/**
* Makes a copy of a vertex neighbour list file with only vertices in given list. Reads in a file
* (typically an vertex neighbour list) and outputs a copy except for any lines whose first entry
* is not in given vertex list.
*
* <p>Each line has entries separated by whitespace. Comment lines are always copied.
*
* @param vertexNameToIndex map from vertex name to an index
* @param fullInputFileName full filename of input file to be copied.
* @param fullOutputFileName full filename of output file
* @param fullOtherInputFileName full filename of second input file needed for synchronisation
* @param cc if, after any white space, first word starts with this string then line is treated as
* a comment line.
* @param sep sepation character, e.g. tab
* @param vertexColumnList list of columns to be compared
* @param convertToIndex convert the vertex names to an index as given by the map provided
* @param forceLowerCase force all strings to be lower case
* @param infoOn true if want more info printed out
*/
public static int synchroniseVertexList(
Map<String, Integer> vertexNameToIndex,
String fullInputFileName,
String fullOutputFileName,
String fullOtherInputFileName,
String cc,
String sep,
Set<Integer> vertexColumnList,
boolean convertToIndex,
boolean forceLowerCase,
boolean infoOn) {
int maxColumn = -1;
for (Integer vertexColumn : vertexColumnList) {
if (vertexColumn < 1)
throw new RuntimeException(
"first column is numbered 1, column with vertices given as " + vertexColumn);
maxColumn = Math.max(vertexColumn, maxColumn);
}
int res = 0; // error code.
boolean dontTestForCommentLine = false;
if (cc.length() == 0) dontTestForCommentLine = true;
// set up input file
TextReader data = FileInput.openFile(fullInputFileName);
if (data == null) {
System.err.println(
"*** synchroniseVertexList input file " + fullInputFileName + " not opened");
return -1;
}
if (infoOn) {
System.out.print("Starting to do vertex synchronisation on columns ");
for (Integer vertexColumn : vertexColumnList) System.out.print(vertexColumn + ", ");
System.out.println(" of file " + fullInputFileName);
if (convertToIndex) System.out.println("Converting names to indices");
}
// Make a copy of vertices and note when they have been found.
TreeSet<String> vertexFoundLL = new TreeSet();
TreeSet<String> vertexNotFound = new TreeSet(); // vertices in vertexNameToIndex not in file
// set up output file
PrintStream PS;
FileOutputStream fout;
try {
fout = new FileOutputStream(fullOutputFileName);
PS = new PrintStream(fout);
} catch (FileNotFoundException e) {
throw new RuntimeException(
"**** synchroniseVertexList output file "
+ fullOutputFileName
+ " not opened, "
+ e.getMessage());
// return -2;
}
int linenumber = 0;
int verticesInput = 0;
int verticesOutput = 0;
String[] vertexFound = new String[vertexColumnList.size()];
try {
// Read the data from the input file.
String[] numbers = new String[1000];
int column = 0; // number of columns on each line
String vertex;
while (data.eof() == false) { // Read until end-of-file.
linenumber++;
column = 0;
// Read until end-of-line.
if (forceLowerCase)
while (data.eoln() == false) numbers[column++] = data.getWord().toLowerCase();
else while (data.eoln() == false) numbers[column++] = data.getWord();
if (maxColumn > column)
throw new RuntimeException(
"on line "
+ linenumber
+ " found "
+ column
+ " columns (first is numbered 1) but synchronising on column "
+ maxColumn);
try {
// next tests for first word starting for comment line string but only if this is a
// nontrivial string
if (dontTestForCommentLine || !numbers[0].startsWith(cc)) {
boolean lineOK = true;
int vc = 0;
for (Integer vertexColumn : vertexColumnList) {
vertex = numbers[vertexColumn - 1];
verticesInput++;
if (!vertexNameToIndex.containsKey(vertex)) {
lineOK = false;
}
vertexFound[vc++] = vertex;
}
if (!lineOK) continue; // the while statement
// try to use vertexFoundLL.addAll; verticesOutput+=vertexColumnList.length;
for (vc = 0; vc < vertexFound.length; vc++) {
verticesOutput++;
vertexFoundLL.add(vertexFound[vc]);
} // eo for vc
} // eo if (dontTestForCommentLine ...
// Now write this line out to output file.
// This will happen if its a comment line or if all vertices in specified columns are in
// the given vertex list
int columnMinusOne = column - 1; // this will be the index of the last column
if (convertToIndex)
for (int c = 0; c < column; c++) {
PS.print(
((vertexColumnList.contains(c)) ? vertexNameToIndex.get(numbers[c]) : numbers[c])
+ (c == columnMinusOne ? "" : sep));
}
else {
for (int c = 0; c < column; c++)
PS.print(numbers[c] + (c == columnMinusOne ? "" : sep));
}
PS.println();
} // eo try
catch (RuntimeException e) {
throw new RuntimeException(
"*** PROBLEM on line " + linenumber + " of input file, " + e.getMessage());
}
} // eofile
for (String name : vertexNameToIndex.keySet())
if (!vertexFoundLL.contains(name)) vertexNotFound.add(name);
if (infoOn) {
System.out.println(
"Finished vertex synchronisation on first column of file "
+ fullInputFileName
+ " producing "
+ fullOutputFileName);
System.out.println(
" "
+ linenumber
+ " input lines found "
+ verticesInput
+ " lines with vertices, wrote "
+ verticesOutput
+ " lines with vertices in given list.");
System.out.println(
" Given "
+ vertexNameToIndex.size()
+ " distinct input vertices, "
+ vertexNotFound.size()
+ " were not used.");
System.out.println(" Output " + vertexFoundLL.size() + " distinct output vertices.");
}
} // eo try
catch (TextReader.Error e) {
// Some problem reading the data from the input file.
res = -3;
throw new RuntimeException(
"*** File Error in "
+ fullInputFileName
+ " or "
+ fullOutputFileName
+ ", "
+ e.getMessage());
} finally {
// Finish by closing the files,
// whatever else may have happened.
try {
data.close();
fout.close();
} catch (IOException e) {
throw new RuntimeException(
"*** File Error closing "
+ fullInputFileName
+ " or "
+ fullOutputFileName
+ ", "
+ e.getMessage());
}
} // eo finally
// write out information file
String infoOutputFileName = fullOutputFileName + ".info.txt";
try {
fout = new FileOutputStream(infoOutputFileName);
PS = new PrintStream(fout);
} catch (FileNotFoundException e) {
res = -2;
throw new RuntimeException(
"**** synchroniseVertexList information file "
+ infoOutputFileName
+ " not opened, "
+ e.getMessage());
}
PS.println(
"Vertex synchronisation on first column of file "
+ fullInputFileName
+ "against file "
+ fullOtherInputFileName
+ " producing "
+ fullOutputFileName);
PS.println(
" "
+ linenumber
+ " input lines found "
+ verticesInput
+ " lines with vertices, wrote "
+ verticesOutput
+ " lines with vertices in given list.");
PS.println(
" Given "
+ vertexNameToIndex.size()
+ " distinct input vertices, "
+ vertexNotFound.size()
+ " were not used.");
PS.println(" Output " + vertexFoundLL.size() + " distinct output vertices.");
if (vertexNotFound.size() > 0) {
PS.println("Vertices in input list not found in synchronised file");
PS.println("name" + sep + "index");
for (String name : vertexNotFound) PS.println(name + sep + vertexNameToIndex.get(name));
}
try {
} // eo try
catch (RuntimeException e) {
res = -4;
throw new RuntimeException(
"*** File Error in information output file "
+ infoOutputFileName
+ ", "
+ e.getMessage());
} finally {
// Finish by closing the files,
// whatever else may have happened.
try {
fout.close();
} catch (IOException e) {
throw new RuntimeException(
"*** File Error closing " + infoOutputFileName + ", " + e.getMessage());
}
}
return res;
}
/**
* Read in list two columns, title then ISSN, strings separated by white space.
*
* <p>Use <tt>(String[]) FileInputreadStringList(fullFileName).toArray()</tt> to get array of
* strings instead of an ArrayList.
*
* @param fullFileName name of file including directories
* @param infoLevel 0 = normal, 2= debugging, -2 = silent
* @return list of journals found.
*/
public TreeSet<Journal> readSimpleJournalData(String fullFileName, int infoLevel) {
TextReader tr = ProcessScopusJournalLists.openFile(fullFileName);
if (tr == null) return null;
if (infoLevel > -2) System.out.println("Starting to read list of strings from " + fullFileName);
ArrayList<String> words = new ArrayList();
TreeSet<Journal> journalList = new TreeSet();
String[] labelList = {JournalTitleLabel, this.ISSNLabel};
int rowNumber = 0;
try {
String[] column;
// first find header row and identify columns needed
String header;
// column = line.split("\\t+"); // split at every tab
int[] columnIndex = null;
while (tr.eof() == false && columnIndex == null) {
rowNumber++;
header = tr.getln();
column = header.split("\\t+"); // split at every tab
columnIndex = testLabelRow(column, labelList);
}
if (columnIndex == null)
throw new RuntimeException("*** no header columns found in fullFileName");
if (infoLevel > -1) System.out.println("... header in row " + rowNumber);
if (infoLevel > 0)
for (int c = 0; c < columnIndex.length; c++)
System.out.println(labelList[c] + " in column " + columnIndex[c]);
// now process main data
String line;
String ISSN;
String title;
Journal journal;
while (tr.eof() == false) {
rowNumber++;
line = tr.getln();
column = line.split("\\t+"); // split at every tab
ISSN = (column.length > 1 ? column[columnIndex[1]] : Journal.SUNSET);
title = (column.length > 0 ? column[columnIndex[0]] : Journal.SUNSET);
journal = new Journal(title, ISSN);
journalList.add(journal);
if (infoLevel > 1) System.out.println(rowNumber + " j=" + title + ", n=" + ISSN);
}
if (infoLevel > -2)
System.out.println(
"Finished reading journals from file "
+ fullFileName
+ " found "
+ journalList.size()
+ " journals");
} // eo try
catch (TextReader.Error e) {
// Some problem reading the noNamedata from the input file.
throw new RuntimeException(
"*** Input Error: readJournalData failed after "
+ journalList.size()
+ " journals, row "
+ rowNumber
+ ", "
+ e.getMessage());
} finally {
tr.close();
}
return journalList;
}