How would you improve this api?

[i30817]

I did a abstraction over both zip and rar to find and extract files. It works ok, but my api is not as elegant as i wanted. I tended to use abstract classes instead of interfaces in the implementation.

The improvements on the api would be mostly on the Selector class 2 posts bellow. Notice that i use a interface on some ordering / selection functions that returns a comparator given an Extractor (the abstract class that extracts characteristics or files from compressed file). I thought about using comparator simply, but i needed access to the extraction functions (that are on the Extractor implementation). The api is “strange” because it receives objects that only each comparator knows about, dunno about that.

I actually found a really niffy way to use a enum as a way to aggregate commonly used implementations with this, so extracting / ordering is sweet and easy ( orderByAscending for ex), and i wonder if fresh eyes would see a way to apply/aggregate that pattern to the other functions (there is one there that uses another enum Inequality).

Notice that i tried to implement a self-reflective api (like lists and the sublist method) by implementing the methods inverseSelector() and subSelector() that allow me to do fairly simple set operations like only select from this subset, or only select from other files, etc. I’d also like to know if i missed something obvious here.

First a sample usage:

[i30817]

        Selector archive = null;
        try {
            //may have to download...
            archive = Extractor.from(origin);
            //can't read, ignore
            if (archive == null) {
                throw new IOException("Can't read " + origin.toString() + " as compressed File");
            }
            //find supported formats inside the file
            StringBuilder regex = new StringBuilder();
            Iterator<String> it = BookLoader.setOfSupportedExtensions().iterator();

            if (it.hasNext()) {
                regex.append("(?:.*\\.").append(it.next()).append("$)");
            }

            while (it.hasNext()) {
                regex.append("|(?:.*\\.").append(it.next()).append("$)");
            }

            archive.selectByRegex(regex.toString(), Pattern.CASE_INSENSITIVE);
            FileView fv = archive.getSelectedMax(Content.FileSize);
            if (fv == null) {
                throw new IOException("No parseable file in compressed file");
            }
            BookLoader loader = BookLoader.forName(fv.getFileName());
            return loader.create(fv.getInputStream(), reparse);
        } finally {
            IoUtils.close(archive);
        }

Then the Selector implementation:

[i30817]

@SuppressWarnings(value = "unchecked")
public final class Selector implements Closeable, Iterable<FileView> {

    private List headers;
    private Extractor extractor;
    private Set workSet;

    public Selector(Extractor extractor) {
        this(extractor, extractor.getFileHeaders());
    }

    private Selector(Extractor extractor, List headers) {
        this.headers = headers;
        this.extractor = extractor;
        this.workSet = new LinkedHashSet(headers.size() * 2);
    }

    public boolean isEmpty() {
        return headers.isEmpty();
    }

    public int selectedSize() {
        return workSet.size();
    }

    public int size() {
        return headers.size();
    }

    public void clear() {
        workSet.clear();
    }

    public void close() throws IOException {
        if (extractor == null) {
            return;
        }
        extractor.close();
        headers = null;
        workSet = null;
        extractor = null;
    }

    public Selector subSelector() {
        if (workSet.size() == headers.size()) {
            workSet.clear();
            return this;
        } else if (workSet.size() == 0) {
            return new Selector(extractor, Collections.EMPTY_LIST);
        }
        return new Selector(extractor, new ArrayList(workSet));
    }
    public Selector inverseSelector() {
        if (workSet.size() == 0) {
            return this;
        } else if (workSet.size() == headers.size()) {
            return new Selector(extractor, Collections.EMPTY_LIST);
        }
        //removing in arraylists is painfull
        LinkedList arr = new LinkedList(headers);
        arr.removeAll(workSet);
        return new Selector(extractor, arr);
    }
    public Selector selectAll() {
        for (Object h : headers) {
            workSet.add(h);
        }
        return this;
    }
    public Selector selectByCRC32(long crc32) {
        for (Object h : headers) {
            Long crc = extractor.getCRC32(h);
            if (crc != null && (crc & 0X00000000ffffffffL) == (crc32 & 0X00000000ffffffffL)) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByCompressedFileSize(Inequality i, long size) {
        if (i == null) {
            return this;
        }
        for (Object h : headers) {
            Long fs = extractor.getCompressedFileSize(h);
            if (i.TEST(fs, size)) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByFileSize(Inequality i, long size) {
        if (i == null) {
            return this;
        }
        for (Object h : headers) {
            Long fs = extractor.getFileSize(h);
            if (i.TEST(fs, size)) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByModificationDate(Inequality i, DateFormat format, String datePattern) {
        if (i == null || format == null || datePattern == null) {
            return this;
        }
        try {
            Date givenDate = format.parse(datePattern);
            for (Object h : headers) {
                Date archiveDate = extractor.getModificationDate(h);
                if (archiveDate == null) {
                    continue;
                }
                Date comparableDate = format.parse(format.format(archiveDate));
                if (i.TEST(comparableDate, givenDate)) {
                    workSet.add(h);
                }
            }
        } catch (ParseException ex) {
            throw new IllegalArgumentException("Formatter can't parse given datePattern", ex);
        }
        return this;
    }
    public Selector selectByModificationDate(Inequality i, Date date) {
        if (i == null || date == null) {
            return this;
        }

        for (Object h : headers) {
            Date archiveDate = extractor.getModificationDate(h);
            if (archiveDate == null) {
                continue;
            }
            if (i.TEST(archiveDate, date)) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByName(String name, boolean caseSensitive) {
        if (name == null) {
            return this;
            //Archived name is never null...
        }
        if (caseSensitive) {
            for (Object h : headers) {
                if (extractor.getFileName(h).equals(name)) {
                    workSet.add(h);
                }
            }
        } else {
            for (Object h : headers) {
                if (extractor.getFileName(h).toLowerCase(Locale.ENGLISH).equals(name.toLowerCase(Locale.ENGLISH))) {
                    workSet.add(h);
                }
            }
        }
        return this;
    }
    public Selector selectByName(Inequality i, String s) {
        if (i == null || s == null) {
            return this;
            //name is never null...
        }
        Comparator<String> natural = Strings.getNaturalComparator();
        for (Object h : headers) {
            if (i.TEST(extractor.getFileName(h), s, natural)) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByRegex(String regex) {
        if (regex == null) {
            return this;
            //name never null
        }
        Pattern p = Pattern.compile(regex);
        Matcher m = p.matcher("");
        for (Object h : headers) {
            m.reset(extractor.getFileName(h));
            if (m.matches()) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectByRegex(String regex, int flags) {
        if (regex == null) {
            return this;
            //name never null
        }
        Pattern p = Pattern.compile(regex, flags);
        Matcher m = p.matcher("");
        for (Object h : headers) {
            m.reset(extractor.getFileName(h));
            if (m.matches()) {
                workSet.add(h);
            }
        }
        return this;
    }
    public Selector selectBySuffix(String suffix, boolean caseSensitive) {
        if (suffix == null) {
            return this;
        }
        //Archived name is never null...
        if (caseSensitive) {
            for (Object h : headers) {
                if (extractor.getFilePath(h).endsWith(suffix)) {
                    workSet.add(h);
                }
            }
        } else {
            for (Object h : headers) {
                if (extractor.getFilePath(h).toLowerCase(Locale.ENGLISH).endsWith(suffix.toLowerCase(Locale.ENGLISH))) {
                    workSet.add(h);
                }
            }
        }
        return this;
    }
    public Selector orderByAscending(FileComparator c) {
        Object[] a = workSet.toArray();
        Arrays.sort(a, c.getComparator(extractor));
        workSet = new LinkedHashSet(a.length * 2);
        for (int i = 0; i < a.length; i++) {
            workSet.add(a[i]);
        }
        return this;
    }
    public FileView getSelected() {
        Iterator it = workSet.iterator();
        if (it.hasNext()) {
            return new FileView(it.next(), extractor);
        } else {
            return null;
        }
    }
    public FileView getSelectedMax(FileComparator c) {
        try {
            Object header = Collections.max(workSet, c.getComparator(extractor));
            return new FileView(header, extractor);
        } catch (NoSuchElementException ex) {
            return null;
        }
    }
    public FileView getSelectedMin(FileComparator c) {
        try {
            Object header = Collections.min(workSet, c.getComparator(extractor));
            return new FileView(header, extractor);
        } catch (NoSuchElementException ex) {
            return null;
        }
    }
    public Selector orderByDescending(FileComparator c) {
        Object[] a = workSet.toArray();
        Arrays.sort(a, Collections.reverseOrder(c.getComparator(extractor)));
        workSet = new LinkedHashSet(a.length * 2);
        for (int i = 0; i < a.length; i++) {
            workSet.add(a[i]);
        }
        return this;
    }
    public ContentsIterator iterator() {
        return new ContentsIterator(workSet.iterator(), workSet.size(), extractor);
    }

    public ContentsIterator iterator(int numberToExtract) {
        int toExtract = Math.max(0, Math.min(numberToExtract, workSet.size()));
        return new ContentsIterator(workSet.iterator(), toExtract, extractor);
    }

    public static final class ContentsIterator implements Iterator<FileView>, Iterable<FileView> {

        private Iterator headerCopy;
        private Extractor extractor;
        private int size, index;

        private ContentsIterator(Iterator headerCopy, int size, Extractor extractor) {
            this.headerCopy = headerCopy;
            this.size = size;
            this.extractor = extractor;
        }

        public Iterator<FileView> iterator() {
            return this;
        }

        public boolean hasNext() {
            return index < size;
        }

        public FileView next() {
            if (hasNext()) {
                index++;
                return new FileView(headerCopy.next(), extractor);
            } else {
                throw new NoSuchElementException();
            }
        }

        public void remove() {
            headerCopy.remove();
        }
    }
}

[i30817]

Another, more hairy use of the api (to extract images based on file name (cover))

        try {
            archive = Extractor.from(file);
            if(archive == null)
                return null;

            regex = "(?:.*\\.jpg$)|(?:.*\\.png$)|(?:.*\\.gif$)|(?:.*\\.bmp$)";

            archive.selectByRegex(regex, Pattern.CASE_INSENSITIVE);
            if (archive.selectedSize() == 0) {
                return null;
            }
            //only png, gif etc
            archive = archive.subSelector();
            regex = "(?:.*rear.*)|(?:.*back.*)";
            archive.selectByRegex(regex, Pattern.CASE_INSENSITIVE);
            //not a back cover
            archive = archive.inverseSelector();
            if (archive.isEmpty()) {
                return null;
            } else {
                //ordered by least false positives.
                regex = "(?:.*fcover.*)|(?:.*front.*)|(?:.*cover.*)|(?:^fc\\..*)";
                archive.selectByRegex(regex, Pattern.CASE_INSENSITIVE);
                imageFileView = archive.getSelected();
            }

            if (imageFileView == null) {
                archive.selectAll();
                archive.orderByDescending(Content.NameSize);
                //introduce some fuzziness for this particular search
                //(remove missing or additional "_" or " " or "'")
                Pattern p = Pattern.compile("(?:_| |')");
                String archiveFileName = p.matcher(file.getName().toLowerCase()).replaceAll("");
                for (FileView fv : archive) {
                    String fileName = fv.getFileName();
                    //remove the extension (for .zip and .rar for now so 4)
                    fileName = fileName.substring(0, fileName.length() - 4);
                    //disregard strings too likely to occur in the archive name
                    if (likelySubString(fileName)) {
                        break;
                    } else if (archiveFileName.contains(p.matcher(fileName.toLowerCase()).replaceAll(""))) {
                        imageFileView = fv;
                        break;
                    }
                }
            }
            if (imageFileView != null) {
                imageInput = imageFileView.getInputStream();
                BufferedImage image = ImageIO.read(imageInput);
                if (image != null && imageTooSmall(image)) {
                    image.getGraphics().dispose();
                    return null;
                }
                return image;
            }
        } finally {
            IoUtils.close(imageInput, archive);
        }

[i30817]

No suggestions eh?
:-\

:-X

:’(

:persecutioncomplex: