_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${NEXT_SHORT_TAG}'))")
PATCH_PEP440=$(python -c "from setuptools.extern import packaging; print(packaging.version.Version('${NEXT_PATCH}'))")
CURRENT_SHORT_TAG_PEP440=$(python -c "from packaging.version import Version; print(Version('${CURRENT_SHORT_TAG}'))")
NEXT_SHORT_TAG_PEP440=$(python -c "from packaging.version import Version; print(Version('${NEXT_SHORT_TAG}'))")
PATCH_PEP440=$(python -c "from packaging.version import Version; print(Version('${NEXT_PATCH}'))")

echo "Preparing release $CURRENT_TAG => $NEXT_FULL_TAG"

Expand Down
28 changes: 28 additions & 0 deletions java/src/main/java/ai/rapids/cudf/JSONOptions.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,8 @@ public final class JSONOptions extends ColumnFilterOptions {
private final boolean allowLeadingZeros;
private final boolean allowNonNumericNumbers;
private final boolean allowUnquotedControlChars;
private final boolean cudfPruneSchema;
private final byte lineDelimiter;

private JSONOptions(Builder builder) {
super(builder);
Expand All @@ -52,6 +54,16 @@ private JSONOptions(Builder builder) {
allowLeadingZeros = builder.allowLeadingZeros;
allowNonNumericNumbers = builder.allowNonNumericNumbers;
allowUnquotedControlChars = builder.allowUnquotedControlChars;
cudfPruneSchema = builder.cudfPruneSchema;
lineDelimiter = builder.lineDelimiter;
}

public boolean shouldCudfPruneSchema() {
return cudfPruneSchema;
}

public byte getLineDelimiter() {
return lineDelimiter;
}

public boolean isDayFirst() {
Expand Down Expand Up @@ -123,6 +135,22 @@ public static final class Builder extends ColumnFilterOptions.Builder<JSONOptio
private boolean mixedTypesAsStrings = false;
private boolean keepQuotes = false;

private boolean cudfPruneSchema = false;
private byte lineDelimiter = '\n';

public Builder withCudfPruneSchema(boolean prune) {
cudfPruneSchema = prune;
return this;
}

public Builder withLineDelimiter(char delimiter) {
if (delimiter > Byte.MAX_VALUE) {
throw new IllegalArgumentException("Only basic ASCII values are supported as line delimiters " + delimiter);
}
lineDelimiter = (byte)delimiter;
return this;
}

/**
* Should json validation be strict or not
*/
Expand Down
36 changes: 31 additions & 5 deletions java/src/main/java/ai/rapids/cudf/Table.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -258,7 +258,9 @@ private static native long readJSON(int[] numChildren, String[] columnNames,
boolean strictValidation,
boolean allowLeadingZeros,
boolean allowNonNumericNumbers,
boolean allowUnquotedControl) throws CudfException;
boolean allowUnquotedControl,
boolean pruneColumns,
byte lineDelimiter) throws CudfException;

private static native long readJSONFromDataSource(int[] numChildren, String[] columnNames,
int[] dTypeIds, int[] dTypeScales,
Expand All @@ -272,6 +274,8 @@ private static native long readJSONFromDataSource(int[] numChildren, String[] co
boolean allowLeadingZeros,
boolean allowNonNumericNumbers,
boolean allowUnquotedControl,
boolean pruneColumns,
byte lineDelimiter,
long dsHandle) throws CudfException;

private static native long readAndInferJSONFromDataSource(boolean dayFirst, boolean lines,
Expand All @@ -284,6 +288,7 @@ private static native long readAndInferJSONFromDataSource(boolean dayFirst, bool
boolean allowLeadingZeros,
boolean allowNonNumericNumbers,
boolean allowUnquotedControl,
byte lineDelimiter,
long dsHandle) throws CudfException;

private static native long readAndInferJSON(long address, long length,
Expand All @@ -297,7 +302,8 @@ private static native long readAndInferJSON(long address, long length,
boolean strictValidation,
boolean allowLeadingZeros,
boolean allowNonNumericNumbers,
boolean allowUnquotedControl) throws CudfException;
boolean allowUnquotedControl,
byte lineDelimiter) throws CudfException;

/**
* Read in Parquet formatted data.
Expand Down Expand Up @@ -1308,6 +1314,10 @@ private static Table gatherJSONColumns(Schema schema, TableWithMeta twm, int emp
* @return the file parsed as a table on the GPU.
*/
public static Table readJSON(Schema schema, JSONOptions opts, File path) {
// only prune the schema if one is provided
boolean cudfPruneSchema = schema.getColumnNames() != null &&
schema.getColumnNames().length != 0 &&
opts.shouldCudfPruneSchema();
try (TableWithMeta twm = new TableWithMeta(
readJSON(schema.getFlattenedNumChildren(), schema.getFlattenedColumnNames(),
schema.getFlattenedTypeIds(), schema.getFlattenedTypeScales(),
Expand All @@ -1321,7 +1331,9 @@ public static Table readJSON(Schema schema, JSONOptions opts, File path) {
opts.strictValidation(),
opts.leadingZerosAllowed(),
opts.nonNumericNumbersAllowed(),
opts.unquotedControlChars()))) {
opts.unquotedControlChars(),
cudfPruneSchema,
opts.getLineDelimiter()))) {

return gatherJSONColumns(schema, twm, -1);
}
Expand Down Expand Up @@ -1404,7 +1416,8 @@ public static TableWithMeta readJSON(JSONOptions opts, HostMemoryBuffer buffer,
opts.strictValidation(),
opts.leadingZerosAllowed(),
opts.nonNumericNumbersAllowed(),
opts.unquotedControlChars()));
opts.unquotedControlChars(),
opts.getLineDelimiter()));
}

/**
Expand All @@ -1426,6 +1439,7 @@ public static TableWithMeta readAndInferJSON(JSONOptions opts, DataSource ds) {
opts.leadingZerosAllowed(),
opts.nonNumericNumbersAllowed(),
opts.unquotedControlChars(),
opts.getLineDelimiter(),
dsHandle));
return twm;
} finally {
Expand Down Expand Up @@ -1465,6 +1479,10 @@ public static Table readJSON(Schema schema, JSONOptions opts, HostMemoryBuffer b
assert len > 0;
assert len <= buffer.length - offset;
assert offset >= 0 && offset < buffer.length;
// only prune the schema if one is provided
boolean cudfPruneSchema = schema.getColumnNames() != null &&
schema.getColumnNames().length != 0 &&
opts.shouldCudfPruneSchema();
try (TableWithMeta twm = new TableWithMeta(readJSON(
schema.getFlattenedNumChildren(), schema.getFlattenedColumnNames(),
schema.getFlattenedTypeIds(), schema.getFlattenedTypeScales(), null,
Expand All @@ -1479,7 +1497,9 @@ public static Table readJSON(Schema schema, JSONOptions opts, HostMemoryBuffer b
opts.strictValidation(),
opts.leadingZerosAllowed(),
opts.nonNumericNumbersAllowed(),
opts.unquotedControlChars()))) {
opts.unquotedControlChars(),
cudfPruneSchema,
opts.getLineDelimiter()))) {
return gatherJSONColumns(schema, twm, emptyRowCount);
}
}
Expand All @@ -1505,6 +1525,10 @@ public static Table readJSON(Schema schema, JSONOptions opts, DataSource ds) {
*/
public static Table readJSON(Schema schema, JSONOptions opts, DataSource ds, int emptyRowCount) {
long dsHandle = DataSourceHelper.createWrapperDataSource(ds);
// only prune the schema if one is provided
boolean cudfPruneSchema = schema.getColumnNames() != null &&
schema.getColumnNames().length != 0 &&
opts.shouldCudfPruneSchema();
try (TableWithMeta twm = new TableWithMeta(readJSONFromDataSource(schema.getFlattenedNumChildren(),
schema.getFlattenedColumnNames(), schema.getFlattenedTypeIds(), schema.getFlattenedTypeScales(),
opts.isDayFirst(),
Expand All @@ -1518,6 +1542,8 @@ public static Table readJSON(Schema schema, JSONOptions opts, DataSource ds, int
opts.leadingZerosAllowed(),
opts.nonNumericNumbersAllowed(),
opts.unquotedControlChars(),
cudfPruneSchema,
opts.getLineDelimiter(),
dsHandle))) {
return gatherJSONColumns(schema, twm, emptyRowCount);
} finally {
Expand Down
24 changes: 19 additions & 5 deletions java/src/main/native/src/TableJni.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1627,6 +1627,7 @@ Java_ai_rapids_cudf_Table_readAndInferJSONFromDataSource(JNIEnv* env,
jboolean allow_leading_zeros,
jboolean allow_nonnumeric_numbers,
jboolean allow_unquoted_control,
jbyte line_delimiter,
jlong ds_handle)
{
JNI_NULL_CHECK(env, ds_handle, "no data source handle given", 0);
Expand All @@ -1646,8 +1647,10 @@ Java_ai_rapids_cudf_Table_readAndInferJSONFromDataSource(JNIEnv* env,
.normalize_single_quotes(static_cast<bool>(normalize_single_quotes))
.normalize_whitespace(static_cast<bool>(normalize_whitespace))
.mixed_types_as_string(mixed_types_as_string)
.delimiter(static_cast<char>(line_delimiter))
.strict_validation(strict_validation)
.keep_quotes(keep_quotes);
.keep_quotes(keep_quotes)
.prune_columns(false);
if (strict_validation) {
opts.numeric_leading_zeros(allow_leading_zeros)
.nonnumeric_numbers(allow_nonnumeric_numbers)
Expand Down Expand Up @@ -1676,7 +1679,8 @@ Java_ai_rapids_cudf_Table_readAndInferJSON(JNIEnv* env,
jboolean strict_validation,
jboolean allow_leading_zeros,
jboolean allow_nonnumeric_numbers,
jboolean allow_unquoted_control)
jboolean allow_unquoted_control,
jbyte line_delimiter)
{
JNI_NULL_CHECK(env, buffer, "buffer cannot be null", 0);
if (buffer_length <= 0) {
Expand All @@ -1700,6 +1704,8 @@ Java_ai_rapids_cudf_Table_readAndInferJSON(JNIEnv* env,
.normalize_whitespace(static_cast<bool>(normalize_whitespace))
.strict_validation(strict_validation)
.mixed_types_as_string(mixed_types_as_string)
.prune_columns(false)
.delimiter(static_cast<char>(line_delimiter))
.keep_quotes(keep_quotes);
if (strict_validation) {
opts.numeric_leading_zeros(allow_leading_zeros)
Expand Down Expand Up @@ -1814,6 +1820,8 @@ Java_ai_rapids_cudf_Table_readJSONFromDataSource(JNIEnv* env,
jboolean allow_leading_zeros,
jboolean allow_nonnumeric_numbers,
jboolean allow_unquoted_control,
jboolean prune_columns,
jbyte line_delimiter,
jlong ds_handle)
{
JNI_NULL_CHECK(env, ds_handle, "no data source handle given", 0);
Expand Down Expand Up @@ -1848,8 +1856,10 @@ Java_ai_rapids_cudf_Table_readJSONFromDataSource(JNIEnv* env,
.normalize_single_quotes(static_cast<bool>(normalize_single_quotes))
.normalize_whitespace(static_cast<bool>(normalize_whitespace))
.mixed_types_as_string(mixed_types_as_string)
.delimiter(static_cast<char>(line_delimiter))
.strict_validation(strict_validation)
.keep_quotes(keep_quotes);
.keep_quotes(keep_quotes)
.prune_columns(prune_columns);
if (strict_validation) {
opts.numeric_leading_zeros(allow_leading_zeros)
.nonnumeric_numbers(allow_nonnumeric_numbers)
Expand Down Expand Up @@ -1908,7 +1918,9 @@ JNIEXPORT jlong JNICALL Java_ai_rapids_cudf_Table_readJSON(JNIEnv* env,
jboolean strict_validation,
jboolean allow_leading_zeros,
jboolean allow_nonnumeric_numbers,
jboolean allow_unquoted_control)
jboolean allow_unquoted_control,
jboolean prune_columns,
jbyte line_delimiter)
{
bool read_buffer = true;
if (buffer == 0) {
Expand Down Expand Up @@ -1957,8 +1969,10 @@ JNIEXPORT jlong JNICALL Java_ai_rapids_cudf_Table_readJSON(JNIEnv* env,
.normalize_single_quotes(static_cast<bool>(normalize_single_quotes))
.normalize_whitespace(static_cast<bool>(normalize_whitespace))
.mixed_types_as_string(mixed_types_as_string)
.delimiter(static_cast<char>(line_delimiter))
.strict_validation(strict_validation)
.keep_quotes(keep_quotes);
.keep_quotes(keep_quotes)
.prune_columns(prune_columns);
if (strict_validation) {
opts.numeric_leading_zeros(allow_leading_zeros)
.nonnumeric_numbers(allow_nonnumeric_numbers)
Expand Down
19 changes: 18 additions & 1 deletion java/src/test/java/ai/rapids/cudf/TableTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,6 @@
import org.apache.parquet.schema.GroupType;
import org.apache.parquet.schema.MessageType;
import org.apache.parquet.schema.OriginalType;
import org.junit.jupiter.api.Tag;
import org.junit.jupiter.api.Test;

import java.io.*;
Expand Down Expand Up @@ -656,6 +655,24 @@ void testJSONValidationUnquotedControl() {
}
}

private static final byte[] CR_JSON_TEST_BUFFER = ("{\"a\":\"12\n3\"}\0" +
"{\"a\":\"AB\nC\"}\0").getBytes(StandardCharsets.UTF_8);

@Test
void testReadJSONDelim() {
Schema schema = Schema.builder().addColumn(DType.STRING, "a").build();
JSONOptions opts = JSONOptions.builder()
.withLines(true)
.withLineDelimiter('\0')
.build();
try (Table expected = new Table.TestBuilder()
.column("12\n3", "AB\nC")
.build();
Table found = Table.readJSON(schema, opts, CR_JSON_TEST_BUFFER)) {
assertTablesAreEqual(expected, found);
}
}

private static final byte[] NESTED_JSON_DATA_BUFFER = ("{\"a\":{\"c\":\"C1\"}}\n" +
"{\"a\":{\"c\":\"C2\", \"b\":\"B2\"}}\n" +
"{\"d\":[1,2,3]}\n" +
Expand Down

0 comments on commit fdfd93a

Please sign in to comment.