Skip to content

feat: add all-banks and sorting options to hipo-diff #1060

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
baltzell wants to merge 4 commits into
base: development
Choose a base branch
from
+176 −101
Open

feat: add all-banks and sorting options to hipo-diff #1060

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ab4cee6
hipo-diff: do all banks, add sorting
baltzell Jan 13, 2026
327699a
cleanup
baltzell Jan 13, 2026
35d7400
cleanup
baltzell Jan 13, 2026
b9c1fe4
cleanup
baltzell Jan 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
277 changes: 176 additions & 101 deletions common-tools/clas-io/src/main/java/org/jlab/utils/HipoDiff.java
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,142 +1,217 @@
package org.jlab.utils;

import java.util.ArrayList;
import java.util.HashMap;
import org.jlab.jnp.hipo4.io.HipoReader;
import org.jlab.jnp.hipo4.data.Bank;
import org.jlab.jnp.hipo4.data.Event;
import org.jlab.jnp.hipo4.data.Schema;
import java.util.HashMap;
import org.jlab.jnp.hipo4.data.SchemaFactory;
import org.jlab.utils.options.OptionParser;

public class HipoDiff {

public static void main(String args[]) {
/**
* Bank sortable by any integer columns.
*/
static class SortedBank extends Bank {
SortedBank(Schema s) { super(s); }
/**
* @param index the bank column indices to sort on
* @return the sorted row indices
*/
int[] getSorted(int... index) {
int[] rows = new int[getRows()];
for (int row = 0; row < rows.length; row++) rows[row] = row;
// bubble sort:
for (int i = 0; i < rows.length - 1; i++) {
for (int j = 0; j < rows.length - i - 1; j++) {
for (int idx : index) {
if (idx >= this.getSchema().getElements()) break;
int x1 = getInt(idx, rows[j]);
int x2 = getInt(idx, rows[j + 1]);
if (x1 > x2) {
int tmp = rows[j];
rows[j] = rows[j + 1];
rows[j + 1] = tmp;
break;
}
else if (x1 < x2) break;
}
}
}
return rows;
}
}

static int nrow = 0;
static int nevent = -1;
static int nentry = 0;
static int nbadevent = 0;
static int nbadrow = 0;
static int nbadentry = 0;
static int[] sortIndex = null;

static int nmax;
static double tolerance;
static boolean verboseMode;
static boolean quietMode;
static Bank runConfigBank;

static ArrayList<SortedBank> banksA = new ArrayList<>();
static ArrayList<SortedBank> banksB = new ArrayList<>();
static HashMap<String, HashMap<String, Integer>> badEntries = new HashMap<>();

public static void main(String args[]) {
OptionParser op = new OptionParser("hipo-diff");
op.addOption("-t", "0.00001", "absolute tolerance for comparisons");
op.addOption("-n", "-1", "number of events");
op.addRequired("-b", "name of bank to diff");
op.addOption("-q", null, "quiet mode");
op.addOption("-Q", null, "verbose mode");
op.addOption("-b", null, "name of bank to diff");
op.addOption("-s", null, "sort on column index");
op.setRequiresInputList(true);
op.parse(args);
if (op.getInputList().size() != 2) {
System.out.println(op.getUsageString());
System.out.println("ERROR: Exactly 2 input files are required.");
System.exit(1);
}

final String bankName = op.getOption("-b").stringValue();
final double tolerance = op.getOption("-t").doubleValue();
final int nmax = op.getOption("-n").intValue();
if (op.getOption("-s").stringValue() != null) {
String[] stmp = op.getOption("-s").stringValue().split(",");
sortIndex = new int[stmp.length];
for (int i = 0; i < stmp.length; i++) sortIndex[i] = Integer.parseInt(stmp[i]);
}
verboseMode = op.getOption("-Q").stringValue() != null;
quietMode = op.getOption("-q").stringValue() != null;
nmax = op.getOption("-n").intValue();
tolerance = op.getOption("-t").doubleValue();

HipoReader readerA = new HipoReader();
HipoReader readerB = new HipoReader();
readerA.open(op.getInputList().get(0));
readerB.open(op.getInputList().get(1));
SchemaFactory sf = readerA.getSchemaFactory();
runConfigBank = new Bank(sf.getSchema("RUN::config"));

Schema schema = readerA.getSchemaFactory().getSchema(bankName);
Bank bankA = new Bank(schema);
Bank bankB = new Bank(schema);

Bank runConfigBank = new Bank(readerA.getSchemaFactory().getSchema("RUN::config"));
Event event = new Event();

int nevent = -1;
int nrow = 0;
int nentry = 0;
int nbadevent = 0;
int nbadrow = 0;
int nbadentry = 0;
HashMap<String, Integer> badEntries = new HashMap<>();

while (readerA.hasNext() && readerB.hasNext() && (nmax<1 || nevent<nmax)) {

if (++nevent % 10000 == 0) {
System.out.println("Analyzed " + nevent + " events");
if (op.getOption("-b").stringValue() == null) {
for (Schema s : sf.getSchemaList()) {
banksA.add(new SortedBank(s));
banksB.add(new SortedBank(s));
}
} else {
banksA.add(new SortedBank(sf.getSchema(op.getOption("-b").stringValue())));
banksB.add(new SortedBank(sf.getSchema(op.getOption("-b").stringValue())));
}

readerA.nextEvent(event);
event.read(bankA);
readerB.nextEvent(event);
event.read(bankB);

event.read(runConfigBank);

if (bankA.getRows() != bankB.getRows()) {
System.out.println("========================= Different number of rows:");
runConfigBank.show();
bankA.show();
bankB.show();
nbadevent++;
System.out.println("=========================");
}
compare(readerA, readerB);
}

else {
for (int i = 0; i < bankA.getRows(); i++) {
boolean mismatch = false;
nrow++;
for (int j = 0; j < schema.getElements(); j++) {
final int type = schema.getType(j);
final String name = schema.getElementName(j);
int element = -1;
String values = "";
nentry++;
switch (type) {
case 1:
if (bankA.getByte(name, i) != bankB.getByte(name, i)) {
element = j;
values += bankA.getByte(name, i) + "/" + bankB.getByte(name, i);
}
break;
case 2:
if (bankA.getShort(name, i) != bankB.getShort(name, i)) {
element = j;
values += bankA.getShort(name, i) + "/" + bankB.getShort(name, i);
}
break;
case 3:
if (bankA.getInt(name, i) != bankB.getInt(name, i)) {
element = j;
values += bankA.getInt(name, i) + "/" + bankB.getInt(name, i);
}
break;
case 4:
if ((!Double.isNaN(bankA.getFloat(name, i)) || !Double.isNaN(bankB.getFloat(name, i)))
&& (!Double.isInfinite(bankA.getFloat(name, i)) || !Double.isInfinite(bankB.getFloat(name, i)))
&& Math.abs(bankA.getFloat(name, i) - bankB.getFloat(name, i)) > tolerance) {
element = j;
values += bankA.getFloat(name, i) + "/" + bankB.getFloat(name, i);
}
break;
}
if (element >= 0) {
System.out.println("mismatch at event " + runConfigBank.getInt("event", 0)
+ " in row " + i
+ " for variable " + name
+ " with values " + values);
mismatch = true;
nbadentry++;
if (badEntries.containsKey(schema.getElementName(element))) {
int nbad = badEntries.get(schema.getElementName(element)) + 1;
badEntries.replace(schema.getElementName(element), nbad);
} else {
badEntries.put(schema.getElementName(element), 1);
}
}
}
if (mismatch) {
nbadrow++;
}
}
}
public static void compare(HipoReader a, HipoReader b) {
Event eventA = new Event();
Event eventB = new Event();
while (a.hasNext() && b.hasNext() && (nmax < 1 || nevent < nmax)) {
if (++nevent % 5000 == 0) System.out.println("Analyzed " + nevent + " events");
a.nextEvent(eventA);
b.nextEvent(eventB);
eventA.read(runConfigBank);
compare(eventA, eventB);
}
System.out.println("\n Analyzed " + nevent + " with " + nbadevent + " bad banks");
System.out.println(nbadrow + "/" + nrow + " mismatched rows");
System.out.println(nbadentry + "/" + nentry + " mismatched entry");
for (String name : badEntries.keySet()) {
System.out.println(name + " " + badEntries.get(name));
}
System.exit(nbadevent + nbadrow + nbadentry);
}

public static void compare(Event a, Event b) {
for (int i = 0; i < banksA.size(); i++) {
a.read(banksA.get(i));
b.read(banksB.get(i));
compare(banksA.get(i), banksB.get(i));
}
}

if (nbadevent + nbadrow + nbadentry > 0) {
System.exit(7);
public static void compare(SortedBank a, SortedBank b) {
if (a.getRows() != b.getRows()) {
System.out.println("========================= Different number of rows:");
runConfigBank.show();
a.show();
b.show();
nbadevent++;
System.out.println("=========================");
return;
}
int[] rowsA = sortIndex == null ? null : a.getSorted(sortIndex);
int[] rowsB = sortIndex == null ? null : b.getSorted(sortIndex);
for (int row = 0; row < a.getRows(); row++) {
boolean mismatch = false;
nrow++;
int rowA = sortIndex == null ? row : rowsA[row];
int rowB = sortIndex == null ? row : rowsB[row];
for (int j = 0; j < a.getSchema().getElements(); j++) {
final int type = a.getSchema().getType(j);
final String name = a.getSchema().getElementName(j);
int element = -1;
String values = "";
nentry++;
switch (type) {
case 1:
if (a.getByte(name, rowA) != b.getByte(name, rowB)) {
element = j;
values += a.getByte(name, rowA) + "/" + b.getByte(name, rowB);
}
break;
case 2:
if (a.getShort(name, rowA) != b.getShort(name, rowB)) {
element = j;
values += a.getShort(name, rowA) + "/" + b.getShort(name, rowB);
}
break;
case 3:
if (a.getInt(name, rowA) != b.getInt(name, rowB)) {
element = j;
values += a.getInt(name, rowA) + "/" + b.getInt(name, rowB);
}
break;
case 4:
if ((!Double.isNaN(a.getFloat(name, rowA)) || !Double.isNaN(b.getFloat(name, rowB)))
&& (!Double.isInfinite(a.getFloat(name, rowA)) || !Double.isInfinite(b.getFloat(name, rowB)))
&& Math.abs(a.getFloat(name, rowA) - b.getFloat(name, rowB)) > tolerance) {
element = j;
values += a.getFloat(name, rowA) + "/" + b.getFloat(name, rowB);
}
break;
}
if (element >= 0) {
if (verboseMode) {
System.out.println("Bank.show " + a.getSchema().getName());
a.show();
b.show();
}
if (!quietMode) {
System.out.println(a.getSchema().getName() + " mismatch at event " + runConfigBank.getInt("event", 0)
+ " in row " + row + " for variable " + name + " with values " + values);
}
mismatch = true;
nbadentry++;
String bankName = a.getSchema().getName();
String elementName = a.getSchema().getElementName(element);
if (!badEntries.containsKey(bankName)) {
badEntries.put(bankName, new HashMap<>());
}
HashMap<String, Integer> m = badEntries.get(bankName);
if (!m.containsKey(elementName)) {
m.put(elementName, 0);
}
m.put(elementName, m.get(elementName) + 1);
}
}
if (mismatch) {
nbadrow++;
}
}
}
}