Java tutorial
/* Copyright (C) 2005-2012, by the President and Fellows of Harvard College. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. Dataverse Network - A web application to share, preserve and analyze research data. Developed at the Institute for Quantitative Social Science, Harvard University. Version 3.0. */ package edu.harvard.iq.dvn.ingest.statdataio.impl.plugins.spss; import java.io.*; import java.util.logging.*; import java.util.*; import java.util.regex.Matcher; import java.util.regex.Pattern; import java.security.NoSuchAlgorithmException; import org.apache.commons.lang.*; import edu.harvard.iq.dvn.ingest.org.thedata.statdataio.*; import edu.harvard.iq.dvn.ingest.org.thedata.statdataio.spi.*; import edu.harvard.iq.dvn.ingest.org.thedata.statdataio.metadata.*; import edu.harvard.iq.dvn.ingest.org.thedata.statdataio.data.*; import edu.harvard.iq.dvn.unf.*; import edu.harvard.iq.dvn.ingest.statdataio.impl.plugins.util.*; import java.text.*; /** * A DVN-Project-implementation of <code>StatDataFileReader</code> for the * SPSS Control Card format. * * @author Leonid Andreev * * implemented based on Akio Sone's implementation of the SPSS/SAV format reader * and his older SPSS control card parser implementation in Perl. */ public class SPSSFileReader extends StatDataFileReader { // static fields: private static String[] FORMAT_NAMES = { "spss", "SPSS" }; private static String[] EXTENSIONS = { "spss", "sps" }; private static String[] MIME_TYPE = { "text/plain" }; // date/time data formats private SimpleDateFormat sdf_ymd = new SimpleDateFormat("yyyy-MM-dd"); private SimpleDateFormat sdf_ymdhms = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); private SimpleDateFormat sdf_dhms = new SimpleDateFormat("DDD HH:mm:ss"); private SimpleDateFormat sdf_hms = new SimpleDateFormat("HH:mm:ss"); private static Logger dbgLog = Logger.getLogger(SPSSFileReader.class.getPackage().getName()); private static final Map<String, String> formatCategoryTable = new HashMap<String, String>(); private static String unfVersionNumber = "5"; // global variables: private int caseQnty = 0; private int varQnty = 0; private char delimiterChar; private Map<String, String> commandStrings = new HashMap<String, String>(); private List<String> variableNameList = new ArrayList<String>(); private Map<String, Integer> unfVariableTypes = new HashMap<String, Integer>(); SDIOMetadata smd = new SPSSMetadata(); DataTable csvData = null; SDIOData sdiodata = null; NumberFormat doubleNumberFormatter = new DecimalFormat(); /** * Constructs a <code>SPSSFileReader</code> instance with a * <code>StatDataFileReaderSpi</code> object. * * @param originator a <code>StatDataFileReaderSpi</code> object. */ public SPSSFileReader(StatDataFileReaderSpi originator) { super(originator); init(); } //String[] variableFormatTypeList= null; //Map<String, String> printFormatTable = new LinkedHashMap<String, String>(); //Map<String, String> printFormatNameTable = new LinkedHashMap<String, String>(); static { } private void init() { doubleNumberFormatter.setGroupingUsed(false); doubleNumberFormatter.setMaximumFractionDigits(340); } void setCaseQnty(int cQ) { caseQnty = cQ; } void setVarQnty(int vQ) { varQnty = vQ; } void setDelimiterChar(char c) { delimiterChar = c; } int getCaseQnty() { return caseQnty; } int getVarQnty() { return varQnty; } char getDelimiterChar() { return delimiterChar; } // Methods ---------------------------------------------------------------// /** * Read the given SPSS Control Card via a <code>BufferedInputStream</code> * object. This method calls an appropriate method associated with the given * field header by reflection. * * @param stream a <code>BufferedInputStream</code>. * @return an <code>SDIOData</code> object * @throws java.io.IOException if a reading error occurs. */ @Override public SDIOData read(BufferedInputStream cardStream, File rawDataFile) throws IOException { dbgLog.fine("***** SPSSFileReader: read() start *****"); // TODO: // catch specific exceptions, provide quality diagnostics getSPSScommandLines(cardStream); smd.getFileInformation().put("mimeType", MIME_TYPE[0]); smd.getFileInformation().put("fileFormat", FORMAT_NAMES[0]); smd.getFileInformation().put("varFormat_schema", "SPSS"); // Now we have the control card pre-parsed and the individual // parts of the card ("commands") stored separately. // Now we can go through these parts and evaluate the commands, // which is going to give us the metadata describing the data set. // // These are the parts of the SPSS data definition card that we are // interested in: // DATA LIST // VAR LABELS // VALUE LABELS // FORMATS (?) // MISSING VALUES // RECODE (?) int readStatus = 0; readStatus = read_DataList(commandStrings.get("DataList")); dbgLog.fine("reading DataList. status: " + readStatus); if (readStatus == 0) { throw new IOException("Invalid control card: " + "No variable definitions found"); } readStatus = read_VarLabels(commandStrings.get("VarLabel")); dbgLog.fine("reading VarLabels. status: " + readStatus); readStatus = read_ValLabels(commandStrings.get("ValLabel")); dbgLog.fine("reading ValLabels. status: " + readStatus); readStatus = read_MisValues(commandStrings.get("MisValue")); dbgLog.fine("reading MisValues. status: " + readStatus); // Now read the data file: CSVFileReader csvFileReader = new CSVFileReader(getDelimiterChar()); BufferedReader csvRd = new BufferedReader(new InputStreamReader(new FileInputStream(rawDataFile))); PrintWriter pwout = createOutputWriter(); int casesRead = 0; //csvData = csvFileReader.read(csvRd, smd); try { casesRead = csvFileReader.read(csvRd, smd, pwout); } catch (IOException ex) { dbgLog.info("Could not read and store CSV data file: " + "IO Exception caught:" + ex.getMessage()); throw new IOException( "Could not read and store CSV data file: " + "IO Exception caught:" + ex.getMessage()); } if (casesRead < 1) { dbgLog.info("Could not read and store CSV data file: " + "Empty or corrupted file?"); throw new IOException("Could not read and store CSV data file: " + "Empty or corrupted file?"); } if (getCaseQnty() > 0) { if (getCaseQnty() != casesRead) { dbgLog.info("Could not read and store CSV data file: " + "Number of cases read doesn't match the number " + "specified in the card."); throw new IOException("Could not read and store CSV data file: " + "Number of cases read doesn't match the number " + "specified in the card."); } } else { setCaseQnty(casesRead); smd.getFileInformation().put("caseQnty", getCaseQnty()); dbgLog.fine("Number of cases not specified in the card; using the number " + "of cases read from the CSV file: " + getCaseQnty()); } //storeTabFileData (csvData, pwout); csvData = readTabDataFile(); // Calculate the datasets statistics, summary and category, and // the UNF signatures: calculateDatasetStats(csvData); // Create and return the SDIOData object: sdiodata = new SDIOData(smd, csvData); dbgLog.fine("***** SPSSFileReader: read() end *****"); return sdiodata; } public boolean isValid(File ddiFile) throws IOException { dbgLog.fine("***** SPSSFileReader: validate() start *****"); BufferedInputStream cardStream = new BufferedInputStream(new FileInputStream(ddiFile)); getSPSScommandLines(cardStream); int readStatus = 0; readStatus = read_DataList(commandStrings.get("DataList")); dbgLog.fine("reading DataList. status: " + readStatus); if (readStatus == 0) { throw new IOException("Invalid control card: " + "No variable definitions found"); } return true; } // This method reads the card file and separates the individual parts (commands) // for further parsing. private void getSPSScommandLines(BufferedInputStream cardStream) throws IOException { dbgLog.fine("start dividing SPSS data definition file"); int counter = 0; BufferedReader rd = new BufferedReader(new InputStreamReader(cardStream)); String line = null; String linesCombined = null; String commandHead = ""; boolean commandTerminated = false; List<String> SPSScommands = new ArrayList<String>(); while ((line = rd.readLine()) != null) { // chop all blanks at the end, replace with a single whitespace: line = line.replaceFirst("[ \t\n]*$", " "); // skip blank, and similar lines: if (line.equals(" ") || line.startsWith("comment") || line.startsWith("finish") || line.matches("^[ \t]*.$")) { dbgLog.fine("skipping line"); } else if (line.startsWith(" ") || line.startsWith("\t") || line.startsWith(".")) { // continuation of a wrapped line. if (linesCombined == null) { throw new IOException("Illegal entry: line " + line + " starts with a white space, but does not appear within a defined SPSS command block"); } if (line.matches("^[ \t]*\\. *")) { // command termination SPSScommands.add(linesCombined); linesCombined = null; } else { line = line.replaceAll("^[ \t]*", ""); linesCombined = linesCombined + line; } } else { // a new command line: if (linesCombined != null) { throw new IOException("line " + line + ": bad formatting; (command " + commandHead + " not terminated properly?)"); } if (line.matches("^[Nn]\\w* [Oo]\\w* [Cc].*")) { // looks like the "Number of Cases" command; // special case -- doesn't need to be terminated by // the "." -- which means we can put it directly // on the command stack: SPSScommands.add(line); } else { commandHead = line; linesCombined = commandHead; } } } rd.close(); if (linesCombined != null) { throw new IOException( "Illegal Control Card Syntax: command " + commandHead + " not terminated properly."); } String regexCommandLine = "^(\\w+?)\\s+?(\\w+)(.*)"; Pattern patternCommandLine = Pattern.compile(regexCommandLine); for (int i = 0; i < SPSScommands.size(); i++) { String commandLine = SPSScommands.get(i); // Note that SPSS commands are not case-sensitive. String command1 = null; String command2 = null; String rest = null; Matcher commandMatcher = patternCommandLine.matcher(commandLine); if (commandMatcher.find()) { command1 = commandMatcher.group(1); command2 = commandMatcher.group(2); rest = commandMatcher.group(3); } dbgLog.fine("command1: " + command1); dbgLog.fine("command2: " + command2); dbgLog.fine("rest: " + rest); // TODO: code below executed only if rest != null -- ? // DATA LIST: if (command1 != null && command2 != null && command1.regionMatches(true, 0, "data", 0, 4) && command2.regionMatches(true, 0, "list", 0, 4)) { if (rest != null) { rest = rest.trim(); dbgLog.fine("saving " + rest + " as a DataList command"); if (commandStrings.get("DataList") == null) { commandStrings.put("DataList", rest); } else { commandStrings.put("DataList", commandStrings.get("DataList") + "/" + rest); } } // VARIABLE LABELS: } else if (command1 != null && command2 != null && command1.regionMatches(true, 0, "var", 0, 3) && command2.regionMatches(true, 0, "lab", 0, 3)) { if (rest != null) { rest = rest.trim(); if (rest.length() > 0 && rest.substring(rest.length() - 1).equals(".")) { rest = rest.substring(0, rest.length() - 1); } dbgLog.fine("saving " + rest + " as a VarLabel command"); if (commandStrings.get("VarLabel") == null) { commandStrings.put("VarLabel", rest); } else { commandStrings.put("VarLabel", commandStrings.get("VarLabel") + " " + rest); } } // VALUE LABELS: } else if (command1 != null && command2 != null && command1.regionMatches(true, 0, "val", 0, 3) && command2.regionMatches(true, 0, "lab", 0, 3)) { if (rest != null) { rest = rest.trim(); if (rest.length() > 0 && rest.substring(rest.length() - 1).equals(".")) { rest = rest.substring(0, rest.length() - 2); } if (rest.length() > 0 && rest.substring(rest.length() - 1).equals("/")) { rest = rest.substring(0, rest.length() - 2); } dbgLog.fine("saving " + rest + "/ as a ValLabel command"); if (commandStrings.get("ValLabel") == null) { commandStrings.put("ValLabel", rest + "/"); } else { commandStrings.put("ValLabel", commandStrings.get("ValLabel") + rest + "/"); } } // MISSING VALUES: } else if (command1 != null && command2 != null && command1.regionMatches(true, 0, "mis", 0, 3) && command2.regionMatches(true, 0, "val", 0, 3)) { if (rest != null) { rest = rest.trim(); if (rest.length() > 0 && rest.substring(rest.length() - 1).equals(".")) { rest = rest.substring(0, rest.length() - 2); } // TODO: // Find out if converting these .toUpperCase() is the // right thing to do. dbgLog.fine("saving " + rest.toUpperCase() + " as the " + i + "-th MisValue command"); if (commandStrings.get("MisValue") == null) { commandStrings.put("MisValue", rest); } else { commandStrings.put("MisValue", commandStrings.get("MisValue") + " " + rest.toUpperCase()); } } // NUMBER OF CASES: (optional -- may not be present) } else if (command1 != null && command2 != null && command1.regionMatches(true, 0, "n", 0, 1) && command2.regionMatches(true, 0, "of", 0, 2)) { if (rest != null) { rest = rest.trim(); if (rest.regionMatches(true, 0, "cases", 0, 5)) { rest = rest.substring(5); rest = rest.trim(); String regexNumberOfCases = "^([0-9]*)"; Pattern patternNumberOfCases = Pattern.compile(regexNumberOfCases); Matcher casesMatcher = patternNumberOfCases.matcher(rest); if (casesMatcher.find()) { setCaseQnty(Integer.valueOf(casesMatcher.group(1))); smd.getFileInformation().put("caseQnty", getCaseQnty()); dbgLog.fine("Number of cases found: " + getCaseQnty()); } } } } else { throw new IOException("Unsupported or illegal command: " + command1 + " (" + commandLine + ")"); } // also: // RECODE? // FORMATS? } } // methods for parsing individual parts of the Control Card: // DATA LIST: int read_DataList(String dataListCommand) throws IOException { int readStatus = 0; // Read the first line (DATA LIST ...) to determine // the field separator: // This line should be "/"-terminated (?) dbgLog.fine("dataList command: " + dataListCommand); List<Integer> variableTypeList = new ArrayList<Integer>(); Set<Integer> decimalVariableSet = new HashSet<Integer>(); List<Integer> printFormatList = new ArrayList<Integer>(); // TODO: move Map<String, String> printFormatNameTable = new LinkedHashMap<String, String>(); // TODO: move String delimiterString = null; //String datalistRegex = "^data\\s+list\\s+list\\('(.)'\\)\\s+?/"; String datalistRegex = "^list\\s*\\('(.)'\\)"; Pattern datalistPattern = Pattern.compile(datalistRegex, java.util.regex.Pattern.CASE_INSENSITIVE); Matcher datalistMatcher = datalistPattern.matcher(dataListCommand); if (datalistMatcher.find()) { delimiterString = datalistMatcher.group(1); setDelimiterChar(delimiterString.charAt(0)); dbgLog.fine("found delimiter: " + delimiterString); } else { throw new IOException( "Invalid SPSS Command Syntax: " + "no delimiter specified in the DATA LIST command."); } // Cut off the remaining lines containing the variable definitions: int separatorIndex = dataListCommand.indexOf("/"); if (separatorIndex == -1) { throw new IOException("Invalid SPSS Command Syntax: " + "missing / delimiter on the DATA LIST line."); // No slash found after the first line of the Data List command. } dataListCommand = dataListCommand.substring(separatorIndex + 1); // Parse the variable section. For a delimited file this should be // a list of variable name + data type pairs. // "fortran" type definitions are assumed. dbgLog.fine("parsing " + dataListCommand + " for variable declarations."); int variableCounter = 0; String varName = null; String varType = null; String varDeclarationRegex = "^\\s*(\\S+)\\s+\\((\\S*)\\)"; Pattern varDeclarationPattern = Pattern.compile(varDeclarationRegex); Matcher varDeclarationMatcher = varDeclarationPattern.matcher(dataListCommand); String stringVarDeclarationRegex = "^[aA]([0-9]+)"; Pattern stringVarDeclarationPattern = Pattern.compile(stringVarDeclarationRegex); String numVarDeclarationRegex = "^[fF]([0-9]+)\\.*([0-9]*)"; Pattern numVarDeclarationPattern = Pattern.compile(numVarDeclarationRegex); int endOfMatches = 0; while (varDeclarationMatcher.find()) { varName = varDeclarationMatcher.group(1); varType = varDeclarationMatcher.group(2); endOfMatches = varDeclarationMatcher.end(); dataListCommand = dataListCommand.substring(endOfMatches); varDeclarationMatcher.reset(dataListCommand); dbgLog.fine("found variable " + varName + ", type " + varType); if (varType == null || varType.equals("")) { throw new IOException( "Invalid variable declaration in DATA LIST command: no type specified for variable " + varName + "."); } variableNameList.add(varName); // unfVariableTypes list holds extended type definitions for the // UNF calculation; // we need to be able to differentiate between Integers and // real numbers, in addition to the "numeric" and "string" values. varType = varType.toUpperCase(); if (varType.startsWith("A")) { // String: Matcher stringVarDeclarationMatcher = stringVarDeclarationPattern.matcher(varType); if (stringVarDeclarationMatcher.find()) { variableTypeList.add(new Integer(stringVarDeclarationMatcher.group(1))); } else { // set to default if the string size is not explicitely // specified: variableTypeList.add(1); } formatCategoryTable.put(varName, SPSSConstants.FORMAT_CATEGORY_TABLE.get("A")); unfVariableTypes.put(varName, -1); printFormatList.add(1); //printFormatNameTable.put(varName, "A"); } else if (varType.startsWith("F")) { // "minimal" format value is 0 -- numeric variableTypeList.add(0); formatCategoryTable.put(varName, SPSSConstants.FORMAT_CATEGORY_TABLE.get("F")); if (varType.equals("F")) { // abbreviated numeric type definition; // defaults to f10.0 // for the purposes of the UNF calculations this is an integer: unfVariableTypes.put(varName, 0); printFormatList.add(5); //printFormatNameTable.put(varName, "F10.0"); } else { Matcher numVarDeclarationMatcher = numVarDeclarationPattern.matcher(varType); if (numVarDeclarationMatcher.find()) { Integer numLength = new Integer(numVarDeclarationMatcher.group(1)); Integer numDecimal = 0; String optionalToken = numVarDeclarationMatcher.group(2); if (optionalToken != null && !optionalToken.equals("")) { numDecimal = new Integer(optionalToken); if ((int) numDecimal > 0) { unfVariableTypes.put(varName, 1); decimalVariableSet.add(variableCounter); printFormatNameTable.put(varName, "F" + numLength + "." + numDecimal); } } printFormatList.add(5); // TODO: verify (should it be 0 instead?) } else { // This does not look like a valid numeric type // definition. throw new IOException("Invalid variable declaration in the DATA LIST command: " + "Illegal or unsupported numeric type definition for variable " + varName); } } } else if (varType.matches("^[1-9]$")) { // Another allowed SPSS abbreviation: // type (N) where N is [1-9] means a numeric decimal with // N decimal positions: variableTypeList.add(0); formatCategoryTable.put(varName, SPSSConstants.FORMAT_CATEGORY_TABLE.get("F")); Integer numDecimal = new Integer(varType); unfVariableTypes.put(varName, 1); decimalVariableSet.add(variableCounter); printFormatList.add(5); // TODO: verify (should it be 0 instead?) printFormatNameTable.put(varName, "F10." + numDecimal); // Check for various date and time formats that we support: } else if (SPSSConstants.FORMAT_CATEGORY_TABLE.get(varType) != null) { //if ( SPSSConstants.FORMAT_CATEGORY_TABLE.get(varType).equals("date") // || SPSSConstants.FORMAT_CATEGORY_TABLE.get(varType).equals("time") // || varType.equals("WKDAY") // || varType.equals("MONTH")) { if (varType.equalsIgnoreCase("DATE") || varType.equalsIgnoreCase("DATETIME")) { variableTypeList.add(1); formatCategoryTable.put(varName, SPSSConstants.FORMAT_CATEGORY_TABLE.get(varType)); unfVariableTypes.put(varName, -1); //printFormatList.add(); // TODO: confirm that this isn't needed. printFormatNameTable.put(varName, varType); } else { throw new IOException("Invalid variable declaration in DATA LIST command: " + "unsupported variable type definition for variable " + varName + "."); } } else { // invalid/unrecognized variable type definition. throw new IOException("Invalid variable declaration in DATA LIST command: " + "unknown or illegal type definition for variable " + varName + "."); } variableCounter++; } //dataListCommand = dataListCommand.substring(endOfMatches); // check to see if we've parsed the entire DATA LIST section: if (!dataListCommand.matches("^[ \t\n\r]*$")) { throw new IOException("Illegal control card syntax: " + "this portion of the DATA LIST command could not be parsed: " + dataListCommand); } smd.getFileInformation().put("varQnty", variableCounter); setVarQnty(variableCounter); dbgLog.fine("varQnty=" + getVarQnty()); smd.setVariableName(variableNameList.toArray(new String[variableNameList.size()])); // "minimal" variable types: SPSS type binary definition: // 0 means numeric, >0 means string. smd.setVariableTypeMinimal( ArrayUtils.toPrimitive(variableTypeList.toArray(new Integer[variableTypeList.size()]))); // This is how the "discrete" and "continuous" numeric values are // distinguished in the data set metadata: smd.setDecimalVariables(decimalVariableSet); //TODO: smd.getFileInformation().put("caseWeightVariableName", caseWeightVariableName); dbgLog.info("<<<<<<"); dbgLog.info("printFormatList = " + printFormatList); dbgLog.info("printFormatNameTable = " + printFormatNameTable); dbgLog.info("formatCategoryTable = " + formatCategoryTable); dbgLog.info(">>>>>>"); smd.setVariableFormat(printFormatList); smd.setVariableFormatName(printFormatNameTable); smd.setVariableFormatCategory(formatCategoryTable); //TODO: verify return variableCounter; } // VARIABLE LABELS: int read_VarLabels(String varLabelsCommand) throws IOException { int readStatus = 0; // List of variable labels. // These are variable name + variable label pairs. dbgLog.fine("parsing " + varLabelsCommand + " for variable labels."); Map<String, String> variableLabelMap = new LinkedHashMap<String, String>(); int labelCounter = 0; if (varLabelsCommand != null && !varLabelsCommand.equals("")) { String varName = null; String varLabel = null; String varLabelRegex = "\\s*(\\S+)\\s+\"([^\"]+)\""; Pattern varLabelPattern = Pattern.compile(varLabelRegex); Matcher varLabelMatcher = varLabelPattern.matcher(varLabelsCommand); while (varLabelMatcher.find()) { varName = varLabelMatcher.group(1); varLabel = varLabelMatcher.group(2); dbgLog.fine("found variable label for " + varName + ": " + varLabel); variableLabelMap.put(varName, varLabel); labelCounter++; } } readStatus = labelCounter; // Create empty placeholder labels for the variables that // do not have labels declared: // (should we make variable labels mandatory?) int labeledVariablesVerified = 0; for (int j = 0; j < getVarQnty(); j++) { String variableName = variableNameList.get(j); if (!variableLabelMap.containsKey(variableName)) { variableLabelMap.put(variableName, ""); } else { labeledVariablesVerified++; } } // Make sure that the labels we found are for legit variables: if (labelCounter != labeledVariablesVerified) { throw new IOException("Invalid control card: Variable label(s) defined for undeclared variables."); } smd.setVariableLabel(variableLabelMap); return readStatus; } int read_ValLabels(String valLabelsCommand) { int readStatus = 0; Map<String, Map<String, String>> valueLabelTable = new LinkedHashMap<String, Map<String, String>>(); Map<String, String> valueVariableMappingTable = new LinkedHashMap<String, String>(); // Value Labels are referenced by the (declared) variable names, // followed by a number of value-"label" pairs; for every variable // the entry is terminated with a "/". dbgLog.fine("parsing " + valLabelsCommand + " for value labels."); if (valLabelsCommand == null || valLabelsCommand.equals("")) { return readStatus; } String varName = null; String valLabelDeclaration = null; String varValue = null; String valueLabel = null; int labelCounter = 0; String valLabelRegex = "\\s*(\\S+)\\s+([^/]+)/"; Pattern valLabelPattern = Pattern.compile(valLabelRegex); Matcher valLabelMatcher = valLabelPattern.matcher(valLabelsCommand); String labelDeclarationRegex = "\\s*(\\S+)\\s+(\"[^\"]*\")"; Pattern labelDeclarationPattern = Pattern.compile(labelDeclarationRegex); while (valLabelMatcher.find()) { varName = valLabelMatcher.group(1); valLabelDeclaration = valLabelMatcher.group(2); dbgLog.fine("found value label declaration for " + varName + ": " + valLabelDeclaration); Map<String, String> valueLabelPairs = new LinkedHashMap<String, String>(); Matcher labelDeclarationMatcher = labelDeclarationPattern.matcher(valLabelDeclaration); int localLabelCounter = 0; while (labelDeclarationMatcher.find()) { varValue = labelDeclarationMatcher.group(1); valueLabel = labelDeclarationMatcher.group(2); dbgLog.fine("found label " + valueLabel + " for value " + varValue); Boolean isNumeric = false; // TODO: ! if (isNumeric) { // Numeric variable: dbgLog.fine("processing numeric value label"); valueLabelPairs.put(doubleNumberFormatter.format(new Double(varValue)), valueLabel); } else { // String variable dbgLog.fine("processing string value label"); varValue = varValue.replaceFirst("^[\"']", ""); varValue = varValue.replaceFirst("[\"']$", ""); valueLabelPairs.put(varValue, valueLabel); } localLabelCounter++; } if (localLabelCounter > 0) { valueLabelTable.put(varName, valueLabelPairs); valueVariableMappingTable.put(varName, varName); labelCounter += localLabelCounter; } // TODO: // Do SPSS cards support shared value label sets -- ? } if (labelCounter > 0) { smd.setValueLabelTable(valueLabelTable); smd.setValueLabelMappingTable(valueVariableMappingTable); readStatus = labelCounter; } // TODO: // Better validation, error reporting. return readStatus; } int read_MisValues(String misValuesCommand) { int readStatus = 0; // Missing Values: // These are declared by the (previously declared) variable name, // followed by a comma-separated list of values in parentheses. // for ex.: FOOBAR (1, 2, 3) dbgLog.fine("parsing " + misValuesCommand + " for missing values."); if (misValuesCommand == null || misValuesCommand.equals("")) { return readStatus; } Map<String, List<String>> missingValueTable = new LinkedHashMap<String, List<String>>(); String varName = null; String misValuesDeclaration = null; String misValue = null; int misValueCounter = 0; String misValuesRegex = "\\s*(\\S+)\\s+\\(([^\\)]+)\\)"; Pattern misValuesPattern = Pattern.compile(misValuesRegex); Matcher misValuesMatcher = misValuesPattern.matcher(misValuesCommand); String misValDeclarationRegex = "\\s*([^,]+)\\s*,*"; Pattern misValDeclarationPattern = Pattern.compile(misValDeclarationRegex); while (misValuesMatcher.find()) { varName = misValuesMatcher.group(1); misValuesDeclaration = misValuesMatcher.group(2); int localMisValueCounter = 0; dbgLog.fine("found missing values declaration for " + varName + ": " + misValuesDeclaration); Matcher misValDeclarationMatcher = misValDeclarationPattern.matcher(misValuesDeclaration); Boolean isNumericVariable = false; // TODO: ! List<String> mv = new ArrayList<String>(); while (misValDeclarationMatcher.find()) { misValue = misValDeclarationMatcher.group(1); dbgLog.fine("found missing value: " + misValue); if (isNumericVariable) { // No support for ranges (yet?) // TODO: find out if SPSS cards support ranges of // missing values. mv.add(doubleNumberFormatter.format(new Double(misValue))); } else { misValue = misValue.replaceFirst("^[\"']", ""); misValue = misValue.replaceFirst("[\"']$", ""); mv.add(misValue); } localMisValueCounter++; } if (localMisValueCounter > 0) { missingValueTable.put(varName, mv); misValueCounter += localMisValueCounter; } } // TODO: // Better validation, error reporting. if (misValueCounter > 0) { smd.setMissingValueTable(missingValueTable); readStatus = misValueCounter; } return readStatus; } // method for creating the output writer for the temporary tab file. // this shouldn't be in the plugins really. PrintWriter createOutputWriter() throws IOException { PrintWriter pwout = null; FileOutputStream fileOutTab = null; try { // create a File object to save the tab-delimited data file File tabDelimitedDataFile = File.createTempFile("tempTabfile.", ".tab"); String tabDelimitedDataFileName = tabDelimitedDataFile.getAbsolutePath(); dbgLog.fine("tab file location: " + tabDelimitedDataFileName); // save the temp file name in the metadata object smd.getFileInformation().put("tabDelimitedDataFileLocation", tabDelimitedDataFileName); fileOutTab = new FileOutputStream(tabDelimitedDataFile); pwout = new PrintWriter(new OutputStreamWriter(fileOutTab, "utf8"), true); } catch (FileNotFoundException ex) { ex.printStackTrace(); } catch (UnsupportedEncodingException ex) { ex.printStackTrace(); } catch (IOException ex) { //ex.printStackTrace(); throw ex; } return pwout; } private void storeTabFileData(DataTable csvData, PrintWriter pwout) { String[] caseRow = new String[getVarQnty()]; for (int i = 0; i < getCaseQnty(); i++) { for (int j = 0; j < getVarQnty(); j++) { caseRow[j] = (String) csvData.getData()[j][i]; } pwout.println(StringUtils.join(caseRow, "\t")); } pwout.close(); } private DataTable readTabDataFile() throws IOException { DataTable tabData = new DataTable(); Object[][] dataTable = null; dataTable = new Object[getVarQnty()][getCaseQnty()]; String tabFileName = (String) smd.getFileInformation().get("tabDelimitedDataFileLocation"); BufferedReader tabFileReader = new BufferedReader(new InputStreamReader(new FileInputStream(tabFileName))); boolean[] isCharacterVariable = smd.isStringVariable(); String line; String[] valueTokens = new String[getVarQnty()]; for (int j = 0; j < getCaseQnty(); j++) { if ((line = tabFileReader.readLine()) == null) { throw new IOException( "Failed to read " + getCaseQnty() + " lines " + "from tabular data file " + tabFileName); } valueTokens = line.split("\t", getVarQnty()); for (int i = 0; i < getVarQnty(); i++) { if (isCharacterVariable[i]) { valueTokens[i] = valueTokens[i].replaceFirst("^\"", ""); valueTokens[i] = valueTokens[i].replaceFirst("\"$", ""); dataTable[i][j] = valueTokens[i]; } else { dataTable[i][j] = valueTokens[i]; } } } tabFileReader.close(); tabData.setData(dataTable); return tabData; } // Method for calculating the UNF signatures. // // It really isn't awesome that each of our file format readers has // its own UNF calculation! It should be format-independent; the // method should be defined in just one place, and, preferably, it should // run on the TAB file and the data set metadata from the database. // I.e., it should be reproducible outside of the ingest. // // TODO: bring this up, soon. // // (the 2 methods below are more or less cut-and-pasted as is from Akio's // SPSS/SAV file reader). private String getUNF(Object[] varData, String[] dateFormats, int variableType, String unfVersionNumber, int variablePosition) throws NumberFormatException, UnfException, IOException, NoSuchAlgorithmException { String unfValue = null; dbgLog.fine("variableType=" + variableType); dbgLog.finer("unfVersionNumber=" + unfVersionNumber); dbgLog.fine("variablePosition=" + variablePosition); dbgLog.fine("variableName=" + variableNameList.get(variablePosition)); switch (variableType) { case 0: // Integer (Long): dbgLog.fine("Integer case"); // Convert array of Strings to array of Longs Long[] ldata = new Long[varData.length]; for (int i = 0; i < varData.length; i++) { //if (varData[i] != null) { try { ldata[i] = new Long((String) varData[i]); } catch (Exception ex) { ldata[i] = null; } //} } unfValue = UNF5Util.calculateUNF(ldata); dbgLog.finer("integer:unfValue=" + unfValue); //dbgLog.finer("sumstat:long case=" + Arrays.deepToString( // ArrayUtils.toObject(StatHelper.calculateSummaryStatistics(ldata)))); smd.getSummaryStatisticsTable().put(variablePosition, ArrayUtils.toObject(StatHelper.calculateSummaryStatistics(ldata))); Map<String, Integer> catStat = StatHelper.calculateCategoryStatistics(ldata); smd.getCategoryStatisticsTable().put(variableNameList.get(variablePosition), catStat); break; case 1: // Double: dbgLog.finer("double case"); // Convert array of Strings to array of Doubles Double[] ddata = new Double[varData.length]; for (int i = 0; i < varData.length; i++) { //if (varData[i]!=null) { try { ddata[i] = new Double((String) varData[i]); } catch (Exception ex) { ddata[i] = null; } //} } unfValue = UNF5Util.calculateUNF(ddata); dbgLog.finer("double:unfValue=" + unfValue); smd.getSummaryStatisticsTable().put(variablePosition, ArrayUtils.toObject(StatHelper.calculateSummaryStatisticsContDistSample(ddata))); break; case -1: // String: // // i.e., this is something *stored* as string; it may still be // a more complex data type than just a string of characters. // Namely, it can be some date or time type that we support. // These should be handled differently when calculating the // UNFs. dbgLog.finer("string case"); String[] strdata = Arrays.asList(varData).toArray(new String[varData.length]); dbgLog.finer("string array passed to calculateUNF: " + Arrays.deepToString(strdata)); if (dateFormats != null) { for (int i = 0; i < varData.length; i++) { if (dateFormats[i] != null && (strdata[i].equals("") || strdata[i].equals(" "))) { strdata[i] = null; dateFormats[i] = null; } } unfValue = UNF5Util.calculateUNF(strdata, dateFormats); } else { unfValue = UNF5Util.calculateUNF(strdata); } dbgLog.finer("string:unfValue=" + unfValue); smd.getSummaryStatisticsTable().put(variablePosition, StatHelper.calculateSummaryStatistics(strdata)); Map<String, Integer> StrCatStat = StatHelper.calculateCategoryStatistics(strdata); //out.println("catStat="+StrCatStat); smd.getCategoryStatisticsTable().put(variableNameList.get(variablePosition), StrCatStat); break; default: dbgLog.fine("unknown variable type found"); String errorMessage = "unknow variable Type found at varData section"; throw new IllegalArgumentException(errorMessage); } // switch dbgLog.fine("unfvalue(last)=" + unfValue); return unfValue; } // This method calculates the summary and category statistics, as well // as the UNF signatures for the variables and the dataset as a whole. private void calculateDatasetStats(DataTable csvData) { String fileUNFvalue = null; String[] unfValues = new String[getVarQnty()]; // TODO: // Catch and differentiate between different exception // that the UNF methods throw. for (int j = 0; j < getVarQnty(); j++) { int variableTypeNumer = unfVariableTypes.get(variableNameList.get(j)); String varFormat = smd.getVariableFormatName().get(smd.getVariableName()[j]); try { dbgLog.finer("j = " + j); // Before we pass the variable vector to the UNF calculator, // we need to check if is of any supported date/time type. // If so, we'll also need to create and pass a list of // date formats, so that the UNFs could be properly calculated. // (otherwise the date/time values will be treated simply as // strings!) if (varFormat != null && (varFormat.equals("WKDAY") || varFormat.equals("MONTH") || "date".equals(SPSSConstants.FORMAT_CATEGORY_TABLE.get(varFormat)) || "time".equals(SPSSConstants.FORMAT_CATEGORY_TABLE.get(varFormat)))) { // TODO: // All these date, time, weekday, etc. values need to be validated! String[] dateFormats = new String[getCaseQnty()]; for (int k = 0; k < getCaseQnty(); k++) { if (SPSSConstants.FORMAT_CATEGORY_TABLE.get(varFormat).equals("date")) { dbgLog.finer("date case"); dateFormats[k] = sdf_ymd.toPattern(); } else if (SPSSConstants.FORMAT_CATEGORY_TABLE.get(varFormat).equals("time")) { dbgLog.finer("time case: DTIME or DATETIME or TIME"); if (varFormat.equals("DTIME")) { dateFormats[k] = sdf_dhms.toPattern(); } else if (varFormat.equals("DATETIME")) { dateFormats[k] = sdf_ymdhms.toPattern(); } else if (varFormat.equals("TIME")) { dateFormats[k] = sdf_hms.toPattern(); } } else if (varFormat.equals("WKDAY")) { // TODO: these need to be validated only. dateFormats = null; } else if (varFormat.equals("MONTH")) { // TODO: these need to be validated only. dateFormats = null; } } unfValues[j] = getUNF(csvData.getData()[j], dateFormats, variableTypeNumer, unfVersionNumber, j); } else { unfValues[j] = getUNF(csvData.getData()[j], null, variableTypeNumer, unfVersionNumber, j); } dbgLog.fine(j + "th unf value" + unfValues[j]); } catch (NumberFormatException ex) { ex.printStackTrace(); } catch (UnfException ex) { ex.printStackTrace(); } catch (IOException ex) { ex.printStackTrace(); //throw ex; } catch (NoSuchAlgorithmException ex) { ex.printStackTrace(); } } dbgLog.fine("unf set:\n" + Arrays.deepToString(unfValues)); try { fileUNFvalue = UNF5Util.calculateUNF(unfValues); } catch (NumberFormatException ex) { ex.printStackTrace(); } catch (IOException ex) { ex.printStackTrace(); //throw ex; } // Set the UNFs we have calculated, the ones for the individual // variables and the file-level UNF: csvData.setUnf(unfValues); csvData.setFileUnf(fileUNFvalue); smd.setVariableUNF(unfValues); smd.getFileInformation().put("fileUNF", fileUNFvalue); dbgLog.fine("file-level unf value:\n" + fileUNFvalue); } public static void main(String[] args) { BufferedInputStream spssCardStream = null; SDIOData processedCard = null; SPSSFileReader spssReader = null; String testCardFile = args[0]; String csvRawDataFile = args[1]; try { spssCardStream = new BufferedInputStream(new FileInputStream(testCardFile)); spssReader = new SPSSFileReader(null); processedCard = spssReader.read(spssCardStream, new File(csvRawDataFile)); } catch (IOException ex) { System.out.println("exception caught: "); System.out.println(ex.getMessage()); if (spssReader == null) { System.out.println("failed to create an SPSS file reader."); } } } }