andi/perfdb
1
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Activity

cleanup after revert

Browse Source
This commit is contained in:
Andreas Huber
2021-05-12 18:20:34 +02:00
parent 7adfc7029f
commit ee79cb0022
4 changed files with 0 additions and 387 deletions
Download Patch File Download Diff File Expand all files Collapse all files

127
pdb-ui/src/main/java/org/lucares/pdbui/CustomExportFormatToEntryTransformer.java
View File

@@ -1,127 +0,0 @@
package org.lucares.pdbui;
import java.io.BufferedReader;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.ArrayBlockingQueue;
import java.util.regex.Pattern;
import org.lucares.pdb.api.Tags;
import org.lucares.pdb.api.TagsBuilder;
import org.lucares.pdb.datastore.Entries;
import org.lucares.pdb.datastore.Entry;
import org.lucares.performance.db.PdbExport;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
/**
*
* File format goals: Minimal size/ minimal repetition while also providing a
* file format that can be used for "normal" ingestion, not just backup/restore.
* It should be easy to implement in any language. It should be easy to debug.
* <p>
* Note: Line breaks are written as {@code \n}.
*
* <pre>
* # // # is the magic byte for the file format used to detect this format
* $123:key1=value1,key2=value2\n // $ marks the beginning of a dictionary entry that says: the following number will be used to refer to the following tags.
* // In this case the tags key1=value1,key2=value2 will be identified by 123.
* // The newline is used as an end marker.
* 1534567890,456,123\n // Defines an entry with timestamp 1534567890, duration 456 and tags key1=value1,key2=value2.
* 1,789,123\n // Timestamps are encoded using delta encoding. That means this triple defines
* // an entry with timestamp 1534567891, duration 789 and tags key1=value1,key2=value2
* -2,135,123\n // Timestamp delta encoding can contain negative numbers. This triple defines an entry
* // with timestamp 1534567889, duration 135 and tags key1=value1,key2=value2
* </pre>
*/
public class CustomExportFormatToEntryTransformer {
private static final int ENTRY_BUFFER_SIZE = 100;
private static final Logger LOGGER = LoggerFactory.getLogger(CustomExportFormatToEntryTransformer.class);
private final Pattern splitByComma = Pattern.compile(",");
private final Map<Long, Tags> tagsDictionary = new HashMap<>();
private long lastEpochMilli;
public void read(final BufferedReader in, final ArrayBlockingQueue<Entries> queue) throws IOException {
Entries bufferedEntries = new Entries(ENTRY_BUFFER_SIZE);
try {
String line;
while ((line = in.readLine()) != null) {
try {
if (line.startsWith(PdbExport.MARKER_DICT_ENTRY)) {
readDictionaryEntry(line);
} else {
final Entry entry = readEntry(line);
if (entry != null) {
bufferedEntries.add(entry);
if (bufferedEntries.size() == ENTRY_BUFFER_SIZE) {
queue.put(bufferedEntries);
bufferedEntries = new Entries(ENTRY_BUFFER_SIZE);
}
}
}
} catch (final Exception e) {
LOGGER.error("ignoring line '{}'", line, e);
}
queue.put(bufferedEntries);
bufferedEntries = new Entries(ENTRY_BUFFER_SIZE);
}
} catch (final InterruptedException e) {
Thread.currentThread().interrupt();
LOGGER.info("aborting because of interruption");
}
}
private Entry readEntry(final String line) {
final String[] timeValueTags = splitByComma.split(line);
final long timeDelta = Long.parseLong(timeValueTags[0]);
final long value = Long.parseLong(timeValueTags[1]);
final long tagsId = Long.parseLong(timeValueTags[2]);
lastEpochMilli = lastEpochMilli + timeDelta;
final Tags tags = tagsDictionary.get(tagsId);
if (tags == null) {
LOGGER.info("no tags available for tagsId {}. Ignoring line '{}'", tagsId, line);
return null;
}
return new Entry(lastEpochMilli, value, tags);
}
private void readDictionaryEntry(final String line) {
final String[] tagsIdToSerializedTags = line.split(Pattern.quote(PdbExport.SEPARATOR_TAG_ID));
final Long tagId = Long.parseLong(tagsIdToSerializedTags[0], 1, tagsIdToSerializedTags[0].length(), 10);
final Tags tags = tagsFromCsv(tagsIdToSerializedTags[1]);
tagsDictionary.put(tagId, tags);
}
public static Tags tagsFromCsv(final String line) {
final TagsBuilder tagsBuilder = new TagsBuilder();
final String[] tagsAsString = line.split(Pattern.quote(","));
for (final String tagAsString : tagsAsString) {
final String[] keyValue = tagAsString.split(Pattern.quote("="));
final int key = Tags.STRING_COMPRESSOR.put(keyValue[0]);
final int value = Tags.STRING_COMPRESSOR.put(keyValue[1]);
tagsBuilder.add(key, value);
}
return tagsBuilder.build();
}
}

12
pdb-ui/src/main/java/org/lucares/pdbui/IngestionHandler.java
View File

@@ -18,7 +18,6 @@ import java.util.zip.GZIPInputStream;
import org.lucares.pdb.datastore.Entries;
import org.lucares.pdb.datastore.Entry;
import org.lucares.pdbui.CsvReaderSettings.ColumnDefinitions;
import org.lucares.performance.db.PdbExport;
import com.fasterxml.jackson.core.JsonParseException;
@@ -58,8 +57,6 @@ public final class IngestionHandler implements Callable<Void> {
if (firstByte == '{') {
in.reset();
readJSON(in);
} else if (firstByte == PdbExport.MAGIC_BYTE) {
readCustomExportFormat(in);
} else if (isGZIP(firstByte)) {
in.reset();
final GZIPInputStream gzip = new GZIPInputStream(in);
@@ -80,15 +77,6 @@ public final class IngestionHandler implements Callable<Void> {
return firstByte == 0x1f;
}
private void readCustomExportFormat(final InputStream in) throws IOException {
final CustomExportFormatToEntryTransformer transformer = new CustomExportFormatToEntryTransformer();
final BufferedReader reader = new BufferedReader(new InputStreamReader(in, StandardCharsets.UTF_8));
transformer.read(reader, queue);
}
private void readJSON(final InputStream in) throws IOException, InterruptedException {
final int chunksize = 100;
Entries entries = new Entries(chunksize);